a, 4 % quehens had 4 vi } f V RW Shey ied Ha ATE EU HB EGED Nie fe srr a ei Vsti {hy taaae ' ; \ wey 1 rane I nga } resp i ! 4,

; , Py Sq rate *) Vahe i 4 Viva ya RO si a Y ; ‘ f RAN SOn Hat tual tal ah Han RN ere ay ay Fan yi tat AY its ; t i qh i i q Hah nie Wa hehthd 2 i Fy F Bios shat Say ; Pe NRA thie af: AACA Gas atl HSM aa freak

fhe

Dacia Pant i Ay iin re CPanel rh 4 f Gone)

Lue fh left 4 eras

aia Pies had Fibs hai Thon i

“

24a hs SALAS pak sd

BARS

=

hs tl

MVeT bent iat

Ne ide meg ind]

Baas) A LysU AG tt mpeg ie Be ae ob i

ized

Sey AWG Tt

eg hay

ea a 1 sti’

iB ty MAPA R eben) ah ft 4454 E

dda Re Cee 13 eth i f ah f s Th a a , EL Te ine el any : RIMES Ee AUPE? Laas et aad tt! ARAL

Pa Hy Piibgdide boa Mead 4 RS Be Pet Pee Wa TN glee

AVS WU 4st By &

blk ded gd ve tik rhe dd yg ay tide tdaae ale na ee ai

OF aM TE Cg FE scone a eal > ARS a ost, « ct nipped aN ED = ye Yo ff i S “2 Wap) = Cape 2 AO 5 ay 2 a) 5 YA = — S Ss as > . ane. ‘ Shines yO ioe) F was WW “”) S WASAly m tings ” m SNOSH w = xe a |

iS saruvadl LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3aluvd

ey m w ” z wn | = S = é 2 Gy & = oe 2 g : 8 Wh? g = i= z E 2, 7 feo ® 2 : Ti iy : : ; : N INSTITUTION NOILALILSNI_NVINOSHLINS _LIBRARI ES SMITHSONIAN INSTI 3 wh Z Z : > ie : “ : z = Yi % a ze C = = Vpoa ie ac & a7 ce for & = am. a = mae 7 es LC Pa a a —t z ih 2 . iS S3IMVYNSIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYY¥ S an 2 to © cS ) . E = = 0 ome = cS a 2 44 a 7m = z Ww = dp) con op) = \N NOILNLILSNI NVINOSHLINS S3IYVYgt LIBRARIES we = = me n O = > Fa

NVINOSHLINS SMITHSONIAN NVINOSHLIWS SMITHSONIAN

a

NS S3IMNVUSIT

Dita Sect oo SMITHSONIAN SMITHSONIAN

SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S314V:

ap) ons WY) : > “ => a ae ” u z ws .o - DB Coe : = rol Z o = ~ Se : : = S e 2 OS = ate Fae = 0) = > Y co

eS S) = Z O = O ie aad a ss) a - rs SNe

AN INSTITUTION NOILNLILSNI NVINOSHLINS SS!1Y¥vudi1 LIBRARIES SMITHSONIAN

- 2 Grae Zz - raga: a ey 2 a ££ ow = Y a a = 2 - Z = pe m 2 m la rm On wn = wn = oD Ss ua

WS SSIYVYUEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS Saly

W c Sais Y Fs ” z w” S Oy X: : S &,: : = Yi Gi oO ao O 23 ye 2) ae Cy ie ON INS 7) YY fi ti ZY" ani aN = E BF as 2 > => - x . Ra) mary hcecrcarll Pa nee, z - = = 2 \) AN INSTITUTION ee LIBRARIES SMITHSONIAN INSTITUT a us = us Ee P Zz a ioe) aN Ww Ea in ud yg P i = : : 2 2 = Uy, ha = < = z yr se mu x TW fas = x oc 4 ; ~ Cc ow Lf Yi, 7. — a m — SAN ~m a aa) ¥ oe ra Ss) se: Oo ies a 2 ~ oO : > Sy Z at z ay a ie WS S3INYVYUgIT LIBRARIES SMITHSONIAN INSTITUTION NOILOLILSNI NVINOSHLINS S3tuve = is = i z ai = hi ° w = Feo) 2 w oO 5 = =e . 5 2 & 5 a Se} = cs Pe) a wv eS E : 2 = 5 = Zz : z Oo z wD ae

IAN INSTITUTION NOILALILSNI NVINOSHLINS S31YVYGI1_ LIBRARIES SMITHSONIAN

:

«

Wb fea LAWN 4 =) ad . z

: ~ : > = -

= AS Ns i id ae “i ca ~ Z i Z z BRARI ES SMITHSONIAN INSTITUTION NOILMLILSNI _NVINOSHLINS SAINVYS17_ LIBRARIES

= a ad = = > .

a B G,

& i : a 2 AN WY 2 = a) = S REND \ ie = = = = ee: HINLILSNI_NVINOSHLINS _LIBRARI ES SMITHSONIAN _INSTITUTION op NOIENLILSNI Pee oS = ; Z |

7) tu Ly tl

ve c = = Gy” =

A Yes oe < Yj pi a <

+ nay tg A a

O a ral is = Ee

2 - Zz ia = Ss 7

INSTITUTION NOILNLILSNI NVINOSHLINS S31uvVyugI] LIBRARIES a rs Se z re z= i Aa Roe = = 2 = oy 2. Salle = > ey 2) ” a oe 2 m S m a | He ai = wn = ” . £ ” ALILSNI 7) ae EIBRARIES SMITHSONIAN INSTITUTION, NO'LNLIISN < = << Ks = & AS =

‘Z eS Oe) Se See

a a wy a SX, 2 Z O SE: Ib

: 8Gye PNK 8 2NX 8 GF

: sme 2 ON: : a 4

Be a es wn eed = bas, Th ees . ARIES SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLINS Sa3luvugi7_“!BRARIES ml : : a & # Z ee. Ve: ae a ANS z =

xt 4%, a < = * SoN < aa

ae ar ad ar BR 02 om

— O — ro) ae O

=! Fr ol = xv =

SA1YVUdIT LIBRARIES SMITHSONIAN” INSTITUTION _ NOILONLILSNI

‘a ell & Pia Fe =

on = as ) — Ss)

: ay 10.0) — OJ ra

a om = = = es

i = et a ez 2

ARIES SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLINS SaluvuaIT_ LIBRARIES z a * Ww z= ~ j = 7

~ ea La ze

OS a OC “o£ O =

= = = e Z =

2 = ca = = . SQ

—NVINOSHLINS _LIBRARIES SMITHSONIAN INSTITUTION , NOILNLILSNI

D : a us 2 :

a: i = 2 = ac

of , me

re Ss = < es <

of = oc bar 2

> © - m. ue. m

O os Oo O a

Way 2 see ail 2 ae Ey IES ZSMITHSONIAN INSTITUTION NOILM.LILSNI” NVINOSHLINS ~ $3 lyvugit at! BRARIES oe = ° = S = iE = a % e* x rs 70 = a 7p) wn” = 2 a Zz Bs Z .

CS Se Aa an Br ALE ew

eal ad ee OITA RK mi aial aK ipo aemeia Kniss a uecae RIMLTITOrtriaatl

i ; Soe a Ey ’ j ; a an) ii Ain j i aan t ie bis OR A Sib aby BL ia aun eit

5

UAT

FEO, OTe

WATSONIA

JOURNAL AND PROCEEDINGS OF THE BOTANICAL SOCIETY OF THE BRITISH ISLES

VOLUME 12

EDITED BY

S. M. COLES, G. HALLIDAY, N. K. B. ROBSON, C. A. STACE AND D. L. WIGSTON

1978-79

PRINTED FOR THE SOCIETY BY Willmer Brothers Limited, Birkenhead

PUBLISHED AND SOLD BY THE BOTANICAL SOCIETY OF THE BRITISH ISLES C/O DEPARTMENT OF BOTANY, BRITISH MUSEUM (NATURAL HISTORY), LONDON SW7 5BD

DATES OF PUBLICATION

Part 1, pp. 1-80, 10th February, 1978

Part 2, pp. 81-200, 28th September, 1978 Part 3, pp. 201-296, 22nd February, 1979 Part 4, pp. 297-400, 6th September, 1979

Printer’s Error

p. 63 line 18 should read:

_. . Luzula sylvatica, a dominant of the ground flora. An area of the South Wood fenced by the Nature. . .

S&

Botanical Society of the British Isles

Patron: Her Majesty Queen Elizabeth the Queen Mother

Applications for membership should be addressed to the Hon. General Secretary, c/o Department of Botany, British Museum (Natural History), Cromwell Road, London, SW7 5BD, from whom copies of the Society’s Prospectus may be obtained.

Officers for 1977-78

Elected at the Annual General Meeting, 7th May 1977 President, Professor D. H. Valentine

Vice-Presidents, Mrs B. H. S. Russell, Dr C. T. Prime, Mr J. P. M. Brenan, Mr J. F. M. Cannon

Honorary General Secretary, Mrs M. Briggs Honorary Treasurer, Mr M. Walpole

Honorary Editors, Dr S. M. Coles, Dr G. Halliday, Dr N. K. B. Robson, Dr C. A. Stace, Dr D. L. Wigston

Honorary Meetings Secretary, Mrs J. M. Mullin Honorary Field Secretary, Miss L. Farrell

Honorary Membership Secretary, Mrs R. M. Hamilton

All enquiries concerning sales of, or advertising space in, the Society’s publications should be addressed to D. H. Kent, 75 Adelaide Road, West Ealing, London, W13 9ED

Watsonia, 12, 1-14 (1978). ]

Morphological variation in the Vicia sativa L. aggregate

E. HOLLINGS 12 Stivichall Croft, Coventry and C. A. STACE

Botanical Laboratories, University of Leicester

ABSTRACT

Wild and cultivated strains of Vicia sativa L. agg. (incl. V. angustifolia L. and V. segetalis Thuill.) and V. lathyroides L. have been examined for 50 primary and many other derived morphological characters. None of the characters commonly used to differentiate the various segregates recognized in the literature or by us was found to be significantly affected by environmental factors, but several of them (e.g. flower-length, pod-length, absolute leaflet- shape) show continuous variation which renders them of very little taxonomic value. Characters such as heterophylly, tendril form, petal colour, pod colour, seed conspicuousness in the pod and seed texture may be used to separate four main taxa: V. lathyroides, V. angustifolia, V. segetalis and V. sativa.

The taxonomic status of these will be discussed in a later paper dealing with breeding and hybridisation data.

INTRODUCTION

Vicia sativa L. agg. is widespread over much of Europe, western Asia and North Africa as a native plant and, due to its frequent cultivation for green manure and fodder, it has become naturalized over an even wider area of the northern hemisphere.

The taxonomy of this species aggregate, and indeed its distinction from some close relatives, has been confused ever since Linnaeus’ time. Much of this confusion is due to the inbreeding which is predominant in the aggregate and, in more recent times, to the large number of cultivated strains which have been bred and then grown on a field scale, and which have subsequently escaped from cultivation and intermingled with the native populations. The work described here is part of an attempt to place the taxonomy of the aggregate on a more logical basis than hitherto, with particular reference to the taxa occurring in the British Isles and adjacent parts of the Continent.

Vicia sativa agg. may be separated from all other European species of the genus by its annual habit; 1- to few-flowered, shortly pedunculate inflorescences; leaves with c 3-8 pairs of leaflets and a terminal, usually branched tendril; glabrous, usually reddish-purple corollas; equal calyx-teeth; and smooth seeds with a hilum occupying } to + the circumference.

Its closest relative, at least in western Europe, is V. lathyroides L., which is smaller in all its parts, has leaves with usually fewer leaflets and unbranched tendrils, and has tuberculate seeds. This species is, however, very often confused with small plants of V. sativa, particularly in the field, and for this reason it has been included in the present survey.

V. lutea L., another closely related species, differs from V. sativa in its unequal calyx-teeth; yellow corollas; strongly pubescent, deep pods; and seeds with a hilum § to 4 the circumference. Variants of V. sativa with white or yellow corollas occasionally occur and have sometimes been misidentified as V. lutea, but the calyx, seeds and pods of the latter species are distinctive.

At maturity most individuals of V. sativa agg. are much-branched, sprawling or climbing plants

2 E. HOLLINGS AND C. A. STACE

bearing 1- to 2-flowered inflorescences in most of the upper leaf-axils. The shape of the leaflets on the lower leaves is often markedly different from that of those on the upper leaves which bear flowers in their axils; plants with this character well developed are termed heterophyllous. The leaves on the primary shoot of the young seedling are usually different again from any of those on the floriferous shoots, both in the shape of the leaflets and in the possession of a simple tendril or in lacking a tendril. The primary shoot never bears flowers; these are formed on the lateral shoots which develop from the lowest few nodes of the primary shoot, and on similar secondary or higher-order laterals. The overall vegetative vigour of a plant is an expression not only of the size of each shoot but of the numbers and orders of the laterals, which soon over-top the short-lived primary shoot.

In many cases the production of high-order laterals continues throughout the growing season. In certain plants very short laterals are formed low down near the ground, late in the season, when most of the older laterals possess ripe or ripening fruits. These short laterals bear small leaves and often very small flowers, which may be cleistogamous and give rise to small, few-seeded pods. In yet other plants, laterals produced very close to ground level become subterranean and develop as whitish shoots with rudimentary leaves and minute, whitish, cleistogamous flowers. The latter form whitish-brown, 1- to 2- seeded, subterranean pods, often in great quantity, which are thus conspicuously different from the earlier epigeal ones and contain seeds often also distinguishable from those formed above ground. — Subterranean seed-production 1s typical of the taxon known as V. amphicarpa Dorthes, and we shall — use the term amphicarpy to describe the feature. bh

V. lathyroides is a less vigorous plant with usually far fewer branches and less pronounced differences in leaf-shape between main and lateral branches. Cleistogamy and amphicarpy have not been observed and, moreover, the primary shoot does on a few occasions bear one or two flowers. Most of these differences are probably related to the ephemeral life-cycle of this species.

PREVIOUS CLASSIFICATIONS OF VICIA SATIVA AGG.

In the past V. /athyroides has almost always been recognized as a species distinct from V. sativa agg. In British Floras the latter has either comprised a single species with or without subspecies and varieties, or two species, V. sativa sensu stricto and V. angustifolia L., which differ in size, leaflet-shape, flower- colour and fruit-shape and -colour. The former treatment was adopted, for instance, in Flora Anglica (Hudson 1762), English botany (Syme 1864) and Flora of the British Isles, 2nd ed. (Tutin 1962); the latter in The English flora (Smith 1825), The British flora (Hooker 1830) and Flora of the British Isles, Ist ed. (Tutin 1952). V. angustifolia, whether treated as a species or a subspecies of V. sativa, was frequently separated into two varieties usually known as var. angustifolia and var. bobartii (E. Forst.) Koch (e.g. Tutin 1952). The former variety is also known as var. segetalis (Thuill.) Koch, and it appears, in fact, that var. bobartii is the type variety and hence the one to be known as var. angustifolia. Var. segetalis is a more robust plant than var. bobartii, and in several characteristics falls somewhere between the latter and V. sativa.

Continental authors have shown a similar degree of variation in their treatment of V. sativa agg., but because of the greater number of variants found in southern Europe the situation is much more confused. Rouy (1899), for example, recognized six distinct taxa in the aggregate, which he called V. communis Rouy. These taxa were termed ‘formes’, intermediate in rank between subspecies and variety, and among them a further ten varieties and three sub-varieties were recognized. Ascherson & Graebner (1909) divided V. sativa into four subspecies, which were largely equivalent to Rouy’s formes (although two of the latter were reduced to synonymy), but in addition they recognized a very large number of varieties and other infra-specific categories.

The most important taxonomic studies of recent years have been made by Mettin & Hanelt (1964), who divided the aggregate into six species: V. sativa, V. angustifolia, V. amphicarpa, V. incisa Bieb., V. cordata Wulfen ex Hoppe and V. macrocarpa (Moris) Bertol. Later (Hanelt & Mettin 1966) they added a seventh species, V. pilosa Bieb. The first two of these are the two species which have been recognized by British authors, and the second of them was divided by Mettin & Hanelt into subsp. angustifolia and subsp. segetalis (Thuill.) Gaud.

Ball (1968) based his treatment for Flora Europaea on the work of Mettin & Hanelt, but preferred to consider the segregates as subspecies of V. sativa. Under this scheme V. angustifolia must be called

MORPHOLOGICAL VARIATION IN THE VICIA SATIVA L. AGGREGATE 3

subsp. nigra (L.) Ehrh. Moreover, Ball included V. pilosa with it, so that he recognized a total of only six subspecies. They are: subsp. sativa, subsp. nigra, subsp. amphicarpa (Dorthes) Asch. & Graeb., subsp. cordata (Wulfen ex Hoppe) Asch. & Graeb., subsp. incisa (Bieb.) Arcangeli and subsp. macrocarpa (Moris) Arcangell.

Much important work on this aggregate has also been carried out by Yamamoto (e.g. Yamamoto 1966). The results of his breeding experiments have important taxonomic implications but Yamamoto has not sought to use these to propose modified systems of classification.

Plitmann’s (1967) valuable contribution on Middle-Eastern species of arinual Vicia includes a careful survey of V. sativa agg. Init he, like Ball (1968), recognized the main variants as subspecies of V. sativa, but subsp. pilosa is not treated (not occurring in the region concerned) and subspp. cordata and incisa are amalgamated as a single subspecies. Moreover, in the main text, V. segetalis is placed with subsp. cordata rather than with subsp. angustifolia (= nigra), although a ‘corrigendum’ (p. 128) reverses that decision. In the Flora of Turkey, Davis & Plitmann (1970) recognized the same five subspecies.

Most previous systems of classification of the Vicia sativa aggregate have been based on a small, subjective selection of characters, both quantitative and qualitative, with little regard to their mode of variation. Some of the most frequently used characters, notably flower and pod length, are in fact continuous variables and the points in the scale of variation used by different authors to separate the taxa has varied a great deal (cf. Figs. 2 & 3). They have probably been selected for use because they are — readily observable on herbarium material, whereas many other characters (often taxonomically more useful) are usually poorly preserved or demand floral dissection.

The work reported here has, therefore, sought to determine the pattern of variation in the morphological characters, as well as their points of discontinuity (if any) and the existence of any correlation between them, and has attempted to ascertain to what extent any groups so defined correspond with the segregates which have been recognized in the past.

For convenience we shall in this paper use the classification (i.e. the delimitation of the segregates and their nomenclature) of Mettin and Hanelt, but in addition shall recognize V. segetalis Thuill. as a distinct species. It is in several ways as distinct from V. angustifolia as are some of the other species and, in the British Isles at least, it has in fact been confused with V. sativa more frequently than with V. angustifolia. It is likely that only V. angustifolia of the above species is native in northern Europe, including the British Isles, but V. sativa and V. segetalis, and to a lesser extent V. cordata and V. amphicarpa, are cultivated there and occur as escapes from cultivation. The last two, however, have not to our knowledge been found wild in the British Isles. The decline of V. sativa as a cultivated plant in the British Isles has been documented by Killick (1975). We shall not be concerned with the other three taxa (V. pilosa, V. incisa and V. macrocarpa).

The main characters which may be used to separate these five segregates of V. sativa agg. and V. lathyroides are best summarized in the following key:

1 Seeds tuberculate .. B 3 a ot a a BH i z V. lathyroides 1 Seeds smooth 2 Plants with subterranean stems bearing cleistogamous flowers and whitish,

few-seeded pods * * - “A Le Mi ». eS vs V. amphicarpa 2 Plants without subterranean stems 3. Pods constricted between seeds .. i. ae is v3 a bes i V. sativa 3. Pods not constricted between seeds 4 Calyx-teeth longer than calyx-tube .. fe bg ie ie ie i eccordata 4 Calyx-teeth shorter than calyx tube 5 Plant markedly heterophyllous .. oe Ki * a i V. angustifolia

5 Plant not or scarcely heterophyllous " Be at ne se .. V. segetalis

- E. HOLLINGS AND C. A. STACE

MATERIALS AND METHODS

Seed samples of wild, known origin were collected personally or by colleagues, or were obtained from the seed exchange schemes of various botanic gardens. The 45 samples so gathered are listed with their chromosome numbers in Hollings & Stace (1974, Table 2). In addition, a large number (c 160) of samples of cultivated material was used, mostly of agricultural origin from various seedsmen and research stations. We feel this is justified because, as previously stated, many of the ‘wild’ populations found in the British Isles today originate as escapes or relics from such sources. For most of the work only about 30 or 40 of the cultivated strains were utilized. Since these plants are annuals, fairly regular re-sowing and seed harvest had to be carried out, although seed viability can last for several years in laboratory conditions. Representatives of all samples used were grown under field conditions and specimens were dried and pressed to provide a comprehensive herbarium. Flowers were more difficult to preserve adequately as they fade and wither very rapidly. Freshly opened flowers were therefore dissected into their component organs and spread out flat on the adhesive surface of transparent, self- adhesive tape, which was then inverted and mounted on stiff cards. These preserve the floral organs indefinitely for measurement and examination.

Measurements were standardized as far as possible, particularly in relation to the part of the plant sampled. Floral measurements are means from at least 10 freshly open flowers; leaf and fruit characters are means of 20 measurements. Seeds were measured with the aid of a Vernier travelling microscope or Vernier screwgauge. Where possible, leaf measurements were taken from the lowest flowering node, unless otherwise noted; stipules were scored from the two nodes below the lowest flowering node (where they are best developed).

Optical densities of wing and standard petals were taken from known weights of tissue (approx. 0-5 g) extracted in 5 ml of 1% methanolic HCl. Optical density at 530nm was determined for each sample using an Eel Spectrometer and corrected to a value for 1g of tissue. 1% methanolic HCl was used as a blank standard.

Cultivation experiments were carried out in crowded (10 plants per 5 inch pot) and isolated (1 plant per 5 inch pot) conditions, in constant (15°C, 16h day, 40% rel. humidity) glasshouse or changing outdoor conditions, and in four types of soil (sand, loam, clay, John Innes No. 1 compost). The cultivation experiments therefore comprised 16 different treatments. The soil types were prepared with regard to texture only; in particular it should be noted that the sand was low in calcium and thus resembled the soil of inland sandy heaths rather than that of most maritime dunes.

DEFINITION OF CHARACTERS INVESTIGATED

The characters chosen for this survey included most of those which had been used taxonomically in the past, plus many which clearly varied and might therefore be of taxonomic value. From this basic list many additional characters were derived, either ratios or measurements from different parts of the plant. The list is clearly by no means exhaustive, as the measurements of a flower such as that of Vicia can be extended almost indefinitely. Moreover, many other, mainly qualitative, characters were omitted from the detailed study, e.g. detailed patterns of testa coloration, presence of a blackish spot (extra-floral nectary) on the stipules, as well as a number of very variable quantitative features, e.g. plant height. The following 50 primary characters were scored, 37 being quantitative and 13 qualitative.

_a. Flowers: 1. Flower length, from base of calyx to apex of lateral petal _ 2. Number of flowers per node 3. Cleistogamy, as one of three categories: cleistogamous flowers absent; aerial cleistogamous flowers present; subterranean cleistogamous flowers present br Cabyx:; 4. Calyx-tube length, from base of mid-ventral tooth to base of calyx-tube 5. Calyx-tube width, across base of calyx-teeth of opened-out calyx 6. Midventral calyx-tooth length, from base to apex along mid-vein 7. Midventral calyx-tooth width, across base

MORPHOLOGICAL VARIATION IN THE VICIA SATIVA L. AGGREGATE 5

c. Standard petal: 8. Petal length, from base of claw to apex of limb along mid-line 9. Petal-limb length, from apex of limb to its junction with claw along mid-line

10. Petal-limb width, across widest point

11. Petal-claw length, from base to its junction with limb

12. Petal-claw apical width, at its junction with limb

13. Petal-claw greatest width, at widest point

14. Petal-notch depth, from tip of petal-lobes to base of notch

15. Optical density at 530nm of pigment from 1g tissue in 5ml 1% methanolic HCI d. Lateral petals:

16. Petal length, from base of longer claw to apex of limb

17. Petal-limb length, from apex of limb to its junction with claws

18. Petal-limb width, across widest point

19. Long petal-claw length, from base to its junction with limb

20. Short petal-claw length, from base to its junction with limb

21. Optical density at 530nm of pigment from 1g tissue in Sml 1% methanolic HCl e. Keel:

22. Keel-pouch length, along greatest length parallel to claws

23. Keel pouch depth, from dorsal opening to ventral suture

24. Keel-claw length, from base to junction with pouch

25. Keel-pouch pigmentation extent, from dorsal opening to further edge of dark pigmentation f. Androecium:

26. Stamen-tube length, along mid-ventral line

27. Stamen-tube width, across widest point when opened out

28. Free filament length, from its junction with stamen-tube to its junction with anther g. Gynoecium:

29. Ovary length, from base to its junction with style

30. Ovary width, across widest point

31. Style length, from its junction with ovary to base of style h. Pod:

32. Pod length, from its junction with pedicel to base of beak

33. Pod depth, from dorsal to ventral suture

34. Pod colour, as five categories: yellow; yellow-brown; brown; brown-black; black

35. Seed conspicuousness, as smooth or moniliform pods

36. Number of seeds per pod

37. Pod pubescence, as glabrous or pilose i. Seeds:

38. Seed length, across widest point parallel to hilum

39. Seed depth, across widest point from dorsal (hilum) to ventral side

40. Seed width, across widest point at right angles to both 38 and 39

41. Testa colour, as five categories: white; yellow; green; brown; black

42. Testa mottling, as mottled or unmottled

43. Testa texture, as smooth or tuberculate j. Leaves:

44. Leaf length, from base of petiole to point of attachment of apical leaflets

45. Leaflet number

46. Leaflet length, from base of apex along midrib, including mucro if present

47. Leaflet width, across widest point

48. Leaflet-apex shape, as five categories: acute; obtuse; truncate; truncate-emarginate; emarginate (Fig. 1)

49. Stipule-tooth number, as mean number of primary teeth per stipule

50. Tendril type, as short and simple or long and branched.

Of the various ratios derived from these primary characters only the heterophylly index, designed to measure the difference in leaf-shape between leaflets on upper and lower leaves, requires further explanation. The heterophylly index is defined as the ratio of average leaflet ratio of the leaf on the eighth node/average leaflet ratio of the leaf on the first node of the first lateral shoot, where average leaflet ratio is the ratio of average leaflet length/average leaflet width.

6 E. HOLLINGS AND C. A. STACE

Acute Obtuse Truncate

Truncate - emarginate Emarginate

FiGurRE 1. Terms used to describe leaflet-apex shapes

RESULTS

MORPHOLOGICAL CHARACTERS As expected, no single character, either primary or derived, affords complete separation of the six taxa defined in the Introduction.

The quantitative characters fall into two main categories, with continuous and discontinuous distributions. In the former case the histograms obtained mostly show normal or near-normal distributions (Fig. 2), although sometimes slightly skewed, but a few show evidence of multimodal] curves (e.g. flower length, Fig. 3). Some of these probably indicate artificial clustering, caused by the rounding up and down of figures, but others are the results of separate curves, with different modes from the various taxa, superimposed on each other. These separate, unimodal curves, however, overlap greatly and such characters are not alone reliable for delimiting taxa, despite the fact that they have been much used by previous authors. In fact they merely serve to illustrate an increase in size of most vegetative and reproductive parts across the aggregate, from V. /athyroides at the lower end of the scale to V. sativa at the upper.

The only quantitative characters showing truly discontinuous distributions are two ratios: optical density of lateral petals/optical density of standard petal (characters 21/15) (Fig. 4), and heterophylly index (Fig. 5). A low optical density ratio (1/8 to 1/13) indicates a conspicuous difference between the pigmentation of the lateral (dark) and standard (light) petals, while a high one (1/1 to 1/4) indicates little difference. A high heterophylly index (4—5S-5) represents a plant with relatively broad leaflets on lower leaves and narrow leaflets on upper (flowering region) leaves, and a low index (0-5—2-:5) represents a plant with relatively little change in leaflet shape. Because of the disjunction of both of these characters, each separating the aggregate into two groups, they can be equally expressed qualitatively as concolorous and bicolorous flowers and as isophyllous or heterophyllous stems, although we do not claim that with wider samples than ours intermediates would not be found in both characters.

Many quantitative characters, especially various floral dimensions and seed measurements, are strongly correlated. Such straight-line relationships illustrate the fairly constant shape of these organs irrespective of absolute size or of the segregate, and are particularly well-marked in characters of the gynoecium, androecium and keel (Fig. 6), probably due to the importance of maintaining the critical floral dimensions needed in the floral biology of the aggregate. They are therefore of little taxonomic use.

MORPHOLOGICAL VARIATION IN THE VICIA SATIVA L. AGGREGATE

“SBIO]] 9AIVRUISIId9I 99IY} Ul

USAIS (JUS) DAYS “4 puR (oI) VIOfisnsuD “4 JO SdduUvI dy) JUSSoIdd1 9SOY} MOjOQ SOUT] UY OY *.YVduvo1ydwuDy "A, JOpUN SIOMO]J SNOUTKZO}JSIgIO Sulonpoid sures |e popnypour ssodind oy} JOJ sAvY 9M je) 1d99x9 (P96) JouRH 2 Une Aq poiujap se (saul] UdyYO1g Ul SosUvI [VUOIdIOXd) So} eS9ISIS XIS 9Y} JO YOVa Aq poygiyXa [e193] PUL INO UI S9BURI 9}BSIPUI WILIZOJSIY 9Y} MOJ9q SOUT] YOIY) OUT, “38e VAIS 1914 JO SyBUd| pod Jo WeIdOsIH{ “7 AUNOIA

‘7 dINBIL Ul SB WIeRISOSTY OY} MOTO SOUT] UTY) puv Yor TL, ‘BBR DANDS DIIIA JO SYISUI| JOMOTJ JO WRISOISIH “¢ AUNOL{ (S896L) IIEd (2961) UlFNL (2S6L) UIFNL

s1jeyobas

2}Epuod

eduesiydwe

eljojizgnBue SopiodAy}e| ————

wuuw «ul YuybuUZ] YaMo|4 OSe8G.0¢5re ce, Og. Sl Ol vi cd Ol S 9

00

N =

ADUDNDII4 %o

©

(8961) IIEd (2961) UNL (cS6b) UlFNL

sijeyobas

e}EPuoD

eduesiudwe

e1oylysnBbue

soplouAyie|

lWuus =ul UuHua pod O OV O€& Og

ADU2NbII4 %

8 FE. HOLLINGS AND C, A. STACE

—

fo)

Ol

. lo Frequency

4 6 8 10 12 14 Ratio of optical densities

Ficure 4. Histogram of ratio of optical densities of lateral and standard petals in Vicia sativa agg. See text for definition.

30

N O

%o Frequency

fo)

1 2 3 a S| 6 Heterophylly index

FIGURE 5. Histogram of heterophylly index in Vicia sativa agg. See text for definition.

14

in mm a= NO

fo)

length

Ovary 0)

O 2 4 6 8 10 12 14 16 Keel length (pouch & claw) in mm

FIGURE 6. Scatter diagram of ovary length/keel length in Vicia sativa agg.

MORPHOLOGICAL VARIATION IN THE VICIA SATIVA L. AGGREGATE 9

Leaflet- apex

O acute @ 144 @ obtuse @ truncate e® ge @ truncate - emarginate @ £ 12] @ emarginate * =i) = td : o = Q 2 2 8 oe iS 0%, o ) 9 6 Q 2 2-0 ’ “88 0 D i goac® c (OyeO) > e) O 4 8 12 16 20 24 28 32

Average leaflet length in mm

FIGURE 7. Scatter diagram of leaflet width/leaflet length in Vicia sativa agg.

On the other hand, most pod and leaf dimensions are rather poorly correlated, indicating that leaflets (Fig. 7) and pods vary greatly in shape. These variations are not closely correlated with the characters used to define the segregates, and are thus no more useful for delimiting them than are the absolute dimensions.

Certain characters are, however, correlated with the two discontinuous variables (heterophylly index and optical density ratio) and serve to further discriminate between separate groups of plants. Plants with a high heterophylly index have mostly smaller pods (Fig. 8), seeds and leaves and fewer stipule- teeth than plants with a low heterophylly index. Furthermore, whereas isophyllous plants exhibit a wide range of flower length, strongly heterophyllous plants fall into two groups with respect to this character: one with flowers 6-9mm and one with flowers 14-19mm long (Fig. 9). Fig. 9 also shows that strongly heterophyllous plants possess concolorous petals, while isophyllous plants possess bicolorous ones. The other floral character which has been used by several previous workers in the aggregate, ratio of calyx-tooth length/calyx-tube length, is not in itself a very useful discriminant (Fig. 10) and is not very well correlated with other characters. For example, although plants with a high heterophylly index all have calyx-teeth/calyx-tube ratios equal to or less than 1-0 (100 in Fig. 11), plants with a low heterophylly index may have ratios well below or well above 1-0 (Fig. 11). The calyx- teeth/calyx-tube ratio is also not well correlated with any of the important qualitative characters such as smooth or moniliform pods.

Many of the qualitative characters, both those systematically scored and many more not so, exhibit a reticulate-type variation, obeying Vavilov’s (1951) Law of Homologous Series. Thus they are found in almost infinite combination with other such characters, and show extensive parallels between related taxa, so that they are of very little value in orthodox taxonomy. Characters of this type are testa colour, patterns of testa mottling, albinism, pubescence, and presence or absence of a black spot on the stipules.

A number of qualitative characters are, however, fairly closely correlated with each other and with various quantitative characters. Plants with tuberculate seeds mainly have small, dark, smooth pods; small, concolorous flowers; a high heterophylly index; and simple tendrils. The rest of the aggregate

index

Heterophylly

index

Heterophylly

25

30

O pod yellow @® pod yellow- brown @ pod brown @ pod brown- black @ pod black

© flower bicolorous. O testa smooth ©O pod smooth

35 40 45 50 55 60 65 70 75

Pod length in mm

FiGure 8. Scatter diagram of heterophylly index/pod length in Vicia sativa agg.

©

Jog

Sooe

O pod yellow

d @ pod yellow-brown 6 fo) @ pod brown @ pod brown-black o a é @ pod black

O flower bicolorous

O testa smooth O-testa tuberculate © pod smooth

OQ pod moniliform O tendril branched -O tendril simple

@ Pe @ 3 20 eo, ° © @ Q 2 20 2? @ rs & @2@ ae & e). @ a? So C0 . @ @

29 doa 280" 28 aa0%8 020% & ‘ ea°e ase ee °° @

10

15 20 25 Flower length in mm

FiGuRE 9. Scatter diagram of heterophylly index/flower length in Vicia sativa agg.

MORPHOLOGICAL VARIATION IN THE VICIA SATIVA L. AGGREGATE 1)

20

16

12

No. specimens

60 70 80 930 100 =: 110 120 130 (ZOR DO... 160 Ratio of calyx- tooth/ calyx-tube length x100

SO

FiGure 10. Histogram of ratio ( x 100) of calyx-tooth length to calyx-tube length in Vicia sativa agg.

index

Heterophylly

50 60 70 80 90 NOL wake 120 ISOR. 40s, 150) | 160 Ratio of calyx-tooth/calyx- tube length x100

FiGuRE 11. Scatter diagram of heterophylly index/ratio ( x 100) of calyx-tooth length to calyx-tube length in Vicia Sativa agg.

have smooth seeds and (usually) branched tendrils. Pods which are moniliform are usually also pale- coloured and pilose; whilst those which are smooth are usually glabrous and dark-coloured, although dark, moniliform pods and pale, smooth pods do occur. The shape of the leaflet apex 1s to some extent correlated with leaflet width, acute and obtuse apices being confined to narrow leaflets, but emarginate apices are found on very narrow as well as very broad leaflets, and leaflet apex shapes are not well correlated with the characters usually used to separate the segregates.

Cleistogamous underground flowers, supposed to be indicative of V. amphicarpa, occur on only one of the living strains studied (from Turkey). On this plant the aerial flowers are also small and cleistogamous, which is not the case in most plants of V. amphicarpa described in the literature. Other strains, with smooth or moniliform and pale or dark pods, with bicolorous or concolorous, large or small flowers, and with isophyllous or heterophyllous stems, produced aerial cleistogamous flowers,

12 E. HOLLINGS AND C: A. STACE

either on the main laterals or on short laterals formed low down late in the season, or produced subterranean whitish branches with flower primordia, which presumably would have later produced subterranean pods. We therefore conclude that neither cleistogamy nor amphicarpy, being scarcely correlated with other characters, are very useful taxonomic discriminants, at least of major taxa.

The best separation of the variants is effected by plotting heterophylly index against flower length, together with a number of qualitative characters (Fig. 9). This demonstrates three main groups of plants:

1. Plants with high heterophylly index; simple tendrils; concolorous flowers; calyx-teeth shorter than calyx-tube; small, smooth, dark pods; and tuberculate seeds. These correspond with V. lathyroides. 2. Plants with high heterophylly index; branched tendrils; larger (14-19mm) concolorous flowers; calyx-teeth shorter than calyx-tube; rather small, smooth, dark pods; and smooth seeds. These correspond with V. angustifolia.

3. Plants with low heterophylly index: branched tendrils; bicolorous flowers; larger pods; and smooth seeds. These comprise: V. sativa, V. segetalis, V. cordata and V. amphicarpa. These vary in the other characters mentioned under groups | and 2, but on the basis of qualitative characters two sub-groups may be recognized: with pale, moniliform, often pilose pods (corresponding to V. sativa); and with dark, smooth, usually glabrous pods (corresponding to V. segetalis and V. cordata). As mentioned above, V. amphicarpa, as usually defined (on the basis of amphicarpy), cannot be accommodated in this system of grouping.

CULTIVATION EXPERIMENTS

Although many of the qualitative characters are known to be genetically determined, the large number of quantitative characters which help distinguish the variants seemed very likely to be affected by environmental factors, especially as the number of variants and the breadth of variation shown by each in any one locality is small. Moreover, sandy heaths and sand-dunes support the smaller, weaker variants, particularly V. lathyroides and V. angustifolia, whilst richer inland soils often support the more luxuriant variants.

Nevertheless, preliminary observations, gained from cultivation of many wild strains in standard conditions, suggested that habitat has little effect on morphology, and Blum (1966) showed that plant density scarcely changed pod and seed size, although it affected the production of lateral shoots.

The results of cultivation experiments on eight selected strains covering the segregates V. sativa, V. segetalis, V. angustifolia and V. lathyroides were remarkably consistent. In all cases plants were more robust and produced more laterals in isolated than in crowded, and in outdoor than in glasshouse, conditions. They were also more robust in potting compost than in loam or clay, and less robust still in sand. Differences were mostly statistically highly significant and were of similar magnitude for all segregates.

On the other hand, detailed analysis of 16 quantitative and qualitative characters, including almost all those considered taxonomically important, in no cases showed any significant differences between treatments. Qualitative characters were completely unchanged and quantitative measurements all fell well within the range expected from previous observations. In particular, environmental factors appear to have negligible or no effect on flower, pod and leaf measurements; heterophylly index; flower, seed and pod colour; seed texture and conspicuousness in pod; and tendril branching.

CONCLUSIONS AND DISCUSSION

The cultivation experiments show that the complex pattern of variation found in V. sativa agg. is not the result of phenotypic responses to different habitats but is genetically determined.

The quantitative characters mostly vary in such a way that the four British segregates, V. lathyroides, V. angustifolia, V. segetalis and V. sativa, differ by increasing size of most of their parts (in the order of taxa given). However, there is a wide measure of overlap so that these characters are not alone sufficient for distinguishing the segregates.

In particular, flower length and pod length have been much used in the past as discriminants. Figs. 2 and 3, besides providing a breakdown of the ranges of variation shown in our material by the individual

MORPHOLOGICAL VARIATION IN THE VICIA SATIVA L. AGGREGATE 13

segregates, also illustrate, by reference to three selected treatments, the very variable interpretation which has been placed on these characters by past workers. It should be noted that, in these selected treatments, V. segetalis was nomenclaturally contained in V. angustifolia, but that a proportion of V. segetalis plants was probably included under V. sativa. Flower length alone may be useful in distinguishing V. /athyroides from the other taxa, except that cleistogamous, late-season flowers of the latter frequently fall into the size range of normal V. lathyroides flowers. In general, vegetative characters, particularly absolute measurements of leaves and the shape of the leaflet apex, are of little taxonomic value.

Two quantitative characters which appear to be good discriminants, heterophylly index and colour difference in standard and lateral petals, may for convenience be scored qualitatively, as explained previously. These, and certain other qualitative characters, notably seed conspicuousness, pod colour and seed texture, are the best diagnostic characters in the aggregate and can be used to separate the four segregates. It is arguable that this effects a very artificial separation, in that a few discontinuous characters have been carefully chosen from a much greater number of continuous variables, but it is the taxonomist’s aim to search for discontinuities and, where they are found, express them in taxonomic terms by naming the discrete units so separated.

V. cordata is usually separated from V. angustifolia and V. segetalis on the basis of the calyx- teeth/calyx-tube ratio and chromosome number, V. cordata having 2n = 10 and a ratio over 1:0(100 in Figs. 10 & 11), the latter two having 2n = 12 and a ratio under 1-0. However, in a previous paper, we (Hollings & Stace 1974) failed to find precise correlation between these two characters, and, moreover, plants of V. sativa also have a calyx-tooth/calyx-tube ratio varying from below to above 1:0. Thus on morphological features alone V. cordata is scarcely worthy of recognition. On the basis of our observations, plants corresponding with V. cordata can be quite easily accommodated in V. segetalis, which agrees well with the preliminary conclusion of Plitmann (1967), but not with his later decisions.

V. amphicarpa, delimited simply by the presence of subterranean pods originating from cleistogamous flowers, is otherwise very variable both in morphology and in karyotype (Hollings & Stace 1974). The recognition by Rouy (1899) of two varieties, var. pseudosativa and var. pseudoangustifolia (amphicarpous variants of V. sativa-like and V. angustifolia-like plants respectively) reinforces the opinion that ‘V. amphicarpa’ as normally understood is not a taxon but a series of variants of other taxa possessing only amphicarpy in common.

The four segregates revealed by our studies may be distinguished by the following diagnoses:

V. lathyroides: Strongly heterophyllous; tendrils simple; flowers 6-9 mm, concolorous (usually dull purple); pods 18-30 mm, brown to black, smooth, glabrous; seeds tuberculate.

V. angustifolia: Strongly heterophyllous; tendrils branched; flowers 14-19 mm, concolorous (usually bright pink); pods 23-38 mm, brown to black, smooth, glabrous; seeds smooth.

V. segetalis: More or less isophyllous; tendrils branched; flowers 9-26 mm, bicolorous (with standard petals paler than lateral petals); pods 28-70 mm, brown to black, smooth, usually glabrous; seeds smooth.

V. sativa: More or less isophyllous; tendrils branched; flowers 11-26 mm, bicolorous (with standard petals paler than lateral petals); pods 36-70 mm, yellowish to brown, moniliform, often pilose; seeds smooth.

As stated in the Introduction, it is likely that only the first two of these are native in the British Isles.

On the basis of morphological characters there is not a strong argument for separating V. /athyroides from the rest of this aggregate. However, there are differences in the suspected pattern of karyotype evolution of V. /athyroides from that of the rest (Hollings & Stace 1974), and there is a strong breeding barrier between it and the other taxa (Hollings & Stace, unpublished).

V. angustifolia, V. segetalis and V. sativa are not genetically isolated, and their rather complex pattern of variation is almost certainly the result of extensive inbreeding and, in this country at least, of the cultivation and subsequent escape of many non-native variants. These segregates (as well as others such as V. macrocarpa, V. pilosa and V. incisa) therefore represent groups of variants representing modes in a spectrum of largely continuous variation. Although they are usually separable by the characters given above, plants of intermediate status, or with one or two anomalous characters, are not

14 E. HOLLINGS AND C. A. STACE

-uncommon. Indeed they are to be expected from an application of the Law of Homologous Series, which proves to be a useful concept in this group of plants and applies to karyotype variation as well as to morphological characters. The most appropriate rank for the main segregates will be discussed in a later paper dealing with breeding and hybridisation data.

ACKNOWLEDGMENT

We are grateful to Miss S. Duffey for preparing the figures.

REFERENCES

ASCHERSON, P. & GRAEBNER, P. (1909). Synopsis der Mitteleuropdischen Flora, 6: 959-975. Leipzig.

BALL, P. W. (1968). Vicia L., in TuTIN, T. G. et al., eds. Flora Europaea, 2: 129-136. Cambridge.

Bum, A. (1966). The influence of plant density on the morphological characters and seed production of common vetch, V. sativa L. Exp. Agric., 2: 61-67.

Davis, P. H. & PLITMANN, U. (1970). Vicia L., in DAvis, P. H., ed. Flora of Turkey, 3: 274-325. Edinburgh.

HANELT, P. & METTIN, D. (1966). Cytosystematische Untersuchungen in der Artengruppe um Vicia sativa L., 2. Kulturpflanze, 14: 137-161.

HOLLinGs, E. & STAcE, C. A. (1974). Karyotype variation and evolution in the Vicia sativa aggregate. New Phytol., 73: 195-208.

Hooker, W. J. (1830). The British flora. London.

Hupson, G. (1762). Flora Anglica, pp. 278-279. London.

KiLuick, H. J. (1975). The decline of Vicia sativa L. sensu stricto in Britain. Watsonia, 10: 288-289.

METTIN, D. & HANELT, P. (1964). Cytosystematische Untersuchungen in der Artengruppe um Vicia sativa L., 1. Kulturpflanze, 12: 163-225.

PLITMANN, U. (1967). Biosystematic study in the annual species of Vicia of the Middle East. Jerusalem.

Rouy, G. (1899). Flore de France, 5: 209-216. Paris.

SmiTH, J. E. (1825). The English flora, 3: 281-283. London.

Syme, J. T. B. (1864). English botany, 3rd ed., 3: 95-99. London.

TutTin, T. G. (1952). Vicia L., in CLAPHAM, A. R., TuTIN, T. G. & WARBURG, E. F. Flora of the British Isles, pp. 442-447. Cambridge.

TutTin, T. G. (1962). Vicia L., in CLAPHAM, A. R., TUTIN, T. G. & WARBURG, E. F. Flora of the British Isles,2nd ed., pp. 354-359. Cambridge.

VAVILOV, N. I. (1951). The origin, variation, immunity and breeding of cultivated plants. Waltham, Massachusetts.

YAMAMOTO, K. (1966). Studies on the hybrids among Vicia sativa L. and its related species. Mem. Fac. Agric. Kagawa Univ., 21: 1-104.

(Accepted May 1977)

Watsonia, 12, 15-21 (1978). | 15

Cytotaxonomic studies on the Cochlearia officinalis L. group from inland stations in Britain

Inde Bi-GikL

Department of Genetics, University of Liverpool

H. A. McALLISTER University of Liverpool Botanic Gardens, Ness, Neston, Wirral and

G. M. FEARN

Department of Chemistry and Biology, Sheffield City Polytechnic

ABSTRACT

The results of a cytotaxonomic survey of Cochlearia Officinalis L. sensu lato from inland sites in Britain are reported. One diploid and two tetraploid cytotypes exist. The diploid (2n = 12) occurs at moderate altitudes and appears to be restricted to base-rich habitats. It corresponds both ecologically and cytologically with C. pyrenaica DC. and, although, in the authors’ experience, it cannot be readily distinguished from the tetraploid C. officinalis L. sensu Stricto on morphological grounds, it is most meaningfully regarded as forming an extension to the range of C. pyrenaica. The two tetraploids are distinguishable from each other both morphologically and cytologically and usually occur at higher altitudes than the diploid, although the 2n = 24 cytotype occasionally is found at lower altitudes. The 2n = 24cytotype, which frequently contains B chromosomes, is assigned to C. officinalis L., while the cytologically constant 2n = 26 cytotype represents C. micacea Marshall. On the basis of the ecology and chromosome numbers of the above three species, it is suggested that C. alpina (Bab.) Wats. is best regarded simply as an inland ecotype of C. officinalis L. The probable post-glacial history of the genus is discussed.

INTRODUCTION

The genus Cochlearia L. is widely distributed in the British Isles, where it occurs in both maritime and inland-upland habitats. It is generally agreed that the coastal populations fall into three or possibly four species. These are C. officinalis L., C. danica L., C. anglica L. and possibly C. scotica Druce. The first three are common round much of the coasts of the British Isles, except that C. anglicais apparently absent from Orkney and Shetland. C. officinalis is an extremely common plant of sea-cliffs and the drier parts of salt-marshes, and it also occurs inland; C. danica is common on the drier parts of sea-cliffs and in sandy places by the sea and occurs rarely inland; and C. anglica is locally abundant on muddy salt- marshes. C. scotica is a plant of the north and west, although it may possibly occur as far south as Berwick in the east. Gill (1971) has, however, cast some doubt on the reported distribution of C. scotica.

The inland-upland populations have been recognized as belonging to C. alpina (Bab.) Wats., C. pyrenaica DC., C. officinalis, or C. micacea Marshall. It has, however, been suggested that neither C. alpina nor C. micacea are entirely montane plants but may descend nearly to sea-level at least in the islands of northern and western Scotland (Druce 1932). The status of the inland populations of Cochlearia have been much disputed, with even the same author recognizing the taxa at different levels at different times—Clapham (1952) recognized both C. alpina and C. micacea but in 1962 submerged both in C. officinalis as C. officinalis L. subsp. alpina (Bab.) Hook. Chater & Heywood (1964) included

16 JAI By GILL SE TALE:

C. alpina in C. pyrenaica. Like Clapham (1962), these authors did not distinguish C. micacea, regarding it only as a narrow-fruited variant of C. pyrenaica. In their treatment of C. officinalis, Chater & Heywood included a note—*Plants from Scotland approach the following species’ [C. pyrenaica]. At least some of the taxonomic confusion which exists in Cochlearia has arisen because the characters which are used to distinguish between the taxa are mostly quantitative or plastic, or both. The literature abounds with comparisons such as leaves fleshy versus leaves not or hardly fleshy (Clapham 1952), silicula ovoid to globose versus silicula ovoid-ellipsoid (Chater & Heywood 1964) or pedicels usually longer than silicula versus pedicels equalling or shorter than the silicula (Chater & Heywood 1964). Occasionally there occurs absolute contradiction between the keys and the descriptions of the species; Clapham (1952) separated C. alpina and C. micacea from the rest of the genus by ‘Inland, usually alpine plants with leaves not or hardly fleshy —C. alpina or C. micacea ‘Maritime plants with fleshy leaves—other species including C. officinalis, yet his description of C. officinalis states ‘reaches 2,800 ft on Ben Creachain’. The taxonomic uncertainties which exist are to some extent reflected in the various chromosome numbers which have been reported for these species (Table 1).

TABLE 1. PREVIOUSLY PUBLISHED BRITISH COUNTS FOR COCHLEARIA ALPINA, C. OFFICINALIS AND C. MICACEA

Species Locality Author 2n

C. alpina Scotland Crane & Gairdner 1923 28 Malham, Mid-W. Yorks, v.c. 64 Gill 1965 12 (+0-2 B) Helvellyn, Cumberland, v.c. 70 Gill 1965 12 (+ 0-1 B)

C. officinalis St Davids, Pembs., v.c. 45 Crane & Gairdner 1923 28 Teesdale, Durham, v.c. 66 Saunte 1955 24 Ben Bulben, Sligo, v.c. H28 Saunte 1955 24 | Black Head, Clare, v.c. H9 Saunte 1955 24 Lauragh, S. Kerry, v.c. H1 Saunte 1955 24 Birdsay, Orkney, v.c. 111 Gill 1973 24 Yecansby, Orkney, v.c. 111 Gill 1973 24 Wick, Caithness, v.c. 109 Gill 1973 24 (+0-2 B) Coast of Isle of Skye, Mid Ebudes, v.c. 104 Gill 1973 24 Banff, Banff., v.c. 94 Gill 1973 24 (+ 0-1 B) Carnoustie, Forfar, v.c. 90 Gill 1973 24 Lamlash, Arran, Clyde Is., v.c. 100 Gill 1973 24 (+0-+1 B) Caernarvon, Caerns., v.c. 49 Gill 1973 24 Hook Farm, Pembs., v.c. 45 Gill 1973 24 Parkgate, Cheshire, v.c. 58 Gill 1973 24

C. micacea Scotland Crane & Gairdner 1923 34-36 Ben Lawers, Mid Perth, v.c. 88 Gill 1973 26 Meall nan Tarmachan, Mid Perth, v.c. 88 Gill 1973 26 Beinn Heasgarnich, Mid Perth, v.c. 88 Gill 1973 26

C. sp. (Probably Various coastal localities in the Gill 1971la 24

C. officinalis) northern and western Scottish islands

As orthodox taxonomic methods had failed to produce any agreement on the status of the inland populations of Cochlearia, the present work was initiated to determine the extent of the morphological and cytological variation in the group.

METHODS Material was collected in the wild and identified on morphological criteria as either C. officinalis or C.

micacea (for criteria see Discussion). The majority of the plants were grown on at the the University of Liverpool Botanic Gardens until root-tips and/or flower-buds could be taken for the determination of

CYTOTAXONOMY OF COCHLEARIA OFFICINALIS L. IN BRITAIN 17

chromosome numbers. These chromosome counts were made by J.J.B.G. who was kept ignorant of both the taxonomic identities and localities of collection of the material until after the chromosome numbers were determined. Chromosome counts were also made by G.M.F. on flower-buds collected in the field or on root-tips and/or flower-buds from plants grown on at the University of Sheffield Experimental Garden. Chromosome counts were made by the methods previously described (Gill 1965, Fearn 1971).

RESULTS

Of all the populations examined only those from Beinn Dearg and Glas Maol were initially identified as C. micacea. These were also the only populations which gave a constant chromosome count of 2n = 26, thus agreeing with the chromosome number for C. micacea previously published by Gill (1973). All the other populations were originally identified as C. officinalis but formed two cytotypes. The 2n = 24 cytotype, which frequently contains B chromosomes, corresponds with C. officinalis sensu stricto and was collected from high-altitude stations (above 800m) and from a single low-altitude locality at Cheddar Gorge. All the high-altitude stations were well flushed and presumably relatively base-rich. The 2n = 12 cytotype, one population of which contains B chromosomes, was collected exclusively from highly base-rich habitats at intermediate altitudes (150—750m).

All the chromosome numbers determined in this investigation are summarized in Table 2 and are included with all previously published counts in Fig. 1.

FiGurE 1. Map showing all published chromosome counts in Cochlearia together with those reported in this paper (all counts made by the present authors blacked in).

= 10 0-28) © 7 = 24¢ = 0-5B) Cl S 2 SS 2G

9¢ € 9-¢ Hd “weg ‘moyjoy dureq ‘o1ue1H 06 ‘OA ‘TejIOJ TORY sejH LL OLE wW000'T 9¢ S ‘SJJt]9 SUIOBY “N Ul SEdIAeID YOo1 Apeys oA SOI “OA ‘ssoy “AA “BIveq UUlog 818°SS7/87 DOIDINA FD w¢/ 6 € ‘QS19A SPISPVOY “OUO}SOUMT] SNOJaJIUOGIeD 9 ‘OA “JOSIIWOG “N “981OH IeppeyD vS 9v/TE (sjueyd Z) (€€ + pz 6b (OA “SuIaeD (az + )pc (AI + Ire “HZ S WOOT] ‘seysny o4ydok1q 19 ‘UopMous ‘ppsusey k UAMSOID ~—_QQQ¢'009/€Z (as + )bz 96 (OA IIABIW (ap + pz z WOOT] ‘S20R] YOOI 194 wey “UleIod UUlog “UINIINID AMO ~— ¢g¢0L0/LZ (as + )vZ Z6 OA “UdEpPIDQY ‘Ss (sjuejd Z) pZ € WOOT ‘seSpoy poysnyj ‘IesvuyooyT “nodg youlg 968 °8SZ/LE (az + )rZ L6 (O°A “SSOUIO}ISO AA (a1 + )pz z WOOT T ‘GJ uo soysny aAydoh1g ‘UOSUIED “epyovoUug,) UR BOD —-—ZE0' 766/87 (dz + bz (syuejd 7) (1 + )pz COL “9° ‘ssOW “M vC Vv WIS /8 “SoSP2] YOOI 19M ‘UISTITY UUlog “Yoresenin eu Woy 96S 858/81 syounyfo ~D (dz + ZI €-L Hd ‘wig6] ‘weans Jo syueq pue cl € poeq sjsulyg ‘“ou0}spues SNOJOJIUOGIe) 6S “OA “SoURT “§ ‘OpISIMIUA 8LEL/PE p-L Hd “wo¢¢ ‘sourur pea] plo jo ZI [e1aAos sdeay [lodg ‘ouojsoul] snosojruoqied LS ‘oA “Skqiog ‘ayey MOWIG 783 °SL/€v 0-L Hd “Wweg9] ‘ayrey Joary ZI I jo syueq pnyy “ou0}sou] snoJojIUOgIeD 9 ‘OA “SYIOX “A-PYAL ‘UO WuUISseIH 79 '66/PE L-9 Hd ‘wy {z ‘sourw pea] plo sesu TI [e1oAas OPISUIVII}S DUOJSOUI] sNOJaJIUOGIeD 69 OA “SHIOXK “M'N ‘TIPPOOM 68°86/PE ZI L WOZp ‘MOPKOU snodIed[ed UI apIsUIedI}S 99 ‘OA “WeyING ‘aepssay, BOE '7TS8/SE Gl! vl WO8y ‘SNooresed “flods ourw pio 99 “oA “ueYyING ‘peoyoud ], L8ELSL/SE wizEeL “Yysny “Al Z apnjiye YSIP{ “oUojsoUN] SNOJayIUOGIeD 69 “OA “pUvIIOUTSIAM ‘BPJSeD UddIH IG WLISS poinuaddd “Dd UZ syueyd $oj0U Je}IQeH Ay]e00 T IOUdIIJOY saroads JOUON Pero)

NOILVOILSHANI SIHL NI AAVW SdIOddS PIYVATHIOD AO SLNNOO ANOSOWOUHD 7 ATEVL

CYTOTAXONOMY OF COCHLEARIA OFFICINALIS L. IN BRITAIN 19

DISCUSSION

It is apparent from Table 2 that in Britain there exist inland populations of Cochlearia with chromosome numbers of 2n = 12 (+ 0 or 2B), 2n = 24(+ 0 —5B) and 2n = 26. These numbers agree with those already published by Gill (1965, 1971a, 1973) for British populations identified as C. pyrenaica, C. officinalis and C. micacea respectively. The counts of 2n = 28 for C. alpina and 2n = 34-36 for C. micacea reported by Crane & Gairdner (1923) were not corroborated, so these early counts must remain very doubtful.

The count of 2n = 28 for C. alpina has been discussed by Saunte (1955), who has pointed out that, in Crane & Gairdner’s drawing of the somatic chromosomes of this species, there are apparent four chromosomes which are smaller than the others. Saunte interpreted these as being accessory chromosomes and therefore regarded the count as in reality 2n = 24 + 4B. This interpretation of course gives Crane & Gairdner’s material the same chromosome number as C. officinalis (Saunte 1955, Gill 1973) and indeed Saunte, who was unaware of the existence of the diploid (2n = 12), did not regard the British inland material known to her as being other than C. officinalis. It has been well established by both Saunte (1955) and by Gill (1971a, 1973) that specimens of C. officinalis containing four B chromosomes, and thus giving chromosome counts of 2n = 28, are common. Crane & Gairdner’s count appears to have been derived from only a single specimen and it is therefore quite possible that it represented an atypical individual from a population of C. officinalis. Crane & Gairdner do not give the locality for the collection of their material and it has not proved possible to trace either this or their voucher specimens, so that no direct check of their count can be carried out. The type locality for C. alpina is Lochnagar (Druce 1904) and our three plants from this locality had chromosome numbers of 2n = 24 (2 plants) and 2n = 29. It would thus appear that our population from the type area was a population of plants with 2m = 24 + 0-SBchromosomes. The situation in C. a/pinais further complicated by the fact that in Smith’s (1811) description of Don’s material from Lochnagar the plant is said to have extremely fleshy basal leaves, while Clapham (1952) separated C. alpina from C. officinalis on the absence of fleshiness in the leaves of the former species. Although the absence of fleshiness in the basal leaves was used by Gill (1965) to separate the diploid material from Malham and Helvellyn from the C. officinalis populations then known to him, it was found that many of the inland 2” = 24 (+ 0-5B) populations also had non-fleshy basal leaves. It is perhaps worth noting here that the fleshiness or non- fleshiness of the leaves is one of those quantitative characters much used in Cochlearia but extremely difficult to apply objectively.

Of all the tetraploid counts only three are from low or moderate altitudes. Two of these (Ben Bulben and Cheddar Gorge) are near to the coast and may be interpreted as being remnant coastal populations of C. officinalis, isolated as inland populations with the falling of the sea-level during the post-glacial period. The count of 2n = 24 from Teesdale (Saunte 1955) should be treated with some suspicion as, despite extensive sampling in that area, the present authors are able to report only diploid counts.

The results suggest, therefore, that the diploid is confined in Britain to apparently base-rich habitats at moderate altitudes (cliffs, streamsides, old mine spoil-heaps), but that near the sea, even on base-rich soils, it may be replaced by the tetraploid. At high altitudes the tetraploid occurs in areas which may or may not be base-rich. Confirmation of such distributions must, however, await more extensive sampling of upland calcareous areas in Scotland and a general survey of upland populations in Ireland.

The status of C. alpina is still difficult to ascertain. The authors are conscious that despite the demonstration that material from the type locality (Lochnagar) is tetraploid they have not yet seen the herbarium specimens of either Don or Crane & Gairdner and therefore cannot with certainty identify C. alpina. They would, however, suggest that, as Lochnagar is not a base-rich habitat, it is unlikely that the diploid would occur there and that it is therefore unlikely that Don’s material was diploid. The extreme difficulty of separating C. alpina from C. officinalis on morphological grounds together with the extremely doubtful differentiation on chromosome number leads them to suggest that C. alpina should be treated only as an inland ecotype of C. officinalis.

The diploid material can be identified, using existing Floras, with equal facility as C. pyrenaica or C. officinalis and it has the same chromosome number as C. pyrenaica (Rohner 1954; Dersch 1962, 1968; Gill 1971b; Kakes 1973). C. pyrenaica in Europe has long been recognized as a plant of base-rich habitats (Hegi 1919, Hiemans 1971) and this also appears true of the diploid populations in Britain. It is thus apparent that, despite the lack of morphological distinguishing characters, the 2n = 12

20 J.J. B. GILL, ET AL.

populations in Britain are best regarded as forming an extension of the known range of C. pyrenaica. The tetraploid 2n = 24(+ 0-SB) populations are best treated as inland populations of C. officinalis.

Those populations which gave a constant count of 2n = 26 as well as comprising a unique cytotype also form a morphological entity and can, in the experience of the authors, be taxonomically distinguished as C. micacea. The chromosome number of this species has been discussed by Gill (1973). The characters which are used to define C. micacea are the much darker green of the foliage of this species when compared with the rest of the inland plants, together with the low-growing habit and strong tendency to produce a perennating woody rootstock with some slight vegetative reproduction by lateral shoots. These characters are, however, all comparative characters and, to use them successfully, it is necessary to be well acquainted with the other inland plants. When, however, these characters are combined with the chromosome number and the known genomic constitution of the taxon (Gill 1973), the authors can see no conclusion other than the maintenance of C. micacea as a separate species. Such a conclusion agrees with that of Pobedimova (1971) and is supported by the constant distinction of C. micacea from C. alpina by McVean & Ratcliffe (1962). The details of the distribution of C. micacea must await further investigation but it appears to be a plant of high altitudes usually above 800m. It may indeed be a British endemic but there is some suggestion that it may also occur in Scandinavia (Hylander 1945).

The extreme morphological similarities of the three species recognized here and the difficulties experienced by orthodox taxonomists are to be expected if the evolutionary relationships between the different chromosome levels are as suggested by Gill (1973). He demonstrated that C. officinalis is essentially an autotetraploid of C. pyrenaica and that C. micacea may be regarded as a primary tetrasomic of C. officinalis. These relationships would result in the only unique genes or alleles possessed by any of the species being those which arose by mutation shortly after speciation. Such mutations would be very rare and, therefore, if the evolutionary pathways suggested by Gill are accepted, it would be expected that the differences between C. pyrenaica, C. officinalis and C. micacea would be those arising from the differing number of gene copies in the three species. Such differences would almost certainly be quantitative rather than qualitative but must account for the differing ecological amplitudes of the taxa.

The tetraploid C. officinalis, if it originated from a highly heterozygous diploid, would have a greatly increased variance for many loci and could, therefore, be much superior to the diploid in its ability to colonize new habitats. The inland distribution of Cochlearia is typical of one much affected by glaciation and it is tempting to suggest, because of the existence of the Polish endemic C. polonica Borb. and the possible endemic status of C. micacea, that much of the speciation in the European members of the genus occurred either during or at the end of the last glaciation. The great number of new habitats made available by the retreating ice would have produced a selective regime in which increased variance would be of considerable advantage. Such a situation would seem to be ideal for the establishment and spread of an autotetraploid species such as C. officinalis. The diploid parent of any autotetraploid would almost certainly be in competition with its tetraploid offspring, but the continued existence of both may also be due to the differences in copies of gene loci between them. The diploid, because of the small number of copies of each locus present, would be able to fix a genetic trait much more quickly than the tetraploid and, in the face of competition from the tetraploid, could retreat into some extreme or relict environment similar to that to which it has already been exposed. In this extreme environment the relative lack of variance of the diploid could well enable it to compete successfully against the tetraploid and eventually, by the accumulation of new mutations, to become so highly and specifically adapted to the particular extreme habitats into which it had retreated that competition from the tetraploid ceased. The habitats at present occupied by C. pyrenaica in northern England would appear to be sufficiently extreme to be accountable for by the events suggested above.

REFERENCES

Cuater, A. O. & Heywoop, V. H. (1964). Cochlearia, in Tutin, T. G. et al., eds. Flora Europea, 1: 313-314. Cambridge.

CLaPHaM, A. R. (1952). Cochlearia, in CLAPHAM, A. R., TUTIN, T. G. & WARBURG, E. F. Flora of the British Isles, pp. 188-192. Cambridge.

CYTOTAXONOMY OF COCHLEARIA OFFICINALIS L. IN BRITAIN DA

CLAPHAM, A. R. (1962). Cochlearia, in CLAPHAM, A. R., TUTIN, T. G. & WARBURG, E. F. Flora of the British isles, 2nd ed., pp. 150-153. Cambridge.

CRANE, M. B. & GaAIRDNER, A. E. (1923). Species crosses in Cochlearia with a preliminary account of their cytology. J. Genet., 13: 187-200.

Derscu, G. (1961). In Lupwie, W., Cochlearia officinalis s.str. und Cochlearia pyrenaicain Hessen. Hess. flor. Briefe, 10: 51-53.

Derscu, G. (1968). Uber einige Chromosomenzaklungen an mitteleuropdischen Bliitenpflanzen. Abh. Ver. Naturk., 52: 9-12.

Druce, G. C. (1904). The life and work of George Don. Notes R. bot. Gdn Edinb., 12: 53-290.

Druce, G. C. (1932). The comital Flora of the British Isles. Arbroath.

FEARN, G. M. (1971). Biosystematic studies of selected species in the Teesdale flora. Ph.D. thesis, University of Sheffield.

GILL, J. J. B. (1965). Diploids in the genus Cochlearia. Watsonia, 6: 188-189.

GILL, J. J. B. (1971a). Cochlearia scotica Druce—Does it exist in northern Scotland? Watsonia, 8: 395-396.

GILL, J. J: B. (1971b). Cytogenetic studies in Cochlearia L. The chromosomal homogeneity within both the 2n = 12 diploids and the 2n = 14 diploids and the cytogenetic relationship between the two chromosome levels. Ann. Bot., 35: 947-956.

GILL, J. J. B. (1973). Cytogenetic studies in Cochlearia L. (Cruciferae). The origins of C. officinalis L. and C. micacea Marshall. Genetica, 44: 217-234.

HEaI, G. (1919). I/lustrierte Flora von Mittel-Europa, 4(1): 330-333. Munich.

HIEMANS, J. (1971). Cochlearia pyrenaica. Gorteria, 5: 153-158.

HYLANDER, N. (1945). Nomenklatorische und systematische Studien tiber nordische Gefasspflanzen. Uppsala & Leipzig.

KAKEs, P. (1973). The chromosome number of Cochlearia pyrenaica DC. near Moresnet (Belgium). Acta bot. neerl., 22: 206-208.

MCcVEAN, D. N. & RATCLIFFE, D. A. (1962). Plant communities of the Scottish Highlands. Monographs of the Nature Conservancy, 1. London.

PoBEDIMOVA, E. (1971). Revisio generis Cochlearia L., 2. Sist. Vyss. Rast. (Leningrad), 7: 167-195.

ROHNER, P. (1974). Zytologische Untersuchungen an einigen schweizerischen Hemi-Oreophyten. Mitt. naturf. Ges. Bern, 11: 43-107.

SAUNTE, L. H. (1955). Cytogenetical studies in the Cochlearia officinalis complex. Hereditas, 41: 499-515.

SmiTH, J. E. (1811). An account of several plants, recently discovered in Scotland by Mr George Don, A.L.S. and not mentioned in the Flora Britannica nor English Botany. Trans. Linn. Soc. Lond., 10: 333-346.

(Accepted June 1977)

et WIATIAG V1 1 NORD RY

ale a verter Bes ae. SANS pent a. es ree

\ " SAVES AiR Sitoaib eds Bs es we TE te eae oe: ay - the a she az Dit see REY A tO RP. RRMA ABET Se

pein aie ge STAs S sdiaohelearienn AbbeanS : Imogene oie Ae iS: suitiestninna aoe ay pene weeny , , ate al i : me sete

ovis. | ws

| ees < | atta Cen ee sinogste ening Peer bate

Cy = wt Shia i od. eedte. Piotaaigit ipeeamee ba t ae npotpet arte Oho: owt ‘aHy. moawlied ileal © hes

pipet Bere Tt ohats ARNG Cie: te digiasodT allele

a] Dae & i Cie ay

Tre)

| via REEM ER Digestion ‘we ata

| Hi Seitiiaat: Sysco Siete) isan) cae

Dy: Ud NAGE AY 1OWO pe cai yee USA sla paint lcanaTNy a hut olabeanccho

Oy, (04, Go rr tt

piaase bed Sid. Acme) VeienaMs 18a OG sonore ahaa TOTAL oft

Seq e yt Sas to ania won cae & fase Wee ait Yersinia ft at Set A: Ss D k . et

Fad ets Sk Sarai t ae Lie eet Re SOP ORY Ua bres

RP INTO PER ‘aatelighviO rif pig feds pet amigo ag naga. tetiagtihieesa ae b

; Jest ine taka Me

At 2—-GlS) he) nial sign prey say 1G mre ia per ii outa “yi :

hie Codey ROC eos pA brbliood niemyneiD viteaodn 2tdeid dates te

OEE -ERE GF Gea Se ppote eel aa ipa ‘ella ainaacttion

$4 hatate Cake ris Cegaemld

; Tek wi pach en ae roe

-

Ae i ih elaelinn The + great laime eb: ue eae amak rs

bP at a L oad Wee Sa eats * Mineo \ 7 “iy min ets

i « ead ie'4 i i oe | AR RES AK i

Watsonia, 12, 23-27 (1978). 23

Notes on British Rubi, 5 E. S. EDEES

23 Dartmouth Avenue, Newcastle, Staffs.

ABSTRACT

Rubus fuscus Weihe & Nees is discussed and two new species are described, viz. R. anglofuscus and R. informifolius. R. nemoralis P. J. Muell. is confirmed as the correct name for R. se/meri Lindeb.

1. Rubus fuscus Weihe & Nees in Bluff & Fingerh., Comp FI. Germ., 1: 681 (1825)

British batologists have misunderstood this taxon. They have applied the name too loosely. Rogers (1900) called R. fuscus ‘variable’ and ‘seldom quite typical with us’ and (note on specimen in BM) ‘notoriously one of our most hopelessly aggregate species.’ The herbarium of W. C. Barton and H. J. Riddelsdell in BM contains about 100 sheets labelled R. fuscus, but, as far as I can ascertain, only one of them, a specimen collected by Focke in 1875 from Weihe’s original station (see below), is correctly named. Focke is partly to blame for the confusion. He knew very well what Weihe meant by R. fuscus, but in later life broadened his interpretation of the name to accommodate related taxa which he had found in England and other European countries. He wrote (Focke 1914): ‘Planta typica species vel prospecies bene distincta videtur, sed innumerae occurrunt formae, in quibus omnes partes leviter variant, ita ut alias species imitentur. Limites R. fusci igitur dubii et variabiles sunt.’ W. C. R. Watson began by trying to define R. fuscus precisely, but ended like Focke in a generalization. In 1949 he correctly distinguished two taxa, one from The Chart, Limpsfield, Surrey, v.c. 17, which he claimed to be the true R. fuscus, and one from the west of England, which he identified with R. fusciformis Sudre. But in 1958 he united them and called them both R. fuscus.

The lectotype of R. fuscus, designated by Professor H. E. Weber in 1976, is a specimen in BR collected by Weihe. It consists of a panicle and a piece of the barren stem with one leaf. The label in Weihe’s handwriting reads as follows: “Rubus fuscus Weihe in montibus circa Altena comitatus Mark.’ It is undated, but, as Weihe is known to have visited Altena only once, there is no doubt that it is one of the specimens referred to in the original description. There are other undated specimens from the same locality in MSTR and yet another in Focke’s herbarium in BREM, but these are mite-infested and therefore less satisfactory. I have not seen the lectotype, but Weber has given me three photographs of it, one showing the whole plant and two showing a flower head and the barren stem enlarged. The specimen in BM collected by Focke at Altena in 1875 matches the photographs very well and so does an undated specimen in K collected by G. Braun (Herbarium Ruborum germanicorum no. 93), ‘Auf mergeligem Boden im Amte Sternberg, Lippe—Detmold, sehr verbreitet.’ Weber visited the Jocus classicus in 1970 without finding R. fuscus there, but he said it was more or less plentiful in southern Westphalia as far as the Lippe district, especially at Blomberg. I have two recent specimens from Blomberg, collected by A. Newton and Weber in 1974, which tally very well with the lectotype and which are, to quote Weber (in /itt. 1975), ‘certainly true R. fuscus with all the characteristic features.’

The following description is based on the Blomberg specimens:-

Stems angled, clothed with numerous, short (c 0-5—1 mm) hairs, acicles and stalked glands, many of the glands (as distinct from some gland-tipped acicles) being shorter than the hairs. Prickles numerous (c 12 every 5 cm), about as long as the stem diameter, subequal, declining.

Leaves quinate (one ternate), pedate. Leaflets green and thinly hairy on both sides; upper surface dark, matt green. Terminal leaflet broadly elliptic to obovate with somewhat straight sides and tapering to a fine point; base emarginate or truncate. Serration uneven, with the principal teeth more prominent than the others and sometimes slightly retrorse. Petiolule about 1/3 length of blade.

24 E. S. EDEES

Inflorescence leafy for 2/3 of its length with ternate and simple leaves subtending peduncles; lower peduncles 6 cm, upper 2 cm long; upper leafless part of inflorescence more or less cylindrical, with short, subequal peduncles; all peduncles divided near the top into short pedicels producing dense clusters of flowers. Rachis densely hairy with many stalked glands almost as long as the hairs. Prickles numerous, slender, declining or curved.

Sepals grey-felted with many acicles and stalked glands, reflexed. Petals pinkish, broadly elliptic; filaments equalling pinkish-based styles; carpels and receptacle pilose.

The description in Weihe & Nees (1826) differs in a few ways, but the full page illustration in that book (t. xxvil) is wonderfully accurate. The prickles on the barren stem are said to be sparse, but the illustration shows them to be numerous. The terminal leaflet is said to be cordate-ovate, but again the illustration shows it as described above, though with a slightly more indented base. The most serious discrepancy is in the shape of the petals, which are said to be suborbicular and are so drawn in the illustration. But Focke (1877), soon after he had studied the living bushes in Weihe’s locality, wrote: ‘Petala obovato-cuneata alba vel rosea.’ The barren stem-piece is very well drawn and shows the short glands, hairs and acicles to perfection.

I am not sure that typical R. fuscus has yet been discovered in the British Isles. Newton and Weber found a bramble at Lee Clump, near Wendover, GR 42/911.046, Bucks., v.c. 24, 17/7/1976, which they both thought was R. fuscus. But the specimen I have seen is not in my judgment identical with the Blomberg plant. For one thing the toothing of the leaflets is not the same. Weber’s latest opinion (in Jitt. 1977) is that itis an extreme form of R. fuscus, though it was not the leaf serration but the relatively long glands and short hairs which seemed to him atypical.

Of the many taxa which have been labelled R. fuscus by British batologists, several deserve recognition as new species. Two which occur in Staffordshire are described below.

2. Rubus anglofuscus E. S. Edees, sp. nov.

Turiones obtuse angulati, in apricis rubescentes, interdum pruinosi, dense pilosi. Aculei numerosi (10-15 per 5 cm), e basi lata declinati, inaequales, impariter dispositi aculeolis tuberculisque saepe interspersi. Aciculae et glandulae stipitatae pilos aequantes vel eis breviores, vulgo sparsiores.

Folia ternata aut quinata pedata. Foliola laete viridia, supra strigosa, subtus leviter pilosa, dentibus latis inaequaliter serrata. Foliolum terminale parum obovatum, breviter acuminatum, basi emarginatum, petiolulo proprio quadruplo longius.

Rami floriferi pilis patentibus dense vestiti. Aculei declinati. Aciculae glandulaeque stipitatae non rarae. Inflorescentiae foliaceae, e ramulis brevibus, adscendentibus, 1-7-floris compositae.

Sepala aculeolata et glandulosa, primo laxe reflexa, demum patentia vel suberecta. Petala alba, late obovata. Stamina alba stylos rubros aequantia vel parum superantia. Carpella et receptacula pilosa.

Stems bluntly angled, red, densely clothed with patent hairs (c 1 mm long), sometimes pruinose. Prickles numerous (10-15 per 5 cm), slightly declining from a broad base, unequal in length and unevenly distributed, often interspersed with pricklets and tubercles. Acicles and stalked glands equal to or shorter than the hairs, usually few and sometimes rare or absent.

Leaves ternate or quinate pedate. Leaflets bright green, strigose above, thinly hairy beneath, biserrate with broad teeth. Terminal leaflet obscurely obovate, with short acuminate point and emarginate base, about four times as long as its petiolule.

Flowering branches clothed with many patent hairs and numerous acicles and stalked glands. Prickles more strongly declining than those of the stems. Panicles with short, ascending peduncles bearing 1-7 flowers, sometimes leafy throughout, but usually with a short leafless extension. Leaflets like those of the stem-leaves in shape, colour, texture and serration.

Sepals aculeolate, glandular, at first reflexed, then patent or suberect. Petals white, broadly obovate. Stamens white, equalling or slightly exceeding the red or reddish-based styles. Carpels and receptacles hairy.

HOLOTYPUs: Wigginton Heath, GR 42/3.3, Oxon, v.c. 23, 24/7/1931, H. J. Riddelsdell, as R. fuscus Weihe & Nees, no. 3007 in herb. Barton & Riddelsdell (BM)

NOTES ON BRITISH RUBI, 5 25

R. anglofuscus differs from R. fuscus in several ways. In R. anglofuscus the stalked glands on the barren stems may be as long as the patent hairs, but are usually few and sometimes rare or even absent. R. fuscus on the other hand has a rather dense clothing of short acicles and stalked glands, which Weihe & Nees (1826) described thus: “Aciculae semilineam vix longae nudoque oculo fere inconspicuae, sed tactu sine negotio distinguendae.’ The leaves of the two taxa have a very different appearance. Those of R. fuscus are usually quinate. Weihe & Nees (1826) said, ‘Folia omnia quinata’. But those of R. anglofuscus are often ternate. However there are too many exceptions on both sides to make this a reliable distinction. There is a much more important difference in the shape and serration of the terminal leaflets. Those of R. fuscus have fine acuminate points and rather straight sides reminiscent of R. pallidus Weihe & Nees, whereas the terminal leaflets of R. anglofuscus have shorter, broader points, more rounded sides and markedly broader teeth. The colour of the foliage is another difference. The leaflets of R. anglofuscus are normally a bright, fresh green contrasting with the brick-red colour of the stems; the leaflets of R. fuscus are darker, ‘saturate viridia’ (Weihe & Nees 1826). The sepals of R. fuscus are reflexed; those of R. anglofuscus become patent or erect after anthesis. The petals of R. fuscus may be pink or white; those of R. anglofuscus seem to be always white.

R. anglofuscus is a common bramble of the west Midlands from Oxford north to Staffordshire. I have specimens in my herbarium (herb. E. S. Edees), all collected by myself, from the following localities: Norton near Stourbridge, GR 32/8.8, Worcs., v.c. 37; Sutton Park, GR 42/1.9, and Edge Hill, GR 42/3.4, Warks., v.c. 38; Trysull and Seisdon, GR 32/8.9, Arley Wood, GR 32/8.8, Enville, GR 32/8.8, Baggeridge Wood, GR 32/8.9, and Wrottesley, GR 33/8.0, Staffs., v.c. 39; Puleston Common, GR 33/7.2, and near Claverley, GR 32/8.9, Salop, v.c. 40; and another collected by myself from the type locality.

Watson’s specimens from The Chart, Limpsfield, Surrey, v.c. 17, which he distributed through the Botanical Exchange Club in 1945 as R. fuscus, are probably R. anglofuscus, though on the two sheets I possess the barren stem is not sufficiently hairy to be typical and the short acicles and stalked glands, which do not exceed 0-5 mm, resemble those of R. fuscus. There are early specimens of R. anglofuscus (all labelled R. fuscus) in CGE from Warwickshire, collected by J. E. Bagnall and others. The earliest I have seen is one in Babington’s herbarium in CGE from Sutton Park, collected by W. Mathews in 1850. In Barton and Riddlesdell’s herbarium in BM there is an excellent series from Wiggington Heath, Oxon., v.c. 23, collected by H. J. Riddelsdell between 1916 and 1935, and a specimen from Seven Springs, Bourton-on-the-Water, GR 42/131.226, E. Gloucs., v.c. 33, collected by Riddelsdell in 1935.

Finally it is probable that this taxon also occurs in France. There is a specimen in MANCH of Sudre’s Rubi rari no. 215, gathered in the Rhine Province in 1911, which seems to me identical with R. anglofuscus. Sudre named it R. erubescens Wirtg., but it is certainly not that.

3. Rubus informifolius E. S. Edees, sp. nov.

Turiones obtuse angulati, in apricis rufescentes, pilis brevibus (c 1 mm) vestiti, aciculis glandulisque stipitatis (c 0-5 mm) obsiti. Aculei numerosi (c 10-15 per 5 cm), inaequales (c 2-6 mm longi), paulo declinati. Glandulae stipitatae nonnunquam sparsae.

Folia quinata pedata. Foliola pallide viridia, supra glabrescentia, subtus ad nervos leviter pilosa sed non tomentosa, non profunde crenata, ad marginem undulata. Foliolum terminale, ubi perfectum, late ovatum vel obovatum vel suborbiculare, sed non raro plane informe, cordatum vel emarginatum, breviter acuminatum.

Rami floriferi dense pilosi, aciculis multis brevibus muniti. Aculei graciles declinati vel curvati. Inflorescentiae subpyramidatae, ramulis inferioribus longis saepe divaricatis instructae, fere usque ad apicem foliosae. Foliola illis turionum pilosiora sed nequaquam tomentosa. Ramuli superiores uniflori vel in medio divisi et 2-4-flori. Pedicelli glandulis breviter stipitatis crebris aculeolisque multis tenuibus flavis praediti.

Sepala albo-marginata, appendiculata, aculeolata, glandulosa, erecto-patentia. Petala alba, late obovata sed non contigua, paulo indentata. Stamina alba stylos virides vix superantia. Carpella subglabra. Receptacula hirsuta.

Stems bluntly angled, reddish in the sun, clothed with short hairs (c 1 mm). Prickles numerous (c

26 E. S: EDEES

10-15 per 5 cm), unequal (c 2-6 mm long), slightly declining: Acicles and stalked glands rather short (c 0-5 mm), not exceeding the hairs. Stalked glands sometimes rare.

Leaves quinate pedate. Leaflets yellowish-green, slightly hairy on both surfaces, but glabrescent above and hard to the touch beneath, without felt; serration shallow; margins undulate. Terminal leaflet broadly ovate or slightly obovate in general outline, but often shapeless with humps and indentations; base emarginate or cordate; point shortly acuminate.

Flowering branches clothed like the stems with many short hairs, acicles and stalked glands. Prickles slender, declining or curved. Panicles subpyramidal, with long, often divaricate, lower branches, leafy nearly to the top. Leaflets hairier than those of the stems, but green on both sides and not felted. Upper peduncles sometimes with a single flower, but more often deeply divided and 2-4-flowered. Pedicels clothed with many fine, yellow acicles and stalked glands of varying length.

Sepals white-bordered, long-pointed, aciculate, glandular, patent to erect. Petals white, broadly obovate but not contiguous, slightly indented. Stamens white, only very slightly longer than the green styles. Carpels subglabrous. Receptacles hairy.

HOLOTYPUs: Cranmoor Park, Wrottesley, GR 33/85.00, Staffs., v.c. 39, 20/7/1958, E. S. Edees 12732 (herb. E. S. Edees)

R. informifolius is common in the Wyre Forest and in the neighbouring parts of Hereford, Worcester, Staffordshire and Salop. In addition to the holotype, the following exsiccata may be cited as representative:

Wood west of Buckenhill Common, Bromyard, GR 32/6.5, Hereford, v.c. 36, 22/7/1952, W. H. Mills, det. W. C. R. Watson, as R. aristisepalus (Sudre) W. C. R. Wats., CGE Tedstone Delamere, GR. 32/6.5, Hereford, v.c. 36, 6/8/1924, H. J. Riddelsdell, herb. Barton &

Riddelsdell, no. 957, BM Clifton (probably Clifton-on-Teme), GR 32/7.6, Worcs., v.c. 37, undated, W. C. R. Watson, as R.

aristisepalus, CGE Southstone Rock, GR 32/7.6, Worcs., v.c. 37, 24/7/1972, E. S. Edees, herb. E.S.E.

Sheep Walks, Enville, GR 32/81.85, Staffs., v.c. 39, 31/7/1954, E. S. Edees, herb. E.S.E.

Wyre Forest, near Cleobury Mortimer, GR 32/6.7, Salop, v.c. 40, 15/7/1953, W.C.R. Watson, as R. fusciformis Sudre, CGE

Whitcliff, Ludlow, GR 32/5.7, Salop, v.c. 40, 8/7/1953, W. C. R. Watson, as R. fuscus Weihe, syn. R.

fusciformis Sudre, herb. E.S.E.

Woodland near Billingsley, GR 32/70.84, Salop, v.c. 40, 9/7/1965, E. S. Edees, herb. E.S.E.

R. informifolius has several distinctive features, viz. broad, often shapeless leaflets with ungainly humps and indentations, which are hard to the touch beneath and which have broad shallow teeth sometimes coalescing near the point of the leaflet; fine rachis prickles, of which a few are strongly curved; rather long, somewhat divaricate peduncles; pedicels with fine, needle-like, yellow prickles; long-pointed, spreading sepals; and a general yellowish-green hue.

4. Rubus nemoralis P. J. Muell., Flora ( Regensb.), 41: 139 (1858) R. selmeri Lindeb., Herb. Rub. Scand. no. 33 (1884)

The object of this note is to justify W. C. R. Watson’s identification of R. se/meri Lindeb. with R. nemoralis P. J. Muell. Watson first declared these two to be the same taxon in 1938, though he advanced no evidence. He had probably seen E. Mueller’s paper, published in 1937, in which the same identification was made. E. Mueller (not to be confused with P. J. Mueller) said that R. se/meri occurred plentifully at Kaiserslautern and was ‘ohne Zweifel der Miillersche Rubus nemoralis’. 1 have a specimen in my herbarium collected by E. Mueller at Kaiserslautern, Rheinpfalz, Germany, 3rd August 1935, which is named in the collector’s handwriting “Rubus nemoralis P. J. Muell. ( = Rubus selmeri Lindeb.)’. This specimen is undoubtedly identical with R. se/meri or at any rate with the common British bramble which has been given that name. The specimen came to me from F. Rilstone and he

NOTES ON BRITISH RUBI, 5 2]

agreed that it was ‘certainly se/meri’. I have not seen the lectotype of R. se/meri designated by Beek (1974), but there is no reason to doubt the identity of the British and Norwegian plants.

The question remains, is R. se/meri identical with R. nemoralis? W. C. Barton (manuscript notes) thought it was not. He admitted that the descriptions were similar but thought that there were differences between the type specimen of R. nemoralis (Reissbach, 20th July 1857, P. J. Mueller, no. 473, LAU) and English specimens of R. se/meri which could not be explained away. However, Professor H. E. Weber has labelled Mueller’s no. 473 “Rubus nemoralis P.J.M. 1858 ( = R. rotundatus P.J.M. ined.) = R. selmeri Lindeberg 1884, holotypus.’ A. Newton and I have examined the specimen independently and have reached the same conclusion that Barton was wrong and Weber right. The specimen is not convincingly the same as R. se/meri at first sight, because the panicle prickles are not strongly curved, most of them being straight and declining. But a close examination reveals the presence of other characters which bring the specimen within the range of R. se/meri. The leaves are digitate, the leaflets glabrous on the upper surface, the sepals loosely reflexed to patent, the stamens short, the carpels pilose and, though the petals are difficult to examine, one of them at least is plainly notched. I have a specimen from Alum Chine, Bournemouth, S. Hants., v.c. 11, collected by W. M. Rogers in 1907, which closely matches Mueller’s type and which Rogers did not hesitate to call R. selmeri.

Therefore will collectors please note that specimens named R. se/meri by me should be re-named R. nemoralis? Both names are taxonomically correct, but R. nemoralis has priority.

ACKNOWLEDGMENTS

I should like to express my thanks to A. Newton, R. J. Pankhurst and Professor H. E. Weber for help in the preparation of this paper and to the curators of the herbaria cited for the loan of specimens.

REFERENCES

BEEK, A. VAN DE (1974). Die Brombeeren des Geldrischen Distriktes innerhalb der Flora der Niederlande. Tilburg.

FockeE, W. O. (1877). Synopsis Ruborum Germaniae. Bremen.

FockeE, W. O. (1914). Species Ruborum. Monographiae generis Rubi, Prodromus, 3. Biblthca bot., 83: 1-274.

MUELLER, E. (1937). Die pfalzischen Brombeeren und ihre pflanzengeographische und klimatologische Bedeutung. Jahresber. Pollichia, ser. 2, 6: 63-112.

Rocers, W. M. (1900). Handbook of British Rubi. London.

Watson, W.C. R. (1938). In Chapple, J. F. G., ed. Report of the distributor for 1937. Rep. bot! Soc. Exch. Club Br. Ish V'655:

Watson, W.C. R. (1947). In Grose, J. D., ed. Report of the distributor for 1945. Rep. botl Soc. Exch. Club Br. Isl., 13: 156.

Watson, W. C. R. (1949). Weihean species of Rubus in Britain. Watsonia, 1: 71-83.

Watson, W. C. R. (1958). Handbook of the Rubi of Great Britain and Ireland. Cambridge.

WEIHE, K. E. A. & NEES VON ESENBECK, C. G. D. (1822-27). Rubi Germanici. Elberfeld.

(Accepted April 1977)

Watsonia, 12, 29-32 (1978). 29

The occurrence of Solanum nigrum L. x S. sarrachoides Sendtn. in Britain A. C. LESLIE

Botany School, Downing Street, Cambridge

ABSTRACT

Solanum x procurrens Leslie, hybr. nov. (= S. nigrum L. x S. sarrachoides Sendtn.), a sterile hybrid, is described, with chromosome counts, from localities in south-eastern Britain.

INTRODUCTION

Gamlingay village lies 18 miles south-west of Cambridge near the boundary of Cambridgeshire, v.c. 29, with Bedfordshire, v.c. 30. Here the Lower Greensand has its botanically most important outcrop in Cambridgeshire and the area formerly held the best acid bogs in the county. These almost entirely disappeared during the last century as the land was drained and ploughed, with the result that the majority of the ground is now devoted to market-gardening, which prospers on the light soil.

To the west of the village, around the area known as the Cinques, several cultivated fields support large mixed populations of two Solanum species, the South American alien S. sarrachoides Sendtn. and the ubiquitous weed S. nigrum L. In September 1975 putative hybrids were found, easily recognized by their large size and abundant flowers. Dr J. M. Edmonds tentatively confirmed that these might be hybrids and further investigation has verified the original determination.

DESCRIPTION OF HYBRID

Stace (1975) recorded no British or European hybrids in the Solanaceae and no other literature records of this hybrid have been traced. Since the hybrid was widespread at Gamlingay, occurs in similar habitats in Bedfordshire and is likely to occur in other established mixed populations, it seems desirable that it be formally described:

Solanum x procurrens Leslie, hybr. nov.

Hybrida inter Solanum nigrum L. et S. sarrachoides Sendtn.

Herba plerumque annua vegeta, pilis glanduliferis brevibus (1-2 cellulis) patentibus numerosis et longis (4-7 cellulis) ascendentibus vel adpressis paucis pilis simplicibus longis (4~7 cellulis) ascendentibus vel adpressis numerosis vestita. Caulis principalis ad 45cm altus, erectus; rami infimi divaricati, in plantis bene evolutis decumbentibus, interdum ad nodos radicantes. Folia 40-120 x 20—45mm, viridia, aliquando ad marginem atrocaerulea, trullata vel late trullata, apice acuta, integra vel sinuato-dentata, basi subtruncata vel late cuneata.

Cymae extra-axillares, racemosae, laxae, 3-7 floribus; pedunculi sub anthesi 9-22mm longi divaricati, interdum post anthesi reflexi; pedicelli sub anthesi leviter arcuati vel erecti, post anthesi reflexi. Calyx vix vel non accrescens; lobi triangulares acuti. Corolla 10-17mm in diametro, alba (interdum extra malvini suffusa); tubo brevissimo, lobis triangularibus vel late triangularibus. Antherae bene evolutae, sed pollinibus irregularibus praecipue (98—100%) sterilibus. Bacca interdum non evoluta vel plerumque parva (2:5—4-5mm lata) nigra caduca sine seminibus, saepe bacca pedicello cadua.

Usually a vigorous annual with numerous short (1-2 celled) spreading glandular hairs, few long (4-7 celled) ascending or adpressed glandular hairs and numerous long (4~7 celled) ascending or adpressed

30 A. C. LESLIE

simple hairs. Stems to 45cm, erect; lowest branches divaricate, long-decumbent in well-developed plants and occasionally rooting at the nodes. Leaves 40-120 x 20-45mm, green, sometimes suffused bluish-black on the margin, trullate or broadly trullate, acute at apex, entire or sinuate-dentate, truncate or broadly cuneate at base.

Cymes extra-axillary, racemose, lax, with 3-7 flowers peduncles 9-22mm long at anthesis, divaricate, sometimes slightly reflexed in fruit. Calyx not or scarcely accrescent; lobes triangular, acute. Corolla 10-17mm diameter, white (sometimes tinged mauve externally); tube very short; lobes triangular or broadly triangular. Anthers well developed but pollen irregular and largely (98-100%) sterile. Berries either not developed or more usually small (2:-5—4-5mm wide), black, caducous, without seeds, and often shed with the pedicels.

HOLOTYPUs: sandy, market-garden field, Gamlingay Cinques, Cambs., v.c. 29, GR 52/228.530, 26 October 1975, A. C. Leslie no. 1029/1975 (CGE)

The epithet (procurrens, extending) was suggested by the habit of well-developed plants. A single specimen grown at the Botany School Field Station in Cambridge made a circular patch over 4m in diameter with very long decumbent branches, before being killed by the first frosts. Both parents can exhibit this behaviour, but neither in the field nor usually in cultivation is this feature so pronounced. Voucher specimens of both parents from the same locality as the holotype of the hybrid have been placed in CGE.

Both parents are variable species and to some extent S. x procurrens refleats this, especially as regards leaf shape. S. nigrum is represented at Gamlingay by subsp. nigrum, and plants of this with either entire or sinuate-dentate leaves occur, as well as plants having either normal or deeply divided corollas with very narrow lobes (var. stenopetalum Doll). A similar, apparently un-named corolla variant of S. sarrachoides occurred in a field at Potton, just over the border in Bedfordshire; this plant also had entire leaves. Specimens of S. sarrachoides with the latter character are frequent in the Gamlingay populations in company with sinuate-dentate plants.

S. x procurrens can be distinguished from its parents by the characters shown in Table 1.

TABLE 1. COMPARISON OF S. NIGRUM L. SUBSP. NIGRUM AND S. SARRACHOIDES SENDTN. WITH THE HYBRID S. x PROCURRENS LESLIE

S. nigrum subsp. nigrum S. X procurrens S. sarrachoides Longest stem hairs Appressed, Ascending or appressed, Patent, glandular eglandular occasionally glandular Calyx (in fruit) Not accrescent Scarcely or not Strongly accrescent accrescent

Calyx-teeth Broadly triangular, Triangular, acute Narrowly triangular,

(at anthesis) obtuse acute Corolla-lobes Longer than broad About as long as broad _ Ass long as broad Berry colour Black Black Green (or purplish-

brown)

Sclerotic granules Absent Absent Present

in berry Pollen fertility (°%) c 100 0-2 c 100 Chromosome number 2n = 72 2n = 48 2n = 24

The hybrid might otherwise be confused with S. /uteum Miller, which differs in its fertile, red, orange or yellow berries and the peduncles usually shorter than the pedicels; or with S. nigrum L. subsp. schultesii (Opiz) Wessely, in which the long glandular hairs are patent and the berries are fertile.

SOLANUM NIGRUM L. x S. SARRACHOIDES SENDTN. IN BRITAIN 31

CHROMOSOME NUMBER

Root-tips of the parents and the hybrid were obtained from individuals transplanted to the University Botanic Garden and the Botany School Field Station (hybrid only) and were pretreated with either a 0-05% solution of colchicine or a saturated aqueous solution of 1—-bromonaphthalene for three hours and fixed in acetic-alcohol (1:3) overnight. Root-tip squashes were then made using the Feulgen method. Three counts from each of the two hybrid plants all gave the tetraploid number of 2n = 48. S. nigrum and S. sarrachoides were confirmed to be hexaploid (2n = 72) and diploid (2n = 16) respectively. The natural occurrence of this tetraploid hybrid is of particular interest in the light of experimental hybridization, which suggests that there may be a S. sarrachoides genome in S. nigrum (Edmonds in press).

DISTRIBUTION OF THE HYBRID

Apart from the fields in the vicinity of the Cinques, S. x procurrens has been noted in the following sites, all on the greensand:

Market-garden field, Mill Hill, south of Gamlingay, Cambs., v.c. 29, GR 52/236.510, 26 October

1975

Market-garden field, Potton, Beds., v.c. 30, GR 52/230.501, 26 October 1975, A. C. Leslie no.

1033/1975 (herb. A.C.L.)

Market-garden field, Sutton, Beds., v.c. 30, GR 52/230.577, 26 October 1975

It undoubtedly occurs elsewhere in this area and should be looked for wherever the two parents are established together; potential sites exist in the East Anglian Breckland and on the sands to the south of Guildford, Surrey, v.c. 17. .

The occurrence of S. sarrachoides at Gamlingay was first formally noted in 1972 (R. J. Pankhurst, card index of the Cambridgeshire flora in CGE), but it has evidently been established for a much longer time and it is known to the local farmers as the ‘White Nightshade’. It has a longer recorded history in Bedfordshire, being listed for Potton by Dony (1953). It may well have been originally introduced with wool shoddy, which is still used irregularly at Sutton. Another South American annual, Galinsoga parviflora Cav., 1s also well established over the whole area.

DISCUSSION

Leslie (1976) showed that most British alien material named S. sarrachoides Sendtn. is in fact S. nitidibaccatum Bitter, whilst S. sarrachoides Sendtn. emend. Bitter is a much rarer plant. Continental authors (e.g. Ooststroom & Reichgelt 1966, Ludwig 1973) have come to similar conclusions. However the relationship of these two taxa is uncertain, for, although European alien material can be clearly divided into two, the variation in this group in South America is complex and other similar taxa occur. In view of the uncertainty about the distinction between the South American taxa S. nitidibaccatum Bitter and S. sarrachoides Sendtn. emend. Bitter, in this paper S. sarrachoides refers to the complex. However, if the two are distinguished, S. x procurrens strictly refers to the hybrid between S. nigrum subsp. nigrum and S. nitidibaccatum.

ACKNOWLEDGMENTS

I should like to thank Dr J. M. Edmonds and Mr R.I.S. Brettell for the chromosome counts, Mr P. D. Sell for his thorough revision of my Latin description, and all three for their critical comments on the manuscript.

REFERENCES

Dony, J. G. (1953). Flora of Bedfordshire, p. 455. Luton. Epmonps, J. M. (In press). Taxonomic studies on Solanum L. subsect. Solanum (Maurella). Bot. J. Linn. Soc. LesLig, A. C. (1976). Solanum sarrachoides. B.S.B.I. News, 12: 13.

32 A. C. LESLIE

LUDWIG,VON W. (1973). Das sidamerikanische Solanum nitidibaccatum, ein Neubiirger in Hessen. Ber. wetterau.

Ges. ges. Naturk., 123: 67-73. OostTsTRooM, S. J. VAN & REICHGELT, T. J. (1966). Solanaceae, in Flora Neerlandica, 4: 141-177. Stace, C. A. (1975). Hybridization and the flora of the British Isles. London.

(Accepted June 1977}

Watsonia, 12, 33-40 (1978). 33

Derelict industrial land as a habitat for rare plants in S. Lancs. (v.c. 59) and W. Lancs. (v.c. 60)

E. F. GREENWOOD

Merseyside County Museums . and R. P. GEMMELL

Joint Reclamation Team of Greater Manchester and Lancashire County Councils

ABSTRACT

The importance of derelict land in S. and W. Lancs. as refugia for regionally rare and local species is demonstrated with particular reference to colonies of marsh orchids (Dactylorhiza species and hybrids). The significance of the sites for research and nature conservation is discussed.

INTRODUCTION

During the last ten years or so a number of sites containing marsh orchids (Dactylorhiza species and _ hybrids) have been found in the Watsonian vice-counties of S. and W. Lancs. (now forming parts of the _ administrative counties of Cheshire, Greater Manchester, Lancashire and Merseyside) where earlier workers (Wheldon & Wilson 1907, Savidge et al.1963) reported them to be rare. Our studies of these sites have revealed that many of the orchid-rich habitats have been formed from derelict industrial land, being either grossly modified by man or entirely man-made. In this study 35 sites containing large populations of marsh orchids were discovered. 22 of these are located in the more industrialized vice-county of S. Lancs, where only two sites, including one of a series of locations on the coastal sand-dunes, are in natural or semi-natural habitats.

Altogether 25 of the sites were found to be entirely man-made or grossly modified by man and it is the significance of these that is discussed.

THE HABITATS

In the semi-natural habitats the substrates were generally base-rich and damp, as in coastal dune- | slacks. Inland habitats were frequently old meadows or commons where, usually through grazing, scrub and woodland development had been prevented. The man-made habitats were always base-rich but not necessarily marshy or damp. They varied from old cultivated fields, reclaimed salt-marshes and disused railway cuttings to tipped industrial _ waste and abandoned clay pits. Table | lists 21 man-made sites and indicates the kind of site involved together with the date of abandonment. None of the sites existed 100 years ago and the earliest records for orchid colonization date from 1915 (Rainford Junction) and 1942 (Cop Lane, Penwortham), but most records are much more recent. It is probably worth noting, however, that at Rainford Junction | and Cop Lane it took 57 and 60 years respectively before the sites were known to be colonized by marsh orchids, whereas in the more recently abandoned sites colonization has occurred much more quickly. _Atsome sites, e.g. Marton railway cutting, colonies have been noticed less than ten years after the site had been abandoned.

G

34 E. F. GREENWOOD AND R. P. GEMMELL TABLE 1. THE ORIGIN OF SOME INDUSTRIAL SITES

Date of abandonment Site Origin (where known) ViCID9 Banks Marsh, GR 34/38.23 Reclaimed salt-marsh after building of 1895 embankment Chat Moss, GR 33/69.97 Formerly raised bog, cut over for peat and dominated by Molinia caerulea. High lime content unexplained but possibly from agriculture Cop Lane, Penwortham, GR 34/52.27 Railway cutting started for sidings 1882 Darcy Lever, Bolton, GR 34/73.07 Leblanc process waste 1920s Failsworth Golf Course, GR 34/88.01 Hollow adjacent to railway cutting —— Hart Common, GR 34/63.05 Damp hollows between colliery spoil heaps _- Haskayne railway cutting, GR 34/35.09 Railway cutting 1953/4 Hightown meadows, GR 34/30.03 Arable fields reclaimed from Link Sands, then c1968 used as hay-meadows Ince Moss, GR 34/59.02 Colliery subsidence, colliery washery waste, boiler 1960s ash, lime waste and peat remnants Kirkless Lane, Ince in Makerfield, GR 34/60.06 Blast-furnace slag and boiler ash 1930s Lightshaw Hall Flood, GR 33/61.99 Colliery subsidence Longton Brickworks, GR 34/48.25 Clay pit used for brickworks 1965 Nob End, Little Lever, GR 34/74.06 Leblanc process waste c1900 Pollard Moor, Hapton, GR 34/79.31 Lime waste and gas-lime tip c1935 Rainford Junction, GR 34/47.02 Triangle of land at railway junction 1858 Rixton clay pits, GR 33/68.90 Clay pits used for brickworks 1960s

Westwood Power Station, Wigan, GR 34/58.04 Power-station ash tipped into subsidence flash c1963 v.c.60

Broadwater, Fleetwood, GR 34/32.45 Waste land, formerly reclaimed salt-marsh —

Heysham Flash, GR 34/40.59 Old sand-dunes enclosed by construction of 1900s Heysham Harbour and tipped boiler ash

Marton railway cutting, GR 34/33.33 Marsh orchids appeared on banks when 1967 management ceased on closure of railway

Salwick Atomic Energy Works, GR 34/46.31 Alkali waste 1963/4

Of particular interest are the habitats associated with tipped waste material; Table 2 lists five principal types of waste that can be recognized. The raw, unweathered wastes from which the tip soils are derived were strongly alkaline and ranged from pH 8.5 (lime waste) to pH 12.7 (Leblanc process waste). Observations of exposed profiles of colonized Leblanc process waste (Table 3) revealed the development of rendzina soils. However, older and more weathered wastes were less base-rich due to leaching and carbonate formation from calcium hydroxide; this has allowed invasion by willow and hawthorn scrub to occur. At the same time, surface humification has encouraged the establishment of various species including legumes, particularly Lotus corniculatus, with a consequent enrichment of soil nitrogen leading to eventual formation of closed grassland communities. Such successional changes may, in time, destroy the floristic richness of the sites.

THE FLORA OF MAN-MADE SITES

The industrial habitats characterized by the marsh orchid populations contained a number of other species regarded as rare according to a recent systematic survey of W. Lancs. (Table 4). In this survey. rare species were defined as those occurring in 10 (2:6%) or fewer tetrads (2 x 2km squares of the National Grid). A similar list of rare species can be prepared for S. Lancs. using Travis’s Flora of South Lancashire (Savidge et al., 1963), but Carex pseudocyperus and Glyceria maxima are more common there whilst Eleocharis uniglumis, widespread in W. Lancs. coastal habitats, is rare. Also Dactylorhiza — incarnata and Blackstonia perfoliata are widespread on the S. Lancs. coastal sand-dunes but inland © they are rare. None of the species is nationally rare (Richards 1972), although a number are either only found occasionally or are at the limits of their distribution in Britain (Perring & Walters 1962).

DERELICT LAND AND RARE PLANTS IN S. & W. LANCS. 35

TABLE 2. THE PRINCIPAL TYPES OF WASTE WHICH PROVIDE HABITATS FOR ORCHID POPULATIONS

Type of waste

Leblanc process Lime waste and gas lime

Power station ash

Blast furnace slag and boiler ash

Colliery washery waste and slurry

Contribution to soil reaction

CagOy Ca(OH)y hy- drolysis products of CaS

CaCO,, hydrolysis of CaS

Hydrolysis of calcium sil- icate minerals

CaCO,, Ca(OH), and hy- drolysis of basic silicates

Carbonate minerals

Type of flora

Open, herb-rich, calcicolous associations. Festuca rubra dominant. Invasion by Crataegus and Salix scrub

Open, herb-rich, calcicolous associations. Festuca rubra dominant. Invasion by Salix scrub

Calcicolous vegetation under Salix scrub. Herb-rich in clearings

Very open, herb-rich, calcicolous associ- ations. Marsh vegetation on boiler ash in damp hollows

Open herb-rich associations. Invasion by Salix scrub

TABLE 3. SOIL PROFILE OF HABITAT ON LEBLANC PROCESS WASTE AFTER 60-80 YEARS EXPOSURE AND COLONIZATION

Depth (cm) pH Profile description 0-5 4) Black, surface humus 5—15 U4 Dark, partially humified waste

15-25 7:8 Yellowish-brown waste stained 25-35 8-0 with deposited ferric salts 35-45 9-4 Yellowish, partially weathered 45-55 On), waste. Ferric salts present 55-65 12:2 White, unweathered waste, 65-75 12-1 calcium hydroxide present

TABLE 4. RARE W. LANCS. SPECIES FOUND IN MAN-MADE SITES IN S. AND W. LANCS.

Epipactis palustris

Glyceria maxima

Gymnadenia conopsea

Ophrys apifera

Orchis morio

Orobanche minor

Osmunda regalis

Pyrola rotundifolia subsp. maritima

Ranunculus trichophyllus

Apium inundatum Blackstonia perfoliata Carex pseudocyperus C. riparia Dactylorhiza incarnata subspp. coccinea and incarnata D. praetermissa D. purpurella Echium vulgare

There is evidence, therefore, that man-made sites are providing additional habitats for regionally rare or localised species, especially in urbanized areas. This is particularly true for the two species

_ Dactylorhiza purpurella and D. praetermissa, which are near the southern and northern limits respectively of their ranges of distribution in Britain. Further, the occurrence of presumed hybrid

_ Swarms between these and other orchid species suggests that the presence of these sites may contribute

36 E. F. GREENWOOD AND R. P. GEMMELL

to the breakdown of isolating mechanisms between species by providing new habitats. So far the following hybrids are thought to have been found: D. fuchsii x D. purpurella, D fuchsii x D. praetermissa, D. fuchsti x D. incarnata, D. incarnata x D. praetermissa, D. incarnata x D. purpurella and D. fuchsii x Gymnadenia conopsea. Much less certain is the occurrence of D. praetermissa x D. purpurella.

TABLE 5. SPECIES CHARACTERISTIC OF ORCHID-RICH INDUSTRIAL HABITATS IN S. AND W. LANCS. (PRESENT IN > 70% OF SITES)

Angelica sylvestris Juncus inflexus Arrhenatherum elatius Lotus corniculatus Carex nigra Plantago lanceolata Centaurea nigra Ranunculus acris Cerastium fontanum Rumex acetosa Cirsium arvense Salix cinerea Dactylis glomerata Senecio jacobaea Epilobium angustifolium Taraxacum officinale Equisetum arvense Trifolium pratense Festuca rubra ; T. repens Heracleum sphondylium Vicia cracca

Holcus lanatus Hypochoeris radicata

Floristic lists, compiled for most of the sites mentioned in Table 1, have revealed several species with a high degree of constancy (Table 5). Of the grasses, Festuca rubra and Dactylis glomerata occurred in over 75% of the sites and were often the dominant or most frequent grasses, as illustrated by the three alkaline waste sites for which species lists are given in Table 6. The most constant and abundant dicotyledons were Centaurea nigra and Tussilago farfara, followed by Angelica sylvestris, Lotus corniculatus and Plantago lanceolata (Tables 5 and 6).

Most of the species found in the man-made sites are common and widespread ruderals which have colonized the sites from nearby habitats. Surprisingly, there were few aliens, although the continued presence of Sisyrinchium bermudiana at Little Lever is noteworthy in view of its rarity at inland locations.

In S. Lancs. there are few natural or semi-natural habitats away from the Irish Sea coast and these are mostly relict raised bogs. These acid ‘mosses’ covered much of the area and, although most of the man-made sites containing marsh orchids are base-rich, acid conditions were found at a few of the sites allowing calcifuge species such as Calluna vulgaris, Carex curta, Eriophorum angustifolium and Molinia caerulea to occur. Of more significance was the occurrence of Empetrum nigrum at Hart Common and Osmunda regalis and Potamogeton polygonifolius at Haskayne. At both sites, natural or semi-natural habitats have disappeared from the vicinity.

Surprisingly, a number of the plants found at the inland, man-made sites are more characteristic of coastal habitats in north-western England (Table 7). Of the species listed (Table 7), 12 are rare in S. and W. Lancs. and of these Pyrola rotundifolia subsp. maritima is especially noteworthy. This subspecies was formerly confined in Britain to the coasts of Lancashire and Norfolk but during this century, and particularly since about 1950, it has been extending its range along west coast sand-dune systems (Kay et al. 1974). The site at Haskayne is, however, only the second inland record for this subspecies. In Cheshire, Lee (1975, 1977) has similarly reported a number of coastal species from inland saline and alkali waste sites.

SIGNIFICANCE OF THE SITES

Kelcey (1975) has presented evidence that industrial development in Britain has contributed much to the continued existence of certain rare and unusual species as well as preventing some common species from becoming rare. It is difficult, however, as Davis (1976) has pointed out, to compare the overall gains and losses through industrial development, although our observations support Kelcey’s

DERELICT LAND AND RARE PLANTS IN S. & W. LANCS.

TABLE 6. DESCRIPTION OF FLORA COLONIZING LIME WASTE HEAPS AT THREE SITES IN S. AND W. LANCS.

Abundance (Domin scale)

Nob End, Species Little Lever Darcy Lever Pollard Moor Festuca rubra 4-6 6-7 5-6 Centaurea nigra 4-6 3 x-] Dactylis glomerata 4-5 4, 2-3 Tussilago farfara 3-5 4 4-5 Linum catharticum 3-4 3-4 3-4 Lotus corniculatus 3-4 1 Agrostis tenuis 2-4 1-2 1 Festuca ovina 2-4 3-4 2-3 Succisa pratensis 2-4 x-1 Sisyrinchium bermudiana 3 Angelica sylvestris 2-3 x-l 1 Centaurium erythraea 2-3 2-3 Crataegus monogyna 2-3 2-3 x-l Dactylorhiza incarnata 2-3 1 3 ~ Deschampsia cespitosa 2-3 1 X D. flexuosa 2-3 _ Erigeron acer 2-3 1-2 _ Euphrasia nemorosa 2-3 3 3 Gymnadenia conopsea 2-3 Pilosella officinarum 2-3 3-4 _ Plantago lanceolata 2-3 2 1 _ Potentilla erecta 2-3 x-1 2 _ Trifolium pratense 2-3 ] X _ T. medium 2-3 _ Aster novi-belgii 1-3 X X _ Carlina vulgaris 2 _ Dactylorhiza purpurella 2 _ Hieracium vulgatum 2 2-3 2 _ Leontodon hispidus 2 X 2 Luzula multiflora 2 1 _ Vicia cracca 2 x-] _ Achillea millefolium 1-2 X x __A. ptarmica 1-2 x _ Agrostis stolonifera 1-2 3 2-4 _ Arrhenatherum elatius 1-2 1-2 1 _ Bellis perennis 1-2 2 X Cirsium arvense 1-2 2-3 2-3 | Dactylorhiza fuchsii 1-2 X 3 Heracleum sphondylium 1-2 x-l 2-3 _ Juncus inflexus 1-2 Senecio jacobaea 1-2 1-2 1-2 _ Agrostis gigantea x—2 _ Arabis hirsuta l ~ Carex flacca 2 Epilobium angustifolium ] 2-3 Hieracium umbellatum 1-2 1 _ Holcus lanatus X X

_ Juncus effusus Orobanche minor

_ Ranunculus acris

_ Solidago canadensis _ Sonchus oleraceus

ee

38

Species

Taraxacum officinale Tragopogon pratensis Trifolium repens Vicia sepium Cerastium fontanum Hieracium glandulosum H. maculatum Hypochoeris radicata Lathyrus pratensis Plantago major

Poa pratensis

Rumex acetosella Rubus fruticosus Acer pseudoplatanus Carex hirta

C. nigra

C. ovalis

Cirsium palustre

C. vulgare Equisetum arvense Festuca arundinacea F. tenuifolia Hieracium vagum Juncus bufonius Lolium perenne Molinia caerulea Nardus stricta

Orchis morio Potentilla anglica Rumex acetosa

R. crispus

R. obtusifolius Sagina procumbens Salix caprea

S. repens

Solidago gigantea Betula pendula Calystegia sepium Chrysanthemum leucanthemum Eupatorium cannabinum Leontodon autumnalis L. taraxacoides Potentilla reptans Prunella vulgaris Salix cinerea Sambucus nigra Solanum dulcamara Sorbus aucuparia Triglochin palustris

E. F. GREENWOOD AND R. P. GEMMELL

TABLE 6—continued.

Nob End, Little Lever

a a TD a

Abundance (Domin scale)

Darcy Lever

mM rs

Pollard Moor

DERELICT LAND AND RARE PLANTS IN S. & W. LANCS. 39

conclusions. We consider that all the sites mentioned in Table | are of botanical interest because of their floristic diversity and significance as refugia for regionally rare and localized species. As such they have been included in a list of 252 sites of scientific interest prepared by the Lancashire Naturalists’ Trust. Man-made sites account for only 16% of these sites in rural W. Lancs. but for 41% of the sites in the more urban S. Lancs., where the more natural sites are largely coastal or rural.

TABLE 7. SPECIES WITH A PREDOMINANTLY COASTAL DISTRIBUTION IN S. AND W. LANCS. BUT FOUND ON INLAND MAN-MADE SITES

Species Characteristic coastal habitat Anthyllis vulneraria Fixed sand-dune Apium inundatum Wet dune-slack Blackstonia perfoliata Fixed sand-dune Carex arenaria Mobile sand-dune Dactylorhiza incarnata subspp.

incarnata and coccinea Damp dune-slack D. praetermissa Damp dune-slack Echium vulgare Fixed sand-dune Eleocharis uniglumis Brackish marsh Epipactis palustris Damp dune-slack Gymnadenia conopsea Fixed sand-dune Juncus gerardii Salt-marsh Oenothera spp. Mobile sand-dune Ophrys apifera Damp dune-slack Orobanche minor Fixed sand-dune Pyrola rotundifolia subsp. Damp dune-slack and

maritima fixed sand-dune Ranunculus trichophyllus Wet dune-slack Salix repens subsp. argentea Damp dune-slack Schoenoplectus tabernaemontani Brackish marsh

The sites also provide valuable opportunities for research, including studies of the origin and development of populations of Dactylorhiza and other species, problems of habitat management, possibilities of deliberate habitat creation and the planning of industrial activities to achieve natural resource benefits.

We conclude from our studies of industrial waste sites and disturbed areas in S. and W. Lancs. that certain types of industrial development and waste disposal provide exciting possibilities for the deliberate creation of habitats of high wildlife interest. To achieve this, new industrial sites, particularly those involving extractive and waste disposal operations, should be carefully planned with a view to habitat development and future natural history interest upon their eventual closure. How to achieve this with different industrial processes, particularly modern ones, opens up a whole new field of research into the use of landscape modelling techniques, placement of materials both during and after industrial operations, different soil fertilizer treatments and soil moisture or drainage control as methods of providing specific habitat types or a variety of conditions for colonization. Such sites provide opportunities for the documented introduction (with the prior approval of the Nature Conservancy Council) of indigenous plant taxa including ones which are isolated geographically from native populations and which do not have effective long-range dispersal mechanisms.

We believe that suggestions of this kind should be given greater attention if the conflicts between industrial development and nature conservation are to be minimized.

ACKNOWLEDGMENTS

We should like to thank many people for drawing our attention to the sites but in particular to members of the Wigan Field Club and the Lancashire Naturalists’ Trust who are now making a detailed survey of some of the orchid populations. One of us (E.F.G.) would also like to thank the British Ecological Society for financial assistance with travelling.

40 E. F. GREENWOOD AND R. P. GEMMELL REFERENCES

Davis, B. N. K. (1976). Wildlife, urbanisation and industry. Biol. Consery., 10: 249-291.

Kay, Q. O. N., RoBerts, R. H. & VAUGHAN, I. M. (1974). The spread of Pyrola rotundifolia L. subsp. maritima (Kenyon) E. F. Warb. in Wales. Watsonia, 10: 61-67.

Ketcey, J. G. (1975). Industrial development and wildlife conservation. Envir. Consery., 2: 99-108.

Lee, J. A. (1975). The conservation of British inland salt marshes. Biol. Conserv., 8: 143-151.

Leg, J. A. (1977). The vegetation of British inland salt marshes. J. Ecol., 65: 673-698.

PERRING, F. H. & WALTERS, S. M., eds (1962). Atlas of the British flora. London.

RICHARDS, A. J. (1972). The code of conduct: a list of rare plants. Watsonia, 9: 67-72.

SAVIDGE, J. P., HEYwoop, V. H. & Gorpon, V., eds (1963). Travis’s Flora of South Lancashire. Liverpool.

WHELDON, J. A. & WILSON, A. (1907). The flora of West Lancashire. Liverpool.

(Accepted June 1977)

Watsonia, 12, 41-46 (1978). 4l

The flora of walls in south-eastern Essex

R. M. PAYNE

49 Galton Road, Westcliff-on-Sea, Essex

ABSTRACT

The flora of 650 walls in south-eastern Essex is analysed, with emphasis on the relative frequency of species on walls of different kinds. Comparisons are made with other surveys of wall-flora.

AIM AND SCOPE OF THE SURVEY

During the four years 1973-1976, I kept 650 walls and series of walls in south-eastern Essex under observation, with the primary aim of discovering the relative frequency of different species of flowering plants and ferns in such habitats. The area covered was that part of S. Essex, v.c. 18, roughly east of a line running south from Chelmsford through Billericay to Tilbury.

Of the 650 walls surveyed, 278 (43%) were urban garden walls (or series of walls), 182 (28°4) were churchyard walls, 47 (7%) railway walls, 41 (6°) walls of secular buildings other than railway buildings, 44 (7%) retaining walls (including some also in the other categories) and 79 (12°) walls in relatively rural areas (other than churchyard, railway and retaining walls). River and sea walls were wholly excluded. Separate lists were made of the species found on these different types of walls.

Little attention was paid to the composition of walls, because, apart from some church walls of Kentish ragstone, almost all walls in south-eastern Essex are made of bricks. The size of the task and limitations of time precluded consideration of ecological factors (aspect, shade, adjacent vegetation, BIC.)

NOTES ON THE METHODS USED

The unit for the survey was either a single wall or a series of adjacent walls of similar construction and age. Thus in churchyards the stone fabric of the church and the brick perimeter wall of the churchyard would be treated as separate units.

Each wall included in the survey supported at least one plant at some time during the four years, and each was visited at least twice, at different seasons, though most of the walls were visited much more often.

Plants growing at the extreme base of a wall were ignored, since these would probably be rooted in the ground and therefore not truly rupestral. Similarly, wall-tops with an obvious accumulation of soil were excluded from the survey.

On garden walls, all plants not likely to have been deliberately planted were included. It was of course usually practicable to list only the plants growing on the outside of garden walls, to which species of horticultural origin would be likely to have spread by natural means.

THE FLORA IN GENERAL

A total of 286 species was recorded, of which 83 (29%) were probably or certainly of horticultural origin; included in this category are not only obvious garden escapes such as Aster cf. novi-belgii, Campanula portenschlagiana and Linaria purpurea, but also such species as Tanacetum parthenium, Fagus sylvatica, Polystichum setiferum and Sorbus aucuparia, which are either not native in south- eastern Essex or, as wall plants, are highly likely to be derived from garden specimens.

42 R. M. PAYNE

Of the 286 species, 200 (70%) were found on fewer than ten of the 650 walls, 164 (57°) on fewer than five walls, and indeed 85 (30%) on only one wall. The order of frequency of the 47 species which occurred on at least 5% of the walls is shown in Table 1.

TABLE 1. PERCENTAGE OCCURRENCES (>5%) ON ALL WALLS

* of garden origin 7 status doubtful

Poa annua 54 Phyllitis scolopendrium* 9-5 Sonchus oleraceus 35 Lamium album ot Senecio vulgaris 33 Capsella bursa-pastoris 8-9 Epilobium adenocaulon 32 Sagina apetala 8-5 Senecio squalidus 29 Epilobium angustifolium 8-3 Hedera helixt 27 Acer pseudo-platanus 8-0 Dryopteris filix-mas* 22 Anthriscus sylvestris 7-4 Antirrhinum majus* 20 Sedum acret 72 Sambucus nigra i, Hordeum murinum 6-9 Corydalis lutea* 18-3 Campanula portenschlagiana* 6:9 Mercurialis annua 18-2 Poa pratensis 6:8 Sagina procumbens 18-2 Pteridium aquilinum 6-6 Cymbalaria muralis 17-4 Lolium perenne 6:2 Lamium purpureum 16-6 Poa angustifolia 6-2 Urtica dioica 16-0 Agropyron repens 6:0 Taraxacum officinale agg. Gy) Ballota nigra 6:0 Stellaria media 14-6 Conyza canadensis 6-0 Rubus fruticosus agg. 14-5 Euphorbia peplus 6:0 Bromus Sterilis 14-3 Holcus lanatus 6:0 Veronica sublobata 14-3 Aster cf. novi-belgii* a7 Galium aparine 11-9 Centranthus ruber* SF Arrhenatherum elatius Ll Convolvulus arvensis 5:2 Dactylis glomerata 11-4 Plantago lanceolata Syl Poa trivialis 10-3

Table 2, which is based on a similar table published by Kent (1961), compares the frequency of the 20 commonest wall plants in the present survey with the frequency of some of them as recorded in wall surveys in Middlesex, the London area, Durham and Cambridge.

Poa annua stands out as the predominant species in four of these lists, but other plants show remarkable dissimilarities in frequency. Durham might be expected to have a different flora from Essex, being 350 km to the north and with a colder, wetter climate. Indeed, nine species among the commonest 14 on Durham walls (Woodell & Rossiter 1959) do not appear in Table 2: these are Epilobium angustifolium, Dactylis glomerata, Epilobium montanum, Acer pseudo-platanus, Lolium perenne, Plantago lanceolata, Poa pratensis agg., Rumex obtusifolius and Senecio jacobaea. (The last- named was found on only four walls in south-eastern Essex). Species prominent in the Cambridge survey (Rishbeth 1948) but considerably less so in south-eastern Essex are Festuca rubra, Arenaria serpyllifolia, Tanacetum parthenium, Acer pseudo-platanus, Capsella and Epilobium angustifolium. Of the two segregates of Arenaria serpyllifolia in Essex, I found A. leptoclados twice as frequently as A. serpyllifolia (as did Grose (1957) on Wiltshire walls); neither occurred on urban garden walls. Epilobium adenocaulon has clearly increased enormously in south-eastern England even in the 15 years since the Middlesex survey, when both E£. angustifolium and E. montanum were much commoner wall species. The high figures shown for Corydalis lutea reflect the greater proportion of garden walls in the Essex survey. Mercurialis annuaand Veronica hederifoliaagg., neither of which figures prominently inany of the other surveys, are abundant urban weeds in south-eastern Essex. All rupestral specimens of the latter that I examined appeared to be the segregate V. sublobata, which Kent (1975) states to be also the usual Middlesex plant.

THE FLORA OF WALLS IN SOUTH-EASTERN ESSEX 43

TABLE 2. PERCENTAGE OCCURRENCES IN VARIOUS SURVEYS OF THE 20 COMMONEST SPECIES IN THE PRESENT SURVEY

650 500 72 66

Essex Middlesex London Durham Cambridge

walls walls walls walls walls* Poa annua 54 48 SPS) 40 Gi) Sonchus oleraceus 35 18-6 1225 Senecio vulgaris 33 9:2 13-9 28 (8) Epilobium adenocaulon 32 4 Senecio squalidus 29 25 22:2 Hedera helix 24 2 Dryopteris filix-mas 22 34-8 27:8 Antirrhinum majus 20 7 4-2 (4) Sambucus nigra 19 5-6 1:4 41 (7) Corydalis lutea 18-3 2 Mercurialis annua 18-2 0-2 Sagina procumbens 18-2 17-4 oF Cymbalaria muralis 17-4 23-6 6:9 (5) Lamium purpureum 16-6 2 Urtica dioica 16 ~ Taraxacum officinale agg. S97; 12-6 11-1 64 (2) Stellaria media 14-6 3-6 Rubus fruticosus agg. 14-5 1:6 23 Bromus sterilis 14-3 2 Veronica sublobata 14-3 0-6 Total number of species 286 204 83 168 186

* order of frequency

Table 3 lists the commonest species on each of the six types of walls studied. In each column only those species which occurred on at least 20% of the walls are included.

URBAN GARDEN WALLS

150 species were recorded from garden walls in built-up areas, a surprisingly large number perhaps partly accounted for by the prevalence in the Southend conurbation of burr walls, i.e. brick walls made of clinkers, which provide a more favourable habitat for seedlings than the conventional smooth brick wall. Among native plants particularly associated with urban garden walls were Mercurialis annua (86% of its sites), Epilobium montanum (73%), Euphorbia peplus (69%), Veronica sublobata (67%) and Lapsana communis (60%); all except Lapsana are common garden weeds in this area. It is perhaps surprising that as many as 47% of the sites for Pteridium aquilinum were walls of this type, since this is neither a garden plant nor acommon garden weed. Calystegia silvatica, slightly commoner throughout the survey than C. sepium, was predominantly (86%) a plant of urban garden walls, whereas C. sepium occurred equally on urban and rural walls.

CHURCHYARD WALLS

Although churchyard walls in south-eastern Essex tend to be rich in numbers of species (160 species in total), I have not found such a marked disparity between these and other walls as Kent (1964) noted in Middlesex, where he ascribed the disparity partly to a colonizing of churchyard walls from relict populations of wild plants still growing in the churchyards though extinct in the surrounding urban areas. This is probably because even now south-eastern Essex is, on the whole, more rural than Middlesex. Nevertheless, the following 13 species were seen only on churchyard walls during the present survey: Cynosurus cristatus, Fragaria vesca, Fraxinus excelsior, Ligustrum vulgare, Oxalis

44 R. M. PAYNE

exilis, Potentilla sterilis, Ranunculus repens, Raphanus raphanistrum, Ribes nigrum, Sonchus arvensis, Tussilago farfara, Urtica urens and Veronica agrestis. In addition, five species were found mainly on churchyard walls: Glechoma hederacea (84% of its sites), Symphoricarpos rivularis (75°%), Parietaria diffusa (67%), Potentilla reptans (61%) and Veronica chamaedrys (57%).

I am uncertain to what extent Hedera helix is a genuine rupestral plant; I endeavoured to record it only where it appeared to be rooted in a wall, but whether it can survive indefinitely without contact with the ground may be doubted.

It is interesting to compare the fern flora of these Essex churchyard walls with that of similar sites in Norfolk, which were investigated by Silverwood (1965). The most striking difference is the much greater frequency of Polypodium in Norfolk. In his survey, Silverwood found ten species of ferns, of which Polypodium vulgare agg. was slightly more numerous than Dryopteris filix-mas, the other species in order of decreasing frequency being Phyllitis scolopendrium, Asplenium adiantum-nigrum, Asplenium ruta-muraria and Asplenium trichomanes, with Polystichum aculeatum, P. setiferum, Ceterach and Pteridium all very scarce. In the rather drier and much more polluted south-eastern Essex, I found eight species on church walls, of which Dryopteris filix-mas was much the commonest, followed by Phyllitis; considerably rarer were Asplenium adiantum-nigrum, A. ruta-muraria, Polypodium and Pteridium, with Asplenium trichomanes and Dryopteris dilatata very scarce. As I suggested in an earlier paper (Payne 1960), the retention of its fronds by Polypodium throughout the year may make it less able to withstand pollution than some other ferns; sulphur dioxide pollution is known to be more intense in the autumn and winter months.

TABLE 3. PERCENTAGE OCCURRENCES (> 20%) ON DIFFERENT TYPES OF WALLS

Urban Church- Railway Retaining Rural Walls of Garden yard Walls Walls Dividing Buildings

Walls Walls Walls Poa annua ip 26-4 39 73 cal Senecio squalidus 36:3 45 36 ~28 46 Sonchus oleraceus 50 26 pis) 43 24 Epilobium adenocaulon 36-0 32 47 4] Senecio vulgaris 48 20 23 43 Hedera helix 52 21 25 28 Dryopteris filix-mas 73) 24 56 Arrhenatherum elatius 24-7 30 32 Urtica dioica PESRB) 23 37 Lamium purpureum WS 32 2 Rubus fruticosus agg. Di 30 D3 Sambucus nigra 25:8 24 Zp Dactylis glomerata 23 25 23 Bromus Sterilis 20 28 Taraxacum officinale agg. 28 24 Lamium album 2)| 23 Antirrhinum majus 4] Corydalis lutea 38 Mercurialis annua 36:3 Cymbalaria muralis 29 Capsella bursa-pastoris 28 Galium aparine 27 Pteridium aquilinum 27 Phyllitis scolopendrium 26 Sagina procumbens 22 Veronica sublobata DD Poa angustifolia 21 Sagina apetala 20

Stellaria media 20

THE FLORA OF WALLS IN SOUTH-EASTERN ESSEX 45 RAILWAY WALLS

Railway walls proved to have amore distinctive flora than those of churchyards. Ofa total of 110 species, the following 15 were found only on railway walls (which comprise both dividing and retaining walls, as well as walls of stations and other railway buildings): Bromus ramosus, Cardamine flexuosa, Ceterach officinarum, Epilobium lanceolatum, Equisetum arvense, Heracleum sphondylium, Hieraceum perpropinquum, Inula conyza, Pastinaca sativa, Senecio erucifolius, Sinapis arvensis, Smyrnium olusatrum, Solidago canadensis, Leucanthemum vulgare and Triglochin maritimum (on a bridge over a brackish dyke). In addition, 75% of Stellaria holostea sites were railway walls, as were 69% of sites for Asplenium trichomanes and 60% for Artemisia vulgaris.

Senecio squalidus is revealed as the railway wall plant par excellence, but a remarkable absentee from this column of Table 3 is Poa annua, which was recorded from only 13% of railway walls. Yet grasses in general, mainly perennial species, were found to occupy six of the top 22 places, compared with only two (both annuals) in the overall list, doubtless because of the proximity of grassy railway banks to most of these wall sites. Ferns are also prominent, Dryopteris filix-mas surprisingly ranking below four other species. Asplenium adiantum-nigrum, though occurring in slightly fewer sites than the other two Spleenworts, is much more numerous where it does occur, hundreds of plants growing on the brick walls of several bridges over the derelict Woodham Ferrers—Maldon line which opened in 1889 and closed in 1939, an interesting parallel to Cambridgeshire, where Walters (1969) records an old railway wall containing ‘hundreds of plants ... more than all the other Cambridgeshire sites put together.’

Of the 47 walls in this category, ten were on derelict lines, and on the whole these had the more varied flora, presumably because they were subject to less disturbance; indeed, every piece of railway brickwork examined on the long-abandoned Maldon line produced plants, whereas very many walls and bridges on active lines were quite barren (and therefore excluded from the survey).

RETAINING WALLS

This category, with a total of 114 species, overlaps with railway walls, and very slightly also with churchyard and garden walls. It supported ten species not found on dividing walls: Bromus ramosus, Cardamine flexuosa, Carex divulsa, Ceterach officinarum, Equisetum arvense, Heracleum sphondylium, Polygonum cuspidatum, Senecio erucifolius, Smyrnium olusatrum and Leucanthemum vulgare. Eight of these were in fact confined to railway retaining walls. As Woodall & Rossiter (1959) remarked, it is difficult to determine whether a plant on a retaining wall has extended its roots through to the soil behind. Equisetum arvense and Polygonum cuspidatum must be particularly suspect here.

RURAL DIVIDING WALLS

This category comprises walls surrounding country estates and large gardens in semi-rural areas, farmyard walls and other walls outside the built-up areas. It doesnot include any churchyard, railway or building walls.

This is another category with a rich (154 species) and distinctive flora. No less than 18 species were found only on these walls, those marked * being of horticultural origin: Asparagus officinalis* , Cerastium semidecandrum, Erodium cicutarium, Fagus sylvatica*, Geranium dissectum, G. lucidum, G. pusillum, Matricaria recutita, Papaver argemone, Pentaglottis sempervirens*, Phleum bertolonii, Ribes rubrum, Rubus laciniatus*, Spergularia rubra, Taxus baccata*, Tragopogon pratensis, Verbascum nigrum and Vinca minor*. A further eight species were found predominantly on these rural walls: Agrostis tenuis (64% of its sites), Arabidopsis thaliana (63%), Erophila verna (75%), Hypochoeris radicata (57%), Ilex aquifolium (67%), Saxifraga tridactylites (80%), Sisymbrium officinale (54%) and Vulpia bromoides (717).

WALLS OF SECULAR BUILDINGS (other than railway buildings)

Dryopteris filix-mas is the commonest species on walls of this category (80 species in total). Pteridium is also frequent and, together with Buddleja davidii, Epilobium angustifolium, Lolium perenne, Hordeum murinum and Conyza canadensis, was more commonly seen on these than on other walls.

46 R. M. PAYNE ACKNOWLEDGMENTS

I am grateful to Mr D. H. Kent who kindly read the paper through while it was in manuscript, and to Mr C. E. A. Andrews who identified Hieracium perpropinquum for me.

REFERENCES

Grose, D. (1957). The flora of Wiltshire. Devizes.

KENT, D. H. (1961). The flora of Middlesex walls. Lond. Nat., 40: 29-43.

KENT, D. H. (1964). Notes on the flora of churchyard walls in Middlesex. Lond. Nat., 43: 13-16.

KENT, D. H. (1975). The historical Flora of Middlesex. London.

Payne, R. M. (1960). The ferns of Epping Forest. Lond. Nat., 39: 25-31.

RISHBETH, J. (1948). The flora of Cambridge walls. J. Ecol., 36: 136-148.

SILVERWOOD, J. H. (1965). Ferns on church walls in Norfolk. Proc. bot. Soc. Br. Isl., 6: 120.

Wacters, S. M. (1969). Cambridgeshire ferns—ecclesiastic and ferroviatic. Nature Cambs., 12: 22-25. WOOoDELL, S. R. J. & Rossiter, J. (1959). The flora of Durham walls. Proc. bot. Soc. Br. Isl., 2: 257-273.

(Accepted June 1977)

Watsonia, 12, 47-52 (1978). 47

Short Notes THE DISTRIBUTION OF CAREX DIGITATA L. IN BRITAIN

Carex digitata L. and C. ornithopoda Willd. (section Digitatae), like C. montana L. of the related section Montanae, are Eurasian sedges that in England reach the north-western limit of their range. The centre of that range, however, is more northern and eastern (i.e. more continental) than that of C. montana. C. digitata and C. ornithopoda are found in Norway, Finland, and northern Russia, and penetrate further into Siberia, while not approaching the Atlantic or the Mediterranean so nearly.

The distribution of Carex digitata in Britain 1s as disjunct as that of C. montana (David 1977) but is more easily explicable, for the requirements of this sedge may be precisely defined: a soil with a high calcium content (pH so far measured ranges between 7:29 and 7-99, with one reading at 6:32), and good drainage yet with some protection against drying out. These conditions are exactly met in the beechwood ‘hangers’ on the Gloucestershire oolite, where the plant is widespread and plentiful in a narrow band from Birdlip in the north-east to Dursley in the south-west, and again, at least formerly, on the southern edge of the North Yorkshire moors from Thirsk eastwards to Scarborough. It is also frequent on shaded pavements of the mountain limestone, for example in the Wye valley and in Silverdale, and occurs, very sporadically, here and there on the limestone ridges of the Midlands. There is a single record from the southern chalk.

In most of its British stations Carex digitata is at risk. Jackdaw Crag at Boston Spa, Aysgarth ‘Freeholds’, and the Avon gorge at Bristol have become public scrambling-grounds. The pavements of Silverdale have been (and still are being) broken up to supply profitable rockery stone. In North Yorkshire, afforestation with conifers, and still more the widening of tracks to facilitate access for the foresters, have reduced the area available to the sedge. Even under natural conditions the crumbling banks favoured by it are a precarious habitat, while the trees or scrub whose shade prevents it from drying out may at last so overshadow it that it cannot flower, or may choke it altogether. It would seem that in any case the individual plants are short-lived as compared with the steadily maturing tussocks of C. montana or of C. humilis, but set seed more certainly and more freely than the former and much more so than the latter. The seed may lie dormant until the appropriate conditions reappear, and this may partly explain the variations in the estimates of particular populations from one year to another and the reappearance of the sedge in localities where it has been given up for lost. The results of the present survey indicate that C. digitata is commoner in Britain than is usually believed. The reason for this belief is that the sedge is easy to overlook when not in flower; even where it flourishes it may be confined to one or two isolated colonies, and such a colony, containing perhaps over a hundred plants, may occupy no more than a few square metres in a terrain that is often difficult to explore. Carex digitata should be looked for on banks and cleared slopes, usually but not invariably south-facing; a favourite habitat is the steep side of a woodland ride where rather more light penetrates than in the wood itself. The plant is best observed in early April when the new shoots, emerging from the overwintered crown, present a characteristic fountain-like outline resembling the green tufts on the heads of the pineapples so beloved by the Rococo decorators. They also show a striking colour-contrast between their pale olive-green and the dark bronzy green of the old leaves. At that season, too, the commoner sedges of the habitat, C. flacca and C. sylvatica, have made no such pronounced growth.

The known British stations are listed below. All that can be located have been surveyed since 1970, and letters indicate the present size of each population: A = | to 20, B = 21 to 100, C = 101 to 1000, D = over 1000. Estimates are approximate, for a reason opposite to that which makes a census of C. montana difficult: C. digitata never makes the composite mats of that species, but the individual plants are often scattered and small so that some are likely to be missed. Where the sedge has not been refound, the date of the last known sighting and the location of herbarium specimens whose authenticity has been confirmed are given.

N. Somerset, v.c. 6: 31/5.7, Leigh Woods (B); 31/7.5, Hinton Abbey, 1838, CGE.

N. Wilts., v.c. 7: 31/8.6, Box, 1869, BM; 31/8.7, Lucknam, 1861, BM; Colerne Park, extinct c 1960 (Horton 1975), BM, CGE, K; Slaughterford, 2 places (B, B). ‘Box’ and ‘Lucknam’ may refer to one of the other stations, but there is still suitable ground for the sedge at least at Box.

48 SHORT NOTES

Dorset, v.c. 9: 30/8.8, Wool, 1912, BMH.

E. Gloucs., v.c. 33: 32/8.0, Toadsmoor Valley, 2 places (B, C); Slad Valley, 2 places (C, C); Painswick (D); 32/8.1, Cranham (C); 32/9.0, ‘between Bisley and Daglingworth, 1955’, record in Biological Records Centre, no specimen traced; 32/9.1, near Cranham, 2 places (B, D); 32/9.2, Thrift Wood, and 42/0.2, Whittington Wood (Notcutt 1862). Specimens from ‘near Cheltenham’ in BIRM, K, OXF. Thrift Wood is much denuded since cattle had access and Whittington Wood was largely grubbed c 1900.

W. Gloucs., v.c. 34: 31/5.7, Clifton (A); 31/5.9, Pen Moel, 4 places(A, A, B, B);31/7.9, Dursley (C); 31/8.9, Nailsworth (A); 32/5.9, Symonds Yat (B); 32/8.0, Woodchester, 3 places (C, C, C); Roborough, Amberley (Riddelsdell et a/. 1948), no specimens traced.

Mon., v.c. 35: 31/5.9, Wyndcliff (B); Blackcliff (B); Itton, 1852, BM, K. Itton is off the limestone and the record may refer to one of the two preceding stations.

Hereford, v.c. 36: 32/5.1, Dowards, 2 areas (C, D); Coppet Hill (Purchas & Ley 1889), no specimen traced; 32/5.3, Capler (B).

Worcs., v.c. 37: 32/7.7, Bewdley North Wood (Lees 1867), MANCH ‘Bewdley 1857’. Wood grubbed c 1970.

Salop, v.c. 40: 33/6.0, Tickwood (B).

Notts., v.c. 56: 43/5.6, Pleasley Wood, 1838, CGE, K.

Derbys., v.c. 57: 43/1.7, Monsal (B); Taddington (B); 43/2.7, Calver (B).

W. Lancs., v.c. 60: 34/4.7, Cringlebarrow, 1904 (Wheldon & Wilson 1907), no specimen traced; Eaves Wood, 2 places (A, C); Gaitbarrows, 2 places (A, C); Leighton Beck (B, not seen by R. W. D.).

N.E. Yorks., v.c. 62: 44/5.5, Clifton Ings, 1860, BM, may be the same as ‘near York, N. Jacke’ in OXF. This water-meadow is an impossible place for C. digitata but the second York record suggest that there may be something other than a mislocation behind the first; 45/5.8, Hawnby (B); Rievaulx (B); Ouldray (A): 45/5.9, Laskill, 1860, BM, now intensely afforested; 45/6.8, Sleightholme Dale, 2 places (A, C); 45/9.8, Ayton (C); 45/9.9, “Hackness’, record in Biological Records Centre, is probably the same as the last, for Hackness itself is off the limestone.

S. W. Yorks., v.c. 63: 43/5.8, Anston Stones (A); 43/5.9, Roche Abbey, 1949 (Sledge 1950), BM, CGE, K; 44/5.0, Levitt Hagg Wood, 1844, CGE.

Mid-W. Yorks., v.c. 64: 34/7.7, Settle, 1834, OXF, 1859, BM; 44/2.6, Mackershaw (C); 44/2.7, Tanfield (A); 44/4.4, Thorp Arch (Boston Spa), 1927 (W. A. Sledge in litt. 1977), BIRM, BM, CGE, K.

N. W. Yorks., v.c. 65: 44/0.8, Aysgarth, 1887, BM.

Westmorland, v.c. 69: 34/3.8, Roudsea Wood (C); 34/4.7, Eggarslack (A, not seen by R. W. D.); Arnside Knott, 1894 (Wilson 1938), no specimen traced; Middlebarrow, 2 places (B, C); Slackhead, 3 places (A, B, B); Beetham Fell, 5 places (A, A, B, B, C); 34/4.8, Sandside (B); Witherslack, 2 places (B, B); 34/4.9, Scout Scar (Wilson 1938) and 34/5.7, Hutton Roof (Baker 1885), no specimens traced. In both localities C. ornithopoda Willd. may have been taken for C. digitata. The finder at Scout Scar, G. E. Martindale, reported that both species were present, but his specimens in LIV and MANCH are all C. ornithopoda, correctly named. I. Hindson’s record from Hutton Roof is pre-1872 and C. ornithopoda was not identified in Britain until 1874. On the other hand, all but one of the records from Slackhead, Beetham, and Sandside were first reported as C. ornithopoda in a survey of limestone pavements recently carried out by the Institute of Terrestrial Ecology.

REFERENCES

BAKER, J. G. (1885). A Flora of the English Lake District, p. 223. London.

Davipb, R. W. (1977). The distribution of Carex montana L. in Britain. Watsonia, 11: 377-378.

Horton, P. J. (1975). Postscript to D. Grose’s ‘A botanical survey of Colerne Park’, in STEARN, L. F., ed. Supplement to the Flora of Wiltshire. Devizes.

Legs, E. (1867). The botany of Worcestershire, p. 57. Worcester.

Notcutt, W. L. (1862). Minutes of the Cheltenham Working Naturalists’ Association (in Cheltenham Public Library).

Purcuas, W. H. & Ley, A. (1889). A Flora of Herefordshire, p. 321. Hereford.

SHORT NOTES 49

RIDDELSDELL, H. J., HEDLEY, G. W. & Price, W. R. (1948). Flora of Gloucestershire, p. 515. Cheltenham. SLEDGE, W. A. (1950). Plant records. Naturalist, Hull, 1950: 32.

WHELDON, J. A. & Witson, A. (1907). The flora of West Lancashire, p. 310. Liverpool.

Wison, A. (1938). The flora of Westmorland, p. 249. Arbroath.

R. W. DaAviIp

CYTISUS STRIATUS (HILL) ROTHM. IN BRITAIN

A few shrubs subsequently identified by B. E. Smythies as Cytisus striatus (Hill) Rothm. were found by me in July 1973 growing with C. scoparius (L.) Link on a roadside bank by the village of Milton, near Drumnadrochit, Easterness, v.c. 96. They differ from the latter species chiefly in the less flattened pods which are densely covered with greyish hairs. Flowering specimens, collected in May 1976, were quite difficult to distinguish from C. scoparius except for the slightly paler yellow corolla and the remains of last year’s pods still persisting here and there. Specimens have been deposited in BM.

C. striatus is a native of Portugal and Spain and has almost certainly been planted at Milton. A. O. Chater (in litt. 1977) reports that this species (det. B. E. Smythies) was planted on a new earth slope above the A487 road between Aberystwyth and Penparcau, Cards., v.c. 46, about 1970, and has since spread so that it is scattered along c 400m of the length of the slope. It flowers and fruits there abundantly.

It is possible that C. striatus has been overlooked in this country and has become naturalized in

similar habitats elsewhere. Uke DUNCAN

CEPHALANTHERA DAMASONIUM (MILL.) DRUCE x C. LONGIFOLIA (L.) FRITSCH

This hybrid has been recorded in France and Germany under the name Cephalanthera x schulzei Camus, Berg. & A. Camus (Camus et a/. 1908); the purpose of this note is to record what appears to be its first known occurrence in the British Isles.

Two plants apparently of this hybrid were found in May 1974, growing on the borders of eastern Hampshire in a woodland site which carries a thriving population of both Cephalanthera damasonium (Mill.) Druce and C. longifolia (L.) Fritsch. The plants were of a similar size, equal in height to surrounding plants of both putative parents. The lower leaves resembled those of C. damasonium, being borne in two ranks, ovate and well ridged, but the upper stem leaves were more elongated. These graded into the bracts, which in the upper flowers were narrow, linear and equalled the ovary.

The larger plant bore ten well-spaced flowers carried more or less parallel to the stem. The flowers were pure white in colour, like those of C. longifolia, with long pointed outer perianth segments, and tended to open fairly well, exposing the yellow epichile of the labellum. This yellow colour closely resembled the colour of the labellum of C. damasonium, but the epichile had three ridges on it, whereas in C. damasonium there are usually five.

One of the most marked features was the conformation of the ovary. In the flowers of C. damasonium examined, the ovary averaged 20mm in length, being broader just before its distal end. The ovary itself was not twisted. In the flowers of C. longifolia the ovary averaged 12mm in length, being narrow and more or less cylindrical. It was twisted anticlockwise through 180°. In the flowers of the putative hybrid the ovary averaged 20 mm in length, being narrow and cylindrical. It was twisted anticlockwise through 180°.

Measurements were taken of the ovary and outer perianth segments of ten flowers each of both the species and of the putative hybrid:

_C. damasonium— Ovary 20 mm, broader near tip, untwisted; outer perianth segments 22 mm x 9 | mm maximum; _C. longifolia— Ovary 12 mm, narrow and cylindrical, twisted through 180° anticlockwise; outer

perianth segments 16 mm x 5 mm maximum;

50 SHORT NOTES

Putative hybrid— Ovary 20 mm, narrow and cylindrical, twisted through 180° anticlockwise; outer perianth segments 25 mm x 7 mm maximum. The two plants recorded here flowered again in May 1975, when a third plant of similar morphology was found nearby. A set of coloured photographs of the hybrid plants is preserved by us.

REFERENCE

Camus, E. G., BERGON, P. & Camus, A. (1908). Monographie des Orchidées de l'Europe, de l'Afrique septentrionale, de l’Asie Mineure et des Provinces Russes transcaspiennes, p. 439. Paris.

D: GC. LANG & J. -bsS. TANSEY

SPINELESS-FRUITED CERATOPHYLLUM

An investigation of the genus Ceratophyllum in the Ouse levels above Newhaven, E. Sussex, v.c. 14, has revealed some curious features. The Spineless Hornwort, C. submersum L. (leaves obscurely denticulate, 3 or 4 times forked), is rare there, but where it does occur it fruits freely and its fruits are warty and spineless and easily recognizable. In contrast the very abundant Hornwort, C. demersum L. (leaves spinose-denticulate, twice forked), seems to fruit hardly at all, even in the hot summer of 1976 when the plants of these ditches did so well. Eventually, however, in September, a short stretch of water, a few yards long, produced C. demersum with well-ripened fruits; these were scarcely warty and smaller than those of its relative, but without the spines at the base which, most of the literature says, are their distinguishing feature. It seemed curious, too, that only the one small area produced plants with fruit.

Inspection of the material of Ceratophyllum in BM confirmed these observations. Specimens of C. submersum there bear plenty of fruits, but those of C. demersum bear relatively few. Among these, a number are cracked and crumbling, and it is impossible to distinguish the character of the fruit; one or two appear to be spined and one collected by E. Milne-Redhead in Beds., v.c. 30, in 1946, is clearly so. But the majority, like those from the Ouse ditches, while corresponding with the description of C. demersum in every other way, have fruits which are totally lacking in spines.

One such specimen, collected at Castle Morton, Worcs., v.c. 37, in August 1914, has attached the following comment by C. E. Moss: ‘This plant is in my opinion C. demersum var. apiculatum = C. apiculatum Cham... . The var. apiculatum is intermediate between C. demersum and C. submersum and it is somewhat arbitrary to refer it to one of the species rather than the other.’ In a way more revealing is an undated page of a letter from R. Brown to A. Bennett on a sheet of C. demersum from Lydiate, 10 - miles north of Liverpool, S. Lancs., v.c. 59: ‘Last July. . . it was fairly well in fruit but with only a specimen here and there with the two spines developed at the base. Last Tuesday I paid another visit. . . expecting to find any quantity of ripe fruit but to my astonishment found matters asleep as if no progress had been made in the two months. In fact I hardly could find any fruit with the developed spines. The plant agrees in all particulars with C. demersum.’ This letter evidently refers to a meeting of the Liverpool Naturalists’ Field Club; their Proceedings (Brown 1888, p. 25) says: “A very small quantity of the fruit appeared to develop the two basal spines. In most cases it seemed to ripen without these appendages.’ The writer remarks further that in 1887 a third Ceratophyllum ‘has been added to the British flora by Mr Alfred Fryer who has discovered C. apiculatum in a ditch by Earith Stanch in _ Huntingdonshire ... The peculiarity of its fruit is that at the base are two minute blunt tubercles instead of spines.’

The note by Scannell (1976) on the fruiting performance of C. demersum in the Grand Canal near Dublin in the summer of 1975 states that, at the end of that warm season, ‘the canal fruits were not in accord with the published descriptions “‘warty, beaked with spines”’,’ and goes on to suggest that the reason for this may be that the fruits were not properly matured. She also quotes the remarks of H. B. — Guppy, referring to the hot summer of 1893, to the effect that the plant needs heat to mature its fruits and only does so in ‘the superheated water of shallow pools’.

SHORT NOTES il

The fruiting specimens of C. demersum from the Ouse washes, of which I have deposited a voucher specimen in BM, were not in shallow water, but in water 2—3 feet deep. Nor did they show the basal tubercles which are suggested as distinguishing characters of C. apiculatum.

While there is an obvious need for further investigation of fruit production and morphology in Ceratophyllum, one may conclude that the presence or absence of spines is clearly not a good character to use 1n separating C. demersum and C. submersum.

ACKNOWLEDGMENTS

I am indebted to Miss M. J. P. Scannell for help with the literature references and to Mr E. J. Clement for reading my draft and suggesting amendments.

REFERENCES

Brown, R. (1888). Additional localities to the Flora of Liverpool. Proc. Liverpool Nat. Fld Club, 1887: 22-27. SCANNELL, M. J. P. (1976). Ceratophyllum demersum L. and fruit performance. Jr. Nat. J., 18: 348-349. E. G. NORMAN

ISOETES ECHINOSPORA DURIEU NEW TO NORTHERN ENGLAND

Isoetes echinospora Durieu was located for the first time in Cumberland, v.c. 70, on December 29th, 1974, when some detached leaves and uprooted plants were recognized in strand-line drift at Cogra Moss reservoir (GR 35/09.19). This is also the first record from England north of Dorset. The identification was confirmed by A. C. Jermy. Jsoetes lacustris L. is abundant there and counts of drift samples suggest a relative abundance of about 10:1, assuming equal susceptibility to dislodgement. /. echinospora grows under 10 cm or more of water, with J. /acustris, either in patches with plants almost touching, or singly among the abundant Juncus bulbosus L. and Eleogiton fluitans (L.) Link. Most of the plants in shallow water are small and stocky, with spreading, curved leaves up to 5 cm long; others have straight, erect leaves up to 13-5cm long. Accompanying plants of J. /Jacustris are similar in size, but reach 23 cm in deep water.

Since 1974, most of the lakes and tarns in Cumberland and Westmorland, v.c. 69, have been visited without any further sites for I. echinospora having been found. It is clear that the plant is extremely scarce in this area and may well be confined to this single station. Whether this is so or not, its presence there presents an interesting problem. The nearest known locality for J. echinosporais in North Wales*, 150 km to the south; the Scottish Highland sites are 200 km or more to the north (Perring & Walters 1962). The reservoir is completely artificial and is the result of a dam built in about 1880 across what then was a boggy valley through which Moss Beck flowed. The nearest standing water was and is 3 km distant at Ennerdale Water to the south and Loweswater to the north-east.

Isoetes spreads by the dissemination of the spores and this is clearly an efficient method of ensuring transport between waters. Birds, such as swans and the vegetarian ducks, and fish eat the leaves and many spores must be ingested. Spores could also be carried externally by birds. The megaspores of both species are covered with rough protuberances and cling readily to feathers, particularly to the downier body feathers. Most waterfowl on the Cumbrian tarns are winter visitors, so that if J. echinospora was brought in by birds it is likely to have come from the north. It is relevant that the plant is widespread in Scandinavia, whence come many of our wintering ducks, and frequent in Iceland and the Scottish Highlands, on the route of our wintering swans.

I. echinospora may be overlooked because of inadequate or inaccurate descriptions in British Floras, and a note on field characters may be useful. Most works describe J. /acustris in some detail and then give the differentiating features of J. echinospora, usually in terms of plant size, number of leaves, leaf posture, and megaspore colour. Although J. lacustris is on average larger, and usually much larger, there is a considerable overlap. The posture of the leaves (spreading or erect) has no specific

* Isoetes echinospora has recently been found in Wigtown, v.c. 74. EDS.

52 SHORT NOTES

significance, depending solely on the situation of the plant. The number of leaves is also variable and quite unreliable as a character. The megaspores, in their natural (wet) state, may be off-white or yellowish in either species, or darker if stained or muddy. When allowed to dry they suddenly turn white, a quality of the siliceous material in the spore walls, and I have found no specific colour difference. Lid (1963), in his excellent account, describes the megaspores as being greyish-white in J. lacustris and chalky-white in J. echinospora. The ornamentation, respectively of warty excrescences or of spines, is the best distinction, but is not a field character. They also differ in size and this is evident, with experience, even without optical aids.

A preliminary character for locating J. echinospora in the field is the grass-green colour of the leaves. The inner immature leaves of J. Jacustris may be as pale a green, but most are much darker. The leaves of the latter species are stiff and brittle, while those of J. echinospora are supple—the one good character given in most descriptions. Lid (1963) describes the leaves of J. echinospora as being stiff at the base, but adhering like a brush when taken out of the water. Young slender leaves of J. /acustris may occasionally adhere and old, outer leaves of J. echinospora, if spreading, may not, but otherwise this is a valid and useful field character.

The best, and a definitive, character is the leaf-shape. This is mentioned by Lid, but not in any English text. The leaves of J. /Jacustris vary considerably in length, thickness and degree of ridging, but in all cases do not taper much for most of their length, narrowing in the last half centimetre or so to an asymmetric point. In contrast, the leaves of /. echinospora taper gradually and evenly throughout to a long, hair-like tip. This is easy to see and is absolutely distinctive in all fresh specimens examined, but not when the plant is dried.

REFERENCES

Lip, J. (1963). Norsk og Svensk Flora, p. 53. Oslo. PERRING, F. H. & WaLtTers, S. M., eds (1962). Atlas of the British flora, p. 2. London. R. STOKOE

Watsonia, 12, 53-58 (1978). 53

Book Reviews

The life of Joshua Gosselin of Guernsey, 1739-1813, Greffier and soldier, antiquary and artist, plantsman and natural historian. David McClintock. Pp. 32, with 5 illustrations. Toucan Press, Guernsey. 1976. Price 75p.

It seems a long time since we had a major biographical resuscitation as a by-product of local Flora work. Apart from Readett on Henry Holden, have there in fact been any since Kent on John Blackstone and Edees on Robert Garner?

Now Mr McClintock has restored another ‘lost’ figure: Joshua Gosselin, known hitherto merely as the author of a lengthy list published in 1815 as an appendix to W. Berry’s History of Guernsey. Begun in 1788, when Gosselin was 41, this runs to no fewer than 528 species (473 of them vascular plants) and constitutes a remarkably comprehensive account of the island’s flora for so early a period. What is more, as has emerged in only the last few years, it is backed up by an extensive collection of voucher specimens with localities (though the localisation is not to particular identifiable gatherings). ‘No other area, writes Mr McClintock with justifiable pride, ‘surely has so magnificent an early start for the recording of its wild flowers.’

Till now no account of Gosselin has extended ‘to more than ten lines of print’. Mr McClintock found there was copious material in family archives, which he has duly quarried with commendable thoroughness, and this 32-page booklet is the result. Local historians must count themselves fortunate that the subject came to the attention of the one person all-familiar with the botanical background and thereby alive to the full scientific implications (though they will grumble at his failure to observe scholarly convention in particularizing his manuscript sources, not least those run to earth in that largely trackless wilderness, the Public Record Office).

Greffier (or Clerk of the Royal Court) as well as an officer in the militia, Gosselin comes to life as a man of wide culture and accomplishments who was meticulous in all he undertook. Considering that he mainly relied on a copy of the utterly antiquated Parkinson, it is a measure of the care and discernment that he brought to his study of the island’s plants that he proves to have missed astonishingly little. Would that his like had existed in every local area at that period!

An unexpected feature, which enhances his interest, is the re-emergence of an enthusiasm for plants among some of his latter-day descendants. One of these was a flower painter of distinction, Charlotte Trower, bequeather of the Trower Fund, which for many years enriched the reports of the Botanical Exchange Club with otherwise prohibitively costly plates; another is a leading horticultural writer, Mrs Betty Massingham. And from a throwaway reference in the Introduction to ‘my cousin Mrs Joan Gosselin’ it would appear that the lineage extends, very fittingly, also to the author himself.

It is sad that the publisher has let the author down with such poorly-reproduced illustrations. But it is a disaster that he has compounded this with a bibliographical enormity: a title on the cover sharply different from the one that appears on the title-page. Confronted with both, who will ever know which of them to cite?

D. E. ALLEN

Index to European taxonomic literature for 1970. Compiled by D. H. Kent & R. K. Brummitt. Pp. 215. Published in conjunction with the Flora Europaea Organisation by the Bentham-Moxon Trustees, Kew. 1977. Price £3:00. Obtainable from the Secretary, Bentham-Moxon Trust, Royal Botanic Gardens, Kew, Richmond, Surrey, U.K.

The Index to European Taxonomic literature was started by Dr Brummitt in 1970, and five parts of it appeared as volumes of Regnum Vegetabile covering the years 1965-1969. In 1971 this Index was subsumed under The Kew record of taxonomic literature, compiled by the staff of the Royal Botanic

54 BOOK REVIEWS

Gardens and closely following the style of the earlier Index but extended to cover the whole world. Although four volumes of the Kew Record have now appeared, the final volume of the Jndex, covering taxonomic literature relevant to Europe published in 1970, has, infuriatingly, had to remain unpublished because of financial problems. With the aid of a grant from the Flora Europaea Organisation, the Bentham-Moxon Trustees have now been able to publish it. The volume is very neatly and readably produced from typed copy and, within these limits, follows the style of the earlier volumes; it matches them exactly in format. It should be bought by anyone who has either the previous volumes of the /ndex, or the succeeding volumes of the Kew Record, or both. The price of £3 is less than one eleventh of the price of the latest volume of the Kew Record, which contains only three times as many entries; it is also less than the price of the 1969 Jndex on its publication in 1972. It is a pleasure to see the work of the indefatigable compilers at last available, and to know that a gap will now be cheaply and satisfactorily filled on the shelves of, one hopes, every botanical library of significance in Europe and beyond.

A. O. CHATER

Dictionary of British and Irish botanists and horticulturists. R. Desmond. Pp. xxvi + 747. Taylor & Francis Ltd, London. 1977. Price £40-00.

Until recently, one of the most useful tools in any botanical library was the Biographical index of deceased British and Irish botanists, generally referred to under the surnames of the two compilers ‘Britten and Boulger’. However, as the last edition was published as long ago as 1931 it became increasingly clear that the volume badly needed bringing up to date. With the publication in April of last year of Ray Desmond’s Dictionary of British and Irish botanists and horticulturists, including plant collectors and botanical artists, ‘Britten and Boulger’ has been superseded by a work of far greater scope, which includes nurserymen, gardeners and horticultural writers as well as botanists, plant collectors and botanical artists.

This fat volume of 747 pages is well bound in a sturdy and attractive green cover. The book is well produced and opens easily. The pages have wide margins enabling manuscript notes to be made. Following the preface, a historical introduction by W. T. Stearn, and an impressive list of books and periodicals consulted in compiling this work, are the biographies. These entries give information such as the dates and places of birth and death of each person listed, short biographical details, a selected list of publications, and locations of plant collections, manuscripts, drawings and portraits. Mention is also made of any genus or species dedicated to a person listed. A new feature is the subject index at the end of the book, which classifies many of the entries under profession, plants or the country where the flora had been collected and studied. Specialists can be found under the name of the subject they studied, such as algology, ecology, lichenology and mycology. British botanists and nurserymen are also listed under the places in the British Isles where they collected or worked. The names under each heading are chronologically arranged.

In a work covering such a wide field it would be impossible not to have some errors and omissions. For example, no mention is made of the important manuscripts concerning Elizabeth Blackwell and her Herbal, preserved in the British Museum (Bloomsbury), and little information is given on Lieutenant Commander Christopher Maitland Stocken D.S.C., R.N., born Bristol 17 April 1922 and killed 23 August 1966 while leading the Royal Naval Expedition to East Greenland. In 1969 his Andalusian flowers and countryside was published posthumously, edited by A. P. Hamilton. Silene stockenii Chater was named after him.

Compared with the entry for F. Kingdon Ward, scant notice has been taken of the two great collectors Frank Ludlow and George Sherriff, who made several expeditions into Bhutan and south-eastern Tibet between 1933 and 1950. Their botanical collections, now in the British Museum (Natural History), amount to over 21,000 gatherings, and their journeys enriched our gardens by the introduction of many plants new to science, such as Aconitum fletcheranum, Gentiana marquandii, Primula ioessa, Rhododendron tsariense and Saxifraga montanella. Six new plants were named after Ludlow and seven after Sherriff.

The compiler of the above Dictionary was formerly Chief Librarian and Archivist at the Royal Botanic Gardens, Kew. He spent nearly eight years in the preparation of the work, and he is to be

BOOK REVIEWS 55

congratulated on the large number of entries it contains—almost four times as many as the publication of 1931. The book fills a great need and to many it will be an indispensable tool. Inevitably the price of the work is high, but it is to be hoped that many public and private institutions, including of course botanical and horticultural libraries, will purchase such a useful book for their reference shelves.

B. HENREY

Flowers of Greece and the Aegean. Anthony Huxley & William Taylor. Pp. vi + 185, with 483 coloured illustrations and 77 line drawings. Chatto & Windus, London. 1977. Price (in U.K.) hardback £6-50, paperback £3-95.

This handy-sized popular guide to the more eye-catching plants of Greek lands follows the same pattern as Polunin & Huxley’s Flowers of the Mediterranean (1965), with the valuable addition of a number of line drawings by Miss Victoria Gordon as text-figures. These are admirably clear and greatly enhance the descriptions of the species so illustrated. In all, a total of 660 species and infraspecific taxa are described in brief, non-technical outline and over 500 are also shown in colour photographs or line-drawings. The authors are to be congratulated on the accumulation and presentation of so many illustrations. Short introductory chapters, one on Geography and Climate and others on the plants particularly associated with the main vegetational types likely to be encountered, paint a general picture of the Greek landscape and flora. As this introduction must be intended for those with little knowledge of the flora, the omission of references to the illustrations against the many plant names mentioned seems strange, requiring as it does use of the index in every case.

The usefulness of this book for identification has its limits set by the ability to cover, in a handy volume, only about a tenth of the flora and by a bias towards certain of the more ‘popular’ families at the expense of others. The orchids are very well represented with ample photographs showing, for the Ophryses, the chief colour and pattern variants likely to be encountered. On the other hand, grasses, which increasingly have their devotees, are almost excluded, whilst the huge legume family gets only a little more space than the Ranunculaceae in the broad sense and the Compositae less than the Liliaceae, Omissions are debatable of course, but the reviewer would have expected to see included Osyris alba, for example, and Erysimum graecum rather than the rare Cretan endemic E. raulinii. More serious than the omissions themselves is the failure to indicate their size, thus leaving the newcomer to the flora without the guidance of knowing whether a family or genus is completely covered or more-or-less largely omitted. Suppose he has a clover to name, unwarned of the 90 or so species not mentioned, he may be misled into thinking it ought to correspond with one of the four species described; or, having a Medicago, fail to recognize it as such at all when, of the 30 species in the area covered, the two depicted are distinctive rather than typical.

Some criticism must be made of the colour photographs, since much depends on them in a book intended for the interested layman and in which no keys are provided to aid comparison of descriptions. It must be said that not all are good enough to lead to easy recognition. Whilst some are admirably clear, many lack either sufficient size (the smallest are 40mm square), clarity or faithful colour rendering to be a useful introduction to the plants portrayed. Surely better and more typical photographs exist of Silene colorata (fig. 28) or Gynandriris sisyrinchium (fig. 402), for example. A general browning mars others and is serious where it gives a false impression of colour contrast between similar species e.g. Paronychia argentea (fig. 24) and P. capitata (fig. 25), Legousia pentagonia (fig. 285) and L. speculum-veneris (fig. 286) or Cyclamen persicum (fig. 193) and C. creticum (fig. 196).

Despite lapses from the ideal the book will be welcomed by flower-lovers visiting Greece and the Islands as guide and souvenir, and for the amusing, if sometimes puzzling, footnotes on the uses, real and mythological, to which the plants have been put from classical times onwards.

S. S. HOOPER

A Flora of the Maltese islands.S. M. Haslam, P. D. Sell & P. A. Wolseley. Pp. 1xxi + 560, with 66 plates and 29 figures. Malta University Press, Msida, Malta. 1977. Price £M3-50 (£4-75).

The successive appearance of four of the five volumes of Flora Europaea has inspired the production of

56 BOOK REVIEWS

new regional Floras and various popular works and pinpointed the principal gaps in our knowledge of European plants. A Flora of the Maltese islands is the latest in a line of regional Floras to adopt its concise, diagnostic style for descriptions; and, although this is a reasonable practical expediency for rapid production, such a procedure has its obvious draw-backs. For example, most of the keys and descriptions for species and genera have been based on, and often lifted straight out of Flora Europaea and are, therefore, distinctly uncritical. The majority of the original descriptions are to be found in the Monocotyledon accounts (which have been compiled by Pat Wolseley with considerable help from Dr Martin Rix and guidance on the Gramineae from Dr Alexander Melderis) and, of course, that of the Compositae, written competently by Peter Sell. Nearly all the cited chromosome numbers are based on | extra-Maltese material, and several of the descriptions mention variation that is not known to occur in Malta. Furthermore, uncritical decisions have led to the inclusion of several taxa of dubious occurrence within the islands.

Nevertheless, this new work is justified in being called a comprehensive Flora since there is such a lot of useful information besides the diagnostic elements. Common English names are given to every species, and Maltese ones too are cited, when available. Distribution data for the islands are based on the localities given by Gulia, Duthrie, Delicata, Sommier, Gatto, Borg, Lanfranco, Kramer and Westra, as well as on the personal observations of Pat Wolseley. The general distribution of each species is also given, and ecological statements are based on published work together with the field notes of Haslam and Wolseley. The illustrations are a highlight, and it 1s always a pleasure to see Pat Wolseley’s rather sparse but distinctive line drawings. These are arranged on 66 plates providing an illustration of at least one species for each genus described in the Flora. However, this is a poor idea, and such expertise would have been better employed by concentrating on taxonomically difficult groups. The introductory chapters provide valuable comments on the history, geography, geology, ecology, vegetation and relationships of the flora. Maltese collaboration is evidenced by a chapter by Prof. J. Borg on fruit-growing, the island’s main agricultural activity, and an account of medicinal plants by the late Prof. Henry Micallef.

The editing has suffered from the attentions of ‘too many cooks.’ At the same time, although transcriptions of Maltese names into English have been, to say the least, variable and spelling or type- setting errors such as Opyhrs and Black Horebound are unforgivable, the general layout and consistency of the work is good and must be commended. The printing quality is medieval, especially for the photographs and the line drawings, which are rendered useless in several instances. The practical life of the book will be short, since there are far too many pages for a paperback edition and the binding is very poor. My copy is falling apart already, which is a serious consideration for a work principally aimed at field students. However, despite all criticism the Flora represents a considerable updating of Borg’s (1927) work, and it is fair comment to say that it will be the standard work of the Maltese islands for the next 50 years.

REFERENCE

Bore, J. (1927). Descriptive Flora of the Maltese islands, including the ferns and flowering plants. Msida, Malta. C. J. HUMPHRIES

Atlas zur Flora von Stidniedersachsen. H. Haeupler. Pp. 367, with 10 plates and 1818 maps. Scripta Geobotanica X, Verlag E. Goltze, Gottingen. 1976. Price DM36.

After the publication of the Atlas of the British flora in 1962, several European Institutions decided to follow suit and produce, along similar lines, an Atlas of the flora of Central Europe. As far as | know, the most comprehensive and best organized project so far is the proposed mapping of the flora of Bavaria (Arbeitsgemeinschaft zur floristischen Kartierung Bayerns), which involves the largest region of the Federal Republic of Germany. The work to be reviewed here is based on a much smaller region, namely southern Lower Saxony, comprising an area of 140km square. In all, the work contains 1746 distribution maps, produced by the dot-grid method. In addition to all species indigenous to the area, it

BOOK REVIEWS Dil

contains numerous microspecies as well as most of the established introductions. The taxonomy is based on Ehrendorfer’s Liste der Gefasspflanzen Mitteleuropas (1967, 1973). The bulk of the records was gathered between the years 1967 and 1971, mainly by field work (90%) but also from literature and from herbarium material.

The introduction contains a map of the area and 30 smaller maps showing geomorphological, climatological and phytogeographical details. The entire work is beautifully produced and printed on high-quality art paper. The colour illustration (Viola odorata) on the cover is splendid indeed. One cannot help wondering on what grounds so much enthusiasm and certainly money was lavished on the production of a distribution atlas of such an insignificantly small area; local patriotism may be the answer. Despite all the effort and industry invested in the project one finds the smallness of the in- dividual maps (six to the page) rather irritating: the topographical background is condensed to such a degree that anyone not familiar with the area finds it hard to recognize the given distribution patterns. Another handicap is the fact that a large part of the area covered is situated within the German Democratic Republic (most of the Harz Mountains and the adjacent plains stretching almost to Magdeburg), to which none of the field workers had any access. The list of co-operators (pp. 28-30) shows a pretty thin representation from the G.D.R. Nevertheless, if this work is a foretaste of things to come and if the projected Atlas of Central Europe is as painstakingly prepared as this book, one should eagerly await its arrival.

E. LAUNERT

Neuste Anweisung, Pflanzen nach dem Leben abzudrucken. E. W. Martius. Wetzlar. 1784. Facsimile print, introduced and edited by Armin Geus (limited edition of 100 copies numbered in Roman and 400 copies in Arabic). Pp. xxvi + 80, with one figure. Basilisken-Presse, Marburg. 1977. Price DM28.

By definition antiques cannot be produced; but due to the rapid expansion of the Scientific community and the subsequent foundation of new universities and scientific libraries, the number of people who want to consult ‘antique books’ is ever increasing and their demands have pushed up the prices of most of the works, which occasionally come on to the open market, to astronomical levels. No wonder that for this reason the reprinting of rare old scientific works has become a major industry. Reprinting not only makes rare books widely available but it often helps to preserve valuable works of reference from being damaged by constant use. In contrast to most publishers who concentrate on the reproduction of ‘useful’ scientific works, the Basilisken-Presse of Marburg has the bibliophile foremost in mind. After their beautiful facsimile edition of Cosmos Conrad Cuno’s Observationes Microscopia, they have, as a second venture, unearthed not only a neglected rarity of the first order but also a most delightful little book: init E. W. Martius, father of the famous author of the Flora Brasiliensis, K. F. Ph. von Martius, ‘ gives the ‘Latest Instructions’ on nature printing. To the historian of scientific illustration this method is of passing interest only, since its shortcomings are obvious. Nevertheless, during the 18th century and even in the first half of the 19th a number of German books were illustrated by this method. Nearer home one should, of course, mention T. Moore’s Ferns of Great Britain and Ireland, which, due to the two-dimensional nature of these plants, showed nature printing at its most successful. For the English reader of the book it is interesting to see Martius disputing the assumption that the credit for the invention of the method should be given to ‘Dr Sherard aus England’. The author, who lived from 1756-1849, was a pharmacist and founded, together with D. H. Hoppe, in 1790 the Regensburger Botanische Gesellschaft (the first botanical society in Europe), wrote his book in Wetzlar after he had studied and acquainted himself with various existing methods of nature printing. In a disarming pedantic fashion—‘von Baumen und allen grossen Pflanzen nimmt man nur Zweige’’—he describes his own method and relates the history of the process in great detail.

This book, provided with a scholarly preface by Prof. A. Geus, printed on high quality paper, furnished with beautiful end papers and attractively bound, can be recommended to every lover of fine books.

E. LAUNERT

58 BOOK REVIEWS

Welsh timber trees, native and introduced. H. A. Hyde, 4th edition, revised by S. G. Harrison. Pp. xii + 165, with 45 half-tone plates and 52 text-figures. The National Museum of Wales, Cardiff. 1977. Price £3-00 plus p. & c. 47p.

This charming little book, with its detailed descriptions and beautiful illustrations, has always had relevance beyond the Welsh borders and must be known to many readers. It is noticed here because the much-enlarged edition of 1960, now out of print, was not reviewed in these pages. The new text is little altered from the earlier one; only the general sections on Welsh forests and woodlands (pp. 17—26) have been rewritten. Some comments on uses and diseases have been added, a fresh distribution map and one Plate substituted and 37 items added to the Bibliography. The statistical Appendix has been dropped, as also, sadly, has the glossary element from the Index; the 37 ‘personal communications’ in the Bibliography might have been better spared.

This edition is therefore largely a reprint. The opportunity has not been taken to increase the forestry content; the text is still more about trees than timber. Mr Harrison could have augmented the conifer section especially; for example, his account of the Mountain Pine has not been brought into line with that in his revision of Dallimore & Jackson (1966), and the Hybrid Larch and Hybrid Cypress, at ieast, deserve fuller treatment. However the book remains a handy, reasonably priced text on the common trees of Britain; and the inclusion of Mr A. F. Mitchell’s records has certainly brought the references to notable specimens up to date.

REFERENCE

DALLIMORE, W. & JACKSON, A. B. (1966). A handbook of Coniferae and Ginkgoaceae, revised by S. G. HARRISON.

London. J. LEwIs

Tree rings and climate. H. C. Fritts. Pp. xi1 + 567, with numerous text figures. Academic Press, London. 1977. Price £16-00.

Dendroclimatology is a branch of dendrochronology, both marginal to the study of the British flora. So, only a short reference to this long book (much of which is technical) is justified here. Nevertheless the subject is fascinating and of ever-growing importance, as is shown by this volume with its 19 pages of references and by the holding last in July 1976 at Greenwich of the five-day symposium on ‘Dendrochronology in Northern Europe’.

The science was pioneered by A. E. Douglass from 1901, who established in 1937 the Centre for Tree- Ring Research in the University of Arizona at Tucson. But the earliest observation on the subject had been made in 1737 by Duhamel and Buffon, who recognized frost damage 29 rings in from the bark. Just a score of years ago, the science revealed the existence of the 4,600 years old Bristle-cone Pine, Pinus longaeva as it was named, and from it a chronology of no less than 8,500 years has now been established.

This book claims to be no more than a progress report (albeit thorough) on a rapidly developing subject, deducing past climates from the size and nature of tree rings, which ‘can give a view of what the climate. . . is most likely to be in the future’. To help to do all this, a close knowledge is required of the physiology of trees, which is here gone into in detail. The book also, of course, sets out the principles and techniques of the science, and mentions some of its practical applications. It seems to me the job has been well done.

D. McCLINTock

_ Watsonia, 12, 9-77 (1968). 59

Reports ANNUAL GENERAL MEETING, MAY 7th, 1977

The Annual General Meeting of the Society was held in the Learning Resources Centre, Plymouth Polytechnic, on Saturday, May 7th, 1977 at 14.00 hours, with 45 members present.

Mr E. L. Swann (retiring President) took the Chair, and Dr D. L. Wigston read a letter of welcome from Lord Morley, Chairman of the Board of Governors of Plymouth Polytechnic, who expressed regret at being unable to join the meeting.

The minutes of the last Annual General Meeting, as published in Watsonia, 11: 271-272 (1977), were passed after the election of Council Members at that meeting (Dr Q. O. N. Kay, Dr J. L. Mason, Mr J. M. Brummitt and Dr J. Dransfield) had been recorded.

REPORT OF COUNCIL The Report of Council for the calendar year 1976 had been circulated to members and was adopted by the meeting.

TREASURER’S REPORT AND ACCOUNTS

The Accounts showed a gain in total net assets of £2,199, but rising costs were a continuing problem. The Treasurer’s Report was carried unanimously, and the Treasurer particularly thanked Dr F. H. Perring, who had taken over responsibility for sales of B.S.B.I. publications, and Miss E. Young, for preparing the tax repayment on covenants.

QUESTIONNAIRE TO MEMBERS RESIDENT IN SCOTLAND

The organization of B.S.B.I. activities in Scotland had been raised at the Annual General Meeting in 1975. Discussion since had led to the sending of a questionnaire to the 159 members in Scotland asking if they would like a B.S.B.I. Committee for Scotland, or if they would prefer the organization in Scotland to remain, as at present, through the Committee for the Study of the Scottish Flora. Mr Swann announced the results: 98 replies had been received, 58 in favour of a B.S.B.I. Committee for Scotland, 38 in favour of the organization as at present, and 2 abstentions.

ELECTION OF PRESIDENT

The retiring President, Mr E. L. Swann, thanked all officers, especially the Honorary General Secretary, for support and help during his presidency and proposed Professor D. H. Valentine for election. This was carried unanimously and Professor Valentine then took the Chair, thanking Mr Swann for his work for the Society.

ELECTION OF VICE-PRESIDENTS Mr J. P. M. Brenan and Mr J. F. M. Cannon were unanimously elected as Vice-Presidents and the retiring Vice-Presidents, Mrs H. R. H. Vaughan and Dr W. T. Stearn, were thanked by the President.

ELECTION OF OFFICERS

The following officers nominated for re-election were elected en bloc: Mrs M. Briggs, Honorary General Secretary; Mr M. Walpole, Honorary Treasurer; Mrs J. M. Mullin, Honorary Meetings Secretary; Miss L. Farrell, Honorary Field Secretary; Mrs R. M. Hamilton, Honorary Membership Secretary; Dr G. Halliday, Dr N. K. B. Robson and Dr C. A. Stace, Honorary Editors. Dr S. M. Coles and Dr D. L. Wigston, nominated for election as new Honorary Editors, were unanimously elected.

ELECTION OF COUNCIL MEMBERS Mr R. W. David, Captain R. G. B. Roe, O.B.E., R.N., and Mrs A. C. M. Duncan had been nominated and were elected unanimously; their order of precedence, as given, was determined by ballot.

60 REPORTS

ELECTION OF HONORARY AUDITORS Messrs Thornton Baker and Co. were gratefully re-elected as Honorary Auditors.

ANY OTHER BUSINESS The suggestion that the thanks of the meeting be sent to Mr F. Perring (Senior) for handling the sales of B.S.B.I. posters and books by post from Oundle Lodge was warmly applauded.

The meeting closed at 14.35.

M. BRIGGS

PAPERS READ AT THE ANNUAL GENERAL MEETING

A NEW COUNTY FLORA FOR DEVON

The decision to start work on a new Flora was taken late in 1969 by the Botanical Section of the Devonshire Association with the cooperation of the Department of Biological Sciences of the University of Exeter. The existing Flora, by the Rev. W. Keble Martin and G. T. Fraser, published in 1939, was not only out of date by modern county Flora standards, but also gave little information on the occurrence of the commoner species. Its records had been collected on a Parish ‘basis over a very considerable period.

The Atlas of the British flora and several new county Floras use the 10km grid square or a subdivision for recording, and the new Devon Flora adopts this basis. Devon essentially represents vice-counties 3 and 4, though boundary changes have occurred. The new county Flora boundary is taken as shown on the 7th edition O. S. maps, scale 1:63,360. The 2 x 2km tetrad is used as the basis for recording flowering plants and ferns. 1843 tetrads are represented within the county boundary as defined, but some very small portions of marginal tetrads are included in the most appropriate adjacent one. Lichens and bryophytes are now also being recorded in 10 x 10km squares.

The data are initially stored on punched cards and then transferred to on-line disc storage using the University of Exeter computer. An associated computer programme allows a range of output data, including a map of the number of species recorded from each tetrad. Records are now sufficiently numerous for maps to reveal individual species distribution patterns, and to identify suspect records.

Woodell (1975) examined the problems of writing a county Flora and the following examine his points in the light of the Devon Flora experience:

Number of visits/visitors per tetrad. Two recorders and two visits to a tetrad are essential, but the most efficient recording is carried out by individuals.

Accessibility. Devon 1s a large county with many narrow lanes, and evenly-spread recording without a car is impossible. The weakest parts are the agricultural areas of the north, centre and south, which are often botanically ‘dull’.

Marginal tetrads. These have presented few problems, dealt with by recorders making specific excursions in conjunction with the distribution maps to ensure that all tetrads are recorded. Identification. This has been the major problem. Grasses and critical genera are under-recorded. The misidentification of common plants is almost impossible to detect, but it is hoped that the numbers of such are small compared with the total and that the less common species with interesting or important distribution patterns are reasonably accurately identified.

FRASER, G. T. & Martin, W. (1939). Flora of Devon. Arbroath. WooDELL, S. R. J. (1975). Five years of a county Flora project. Watsonia, 10: 265-272.

R. B. IvimMgEy-Cook

THE FLORA OF THE TAVISTOCK WOODLANDS

The Tavistock Woodlands were part of the Duke of Bedford’s Devonshire Estates from 1539 to 1959. There is a remarkably complete series of records for the Estate which allow interpretation of the changing patterns of coppice, coppice-with-standards, and clear-fell plantation forestry (Wigston 1976). Dates of introductions can often be identified, such as Fagus sylvatica (not native to the south- west peninsula) for pure beech stands, Quercus cerris and Rhododendron spp. (apparently for

REPORTS 61

amenity), and Picea sitchensis with the initiation of plantation forestry. The variety of tree species compartments and their range of age-structure also leads to a diversity of under-storey and epiphyte species, such as Vaccinium myrtillus and Luzula sylvatica dominating under oak coppice, and Rubus

fruticosus agg. and Pteridium aquilinum under old spruce canopy. Species lists are currently being compiled and phenological changes under various canopies being monitored by Mrs M. Harris. A ecent and unexpected record is of Erica vagans on the edge of a ride.

In 1959 the Woodlands were purchased by the Earl of Bradford, who has initiated a selection forestry scheme in the estate—the Bradford-Hutt continuous canopy system (Hutt 1975). Under this management, tree-species diversity, both of conifers and hardwoods, begins to develop as soon as the system is introduced into existing stands (which may be of conifers, closed canopy hardwoods, or relict coppice). Self-seeding of non-crop species (e.g. native oak) occurs, and is allowed to remain until the appropriate thinning, or, if suitable, it may be incorporated into the management programme. Species lists show that under-storey species diversity is also encouraged. For example, under old stands of Pinus nigra var. maritima, only three understorey and two epiphyte species were recorded; under canopy of the same species following 15 years of Bradford-Hutt management, five introduced trees were recorded (including the hardwood Nothofagus procera), with three self-sown hardwoods (Q. robur, Q. cerris, Sorbus aucuparia), thirteen understorey flowering plants and ferns, and seven epiphytic bryophytes and lichens.

Floral diversity under the system is currently being investigated. Initial impressions suggest that such management is a remarkable combination of economic exploitation and ecological conservation. Compared with normal coniferous forest plantation, the Tavistock Woodlands Estate is an area of considerable interest for botanists.

Hutt, P. (1976). The Bradford Plan continuous canopy forestry system. Practical Education (South West), 72: 20-24.

WiastTon, D. L. (1976). Ecological aspects of woodland management and forestry practice (paper read to the Institute of Environmental Science and Forestry Commission symposium Forest Ecology, Farnborough College of Technology, December 1976).

D. L. WIGSTON

GALLS ON DEVON OAKS

Plant-galls are caused by a number of different organisms including aphids, mites, beetles and Hymenopteran wasps. The gall structure itself is plant tissue laid down in response to a chemical initiation by the gall causer. Galls occur on most parts of the host tree—on roots, leaves, catkins, buds, bark, twigs and, in the case of oaks, acorns and acorn-cups. By far the most differentiated galls are those caused by the Hymenopteran family Cynipidae, some 30 species of which are known to cause galls on the indigenous British oak species, Quercus robur, Q. petraea and hybrids. Many wasps have alternate sexual and agamic generations which give rise to different gall structures; these may occur on the same host taxon or on different Quercus species. For example, the spread of the marble gall, Andricus kollari, following its introduction in the 1830s for tannin production, was probably facilitated by the introduction of its sexual generation host, Q. cerris, in the 1720s.

A number of Quercus robur, Q. petraea and hybrid sites in Devon are being monitored for gall occurrence and distribution, both within individual tree canopies and within and between populations. Initial findings from the study are:

(a) exposed oaks are more heavily infested than sheltered trees;

(b) Q. robur is attacked by larger numbers of gall-causing species and in higher densities than is Q. petraea or hybrids;

(c) one or both generations of some gall causers may be host-specific. Thus, the knopper gall, A. quercus-calicis, appears to be restricted to Q. robur, as is the catkin gall, A. quercusramuli;

(d) gall distribution is largely restricted to the lower canopy.

The spread of A. quercus-calicis, first reported in Britain in 1962, has been followed with particular interest in view of its possible effect on the natural regeneration potential of British oaks. A previously unreported observation is the galling of the acorn itself in addition to the cup. The distribution of this gall increased considerably in Devon during 1976, reaching as far west as Plymouth. As much as 80% of a tree’s mast can be galled, and in germination experiments 60% or more of uninfested acorns

G62 REPORTS

developed into seedlings, but only 7% of infested (knoppered) acorns. No chalcid parasites of this gall (known to occur on the Continent) have been found in this study.

M. H. MARTIN

NATIONAL VEGETATION SURVEY—A PROGRESS REPORT ON THE SOUTH-WEST REGION

The major treatise on the vegetation of the British Isles is still that of Tansley (1939), the last printing of which appeared in 1949. Although it will remain a landmark in British ecology, the work is out of date; much detailed work is now available on areas of British vegetation, but, with the exception of McVean & Ratcliffe’s (1962) work on the Scottish Highlands, there has been little attempt at a synthesis to follow the work of Tansley. In 1974 the Nature Conservancy Council issued a contract to produce a ‘dictionary’ of vegetation types covering the plant communities of all natural, semi-natural and major artificial habitats occurring in Great Britain. The dictionary is intended to be a reference description of discrete, named and systematically arranged vegetation units, but of a form to be widely useable by workers in plant and animal taxonomy and ecology, nature conservation, land-use planning and description.

The survey is based on five essentially climatic and geographical regions; the South-West Region, based at Exeter, consists of South Wales in addition to the south-western peninsula. Each region has a supervisor and a research assistant. The experience of the supervisor and the increasing familiarity of the assistants with their region is complemented by the local experience of research workers, B.S.B.I. recorders, Naturalists’ Trusts, etc.

The releve approach is used, and the initial work has involved establishing the range of homogeneous vegetation types within the region, from floristic lists and associated site data. It is already clear that the varied geography and history of the South-West Region has produced a multiplicity of vegetation types whose limits remain to be defined. All flowering plants, gymnosperms, pteridophytes, bryophytes and macrolichens are recorded. A list of critical taxa has been compiled, including Rosa, Rubus fruticosus agg., Hieracium, Quercus robur/petraea, and exchange of information with the editors of the projected new Flora of Great Britain and Ireland is intended. Voucher material of these taxa has been collected from the outset.

TANSLEY, A. G. (1939). The British Islands and their vegetation. Cambridge. McVEAN, D.N. & Ratcuirre, D. A. (1962). Plant communities of the Scottish Highlands. Monographs of the Nature Conservancy, 1. London.

M. C. F. Proctor & P. WILKINS

EXCURSIONS HELD IN CONNECTION WITH THE ANNUAL GENERAL MEETING

MOUNT EDGCUMBE PARK, CORNWALL. May 7th

Early arrivals for the Annual General Meeting gathered at Admiral’s Hard, Stonehouse, for the 10.15 Cremyll Ferry across the mouth of the river Tamar to Cornwall and Mount Edgcumbe Country Park. The formal gardens were visited first, and fine specimens of Ginkgo biloba and Quercus suber were observed, and also the unique Q. ilex hedge. Particularly admired was the ‘English Garden’, restored as an Elizabethan Knot Garden. In addition to the formal plantings, interesting self-introductions were seen, including Polypodium australe on the sea-wall.

From the formal gardens the party went on to the open parts of the estate. Of special interest were characteristic specimens of U/mus glabra, U. procera and U. carpinifolia within a few yards of one another. The pond below Edgcumbe house was covered by Azolla filiculoides, which in one year had largely ousted Lemna spp. and Wolffia arrhiza.

The visit was necessarily short, but lunch in the Edgcumbe Arms suitably prepared the members for the afternoon Annual General Meeting.

D. L. WIGSTON

a ee

ee an ee ees

REPORTS 63

WISTMAN’S WOOD, DEVON. May 8th

Some 25 members of the Society gathered at Two Bridges, Dartmoor, to take the footpath to the highest Quercus robur population in the British Isles, Wistman’s Wood. Ominous rainclouds suggested that we were to receive some of the 60 or more inches of rain that the Wood experiences each year.

Beyond Crockern Farm the party paused to view the Forestry Commission plantation at Beardown. The management strategy of single-species, uniform-age-compartment, clear-fell forestry was clearly seen. The associated paucity of understorey species and severe gully-erosion on slopes was noted.

On reaching Wistman’s Wood, the late flushing of these high-level oaks was evident, although there was much less attack by Tortrix viridana and other defoliators this year. Expanded leaves clearly showed the features of Q. robur. The unusual growth-form and population structure of the oaks was examined; the distinction between old, prostrate, low-branching forms and younger upright trees was seen. The party was fortunate in including Dr M. & Mr G. Spooner, who have conducted much research into the visible changes in Wistman’s Wood from photographs dating from Victorian times onwards. In particular Mr Spooner was able to locate positions from which A. G. Tansley obtained photographs, and showed that the Wood was expanding and regenerating vigorously. The party went on to examine the rich understorey and epiphyte flora for which the Wood is renowned. Sorbus aucuparia was noted as an epiphyte on some older oaks! Polypodium vulgare was another common epiphyte along with many bryophyte species. Grazing within the wood was clear from the cropped accompanied by Equisetum fluviatile. A Dactylorchis was subsequently determined by Mr Wood as Conservancy to eliminate the effects of grazing, was examined. L. sy/vatica was particularly luxuriant, but impenetrable Rubus fruticosus agg. was dominant. The lower part of the enclosure extending below the woodland boundary contained many young oaks, established after the erection of the fence. Although the Wood is expanding, grazing of seedlings must be a major population-regulating factor. The party then proceeded to the junction of the South and Middle Woods; with the expansion of the population these two areas, distinct in Tansley’s time, have now nearly merged. The party lunched bya young prostrate oak regarded by Mr Spooner as typical of the growth-form of most of the oaks earlier this century.

The North Wood was then visited, and the members moved on to examine the flora of Longford and Littaford Tors. Typical habitats for plants as diverse as Lycopodium selago, Hymenophyllum wilsonii, and Endymion non-scriptus were observed, but not the plants themselves. The weather had proved remarkably kind, and the party returned to Two Bridges and thence to the Spooners’ house at Yelverton for tea, where photographs of Wistman’s Wood at various dates were displayed. Many members were delighted by the garden, which contained, amongst other plants, Thelypteris oreopteris, Erica vagans and Polygonatum multiflorum.

The hour was late and the party began to break up. One stalwart went on to see H. tunbrigense at Shaugh Prior, and those members staying in Plymouth returned to see the Plymouth Pear at Estover, and the Plymouth Thistle and Bladder Campion on Plymouth Hoe.

D. L. WIGSTON

SLAPTON LEY, DEVON. May 9th

The previous day’s good fortune with the weather was not repeated. Some 15 members gathered in intense rain on Slapton Sands. Undaunted, they were met by John Griffin, the Deputy Warden of Slapton Ley Field Studies Centre, who led the party along the shingle ridge to observe characteristic species such as Calystegia soldanella, Glaucium flavum, Euphorbia paralias and Cakile maritima. The sea pea, Lathyrus japonicus, was observed—a recent introduction following a Royal Marines exercise from Portland.

The backslope of the ridge was seen to be a more stable habitat, with Silene maritima and Armeria maritima. The considerable variability in the form of Geranium molle was discussed. As we proceeded across the bridge between the Higher and Lower Ley, the dense reed beds could be clearly seen. Along the edge of the Ley the locally rare Lusitanian species Taraxacum hispanicum was collected for the British Museum herbarium. The exterior bracts of this plant serve to distinguish it from all other members of the genus. The Lower Ley Marsh was characterized by the dominance of Phragmites and Sallows; Equisetum fluviatile and Oenanthe crocata were much in evidence.

64 REPORTS

As we passed into France Wood, Allium ursinum was apparent both visually and aromatically, and later garnished one member’s sandwiches. The management of the woodland to maintain its mixed character was pointed out.

D. L. WIGSTON

CONFERENCE REPORT

THE POLLINATION OF FLOWERS BY INSECTS

This symposium was organized by Dr A. J. Richards at Castle Leazes Halls, University of Newcastle upon Tyne, from April 14th—17th, 1977, for the B.S.B.I. in association with the Linnean Society. Of the 150 participants, no less than 40 came from abroad, and nearly half of those from Holland, but representatives from Norway, Denmark, Austria, Israel, Eire, South Africa, New Zealand, Canada and the U.S.A. were also present. British members included many amateurs, some with anthecological or ambrosial leanings, but also a strong professional contingent, and about 25 students.

During a protracted arrival on the Thursday, participants learnt the complex infrastructure of the building, so typical of modern campus architecture. The conference started on Friday morning in the Junior Common Room of Freeman’s Hall. This commodious room also doubled as a bar and area fora flower display: meals were next door and exhibits immediately downstairs. Professor K. Faegri gave an introductory address on trends in research in pollination ecology, after welcomes by Professor S. L. Ranson, on behalf of the University, and Professor D. H. Valentine, on behalf of the B.S.B.I. This was followed, after coffee, by papers on pollinator behaviour by DrS. R. J. Woodell (directionality in coastal bumble bees), Dr N. B. M. Brantjes (on how moths find flowers), and Dr Sarah Corbet (on variation in nectar content in Echium and its effect on pollinators).

After lunch, Dr M. C. F. Proctor, who had brought with him a series of his magnificent photographs of pollination, one of which had formed the motif for the conference, talked on the evolution of insect- pollinated flowers with respect to habitat-type, followed by Dr P. Kevan who gave a most enlightening lecture on how insects see flower colours. After tea, Professor S. Vogel, who astounded participants throughout the conference with his encyclopaedic knowledge, talked on deception in pollen flowers, followed by Professor A. D. J. Meeuse, who gave a graphic illustration of the heat generation in aroids.

Participants then walked to the Department of Plant Biology, relishing the fresh air and exercise, where they were entertained with sherry. After dinner, we saw three films: one by Professor Valentine on Impatiens; a short but exciting view of hawk-moths pollinating Echium in the Breck by Dr Corbet; and a long film on pollination of Rhinanthus by Dr N. Kwak. An even longer, but no less professional, film covering the whole of the Rhinanthoideae was shown by Dr Kwak to interested parties the following night.

We had hoped to see some 8mm films, but the lack of a suitable spool prevented this for nearly an hour. Some Dutch ingenuity finally allowed us to see Professor Vogel’s film on oil-collecting wasps in Lysimachia, and Dr Brantjes’ excellent sequences of a seed-eating moth-larva.

The next day concentrated on population biology, and opened with Dr A. J. Richards reading Professor D. A. Levin’s important, if somewhat weighty, paper on the effect of pollinator behaviour on population structure (Professor Levin was unfortunately not able to be present). Two papers on the effect of pollination on flower-colour polymorphism by Dr Q. O. N. Kay, and his ex-student, Dr D. Mogford, were followed by Dr Richards’ and Miss H. Ibrahim’s work on population size in Primula veris, and Dr A. J. Beattie’s most excellent paper on seed and pollen dispersal in Viola.

After lunch, Professor L. van der Pijl (who regaled us throughout the conference with a succession of entertaining stories) talked on floral functions, their integration and sexual disharmony, followed by Professor Valentine, who gave an account of the biology of the British /mpatiens. After tea, Dr D. Eisikowitch showed that even the Israel sea-shore is a windy place, and that plants seem to be dwarf in response to the low-flying requirements of pollinators. Two final papers, discussing insect pollination in apparently wind-pollinated plants (Plantago and Salix), from Professor Meeuse and P. Stelleman, brought the papers read to 18 and the paper-reading sessions to a close. After a sherry party,

REPORTS ~ 65

given by the University and hosted by Professor D. H. Whiffen, Dean of Science, came the conference dinner, attended by about 70 and followed by some memorable speeches.

The weather had been cold and the season late, even by north-eastern standards, so it was with some trepidation that about 40 people set out for the Farne Islands. In the event, it was indeed very cold, and rising wind caused a hasty and premature evacuation, but close views of many of the nesting birds and the seals were afforded to many, even if pollinators were scarce! About 25 went to Corbridge, where ‘Kilbryde’, Randall Cooke’s old garden, now the property of the University of Newcastle, was examined with interest, although, due to the very late season, the amount in flower was disappointing. A visit to the Roman fort at Corstopitum was very cold, but enlivened by a guide who is researching at the site. On the way home a hybrid population of Primula was visited briefly.

It remains to comment that proceedings throughout were lively and interesting, and socially as well as intellectually enjoyable, and thanks are due to the authorities at Castle Leazes Halls for making arrangements so efficiently, as well as to speakers, exhibitors, and volunteer helpers from the Department of Plant Biology, without whom the conference would not have been possible. One name must be selected for special mention: E. M. Caldwell, administrator of the Department of Plant Biology, who undertook much of the strenuous and tedious work of registration, booking and finance.

A. J. RICHARDS

FIELD MEETINGS 1976

CHANNEL ISLANDS

GUERNSEY. MAY 27TH—JUNE 2ND

The B.S.B.I. and the Wild Flower Society joined forces with La Societe Guernesiaise for six days in the Guernsey bailiwick. The occasion was launched by Old Government House bursting at the gunwales to hear a talk on the local wild flowers illustrated by some 150 of the superb slides of John Bichard. The object of the visit was to see how the recent Wild flowers of Guernsey worked in the field. To the gratification of the author, it did this well; not that some 30 visitors, plus an always welcome daily posse from La Societe (in Sark also from the newly-founded Société Serquaise), did not add to the records.

For example, Gaudinia fragilis was spotted well-naturalized in satisfactory quantity in a fresh area of the central pasture of La Grande Mare; and more plants were seen of Centaurea aspera, near what had seemed the last, ailing, three. Near by, at Vazon, Ophrys apifera was most happily refound in a long- lost locality. Tragopogon porrifolius, even if it was only one plant, was a nice find in a lane at St Peter’s, at the opposite end of the island from its small but long-established colony. The Amaranthus deflexus by the St Sampson’s power station was not only flourishing despite attrition by the tidiers, but one plant had blotched leaves; and Polygonum rurivagum was near by.

In Sark, Festuca caesia (glauca) was on a hot, dry cliff by La Coupee, and Geranium purpureum down the path to the west. Sibthorpia europaea turned up in a gully near Le Fort; and Oxalis exilis, Euphorbia lathyrus and Papaver somniferum confirmed at Clos Jaon, all new to the island. So is Potentilla argentea, many plants of which were seen in a bed and path at the Seigneurie garden.

Herm had Crambe maritima on both western and eastern sides (it seems to be increasing locally); there were two good plants of Rumex rupestris on the west coast, and Polygonum maritimum surviving ball games on the east.

Alderney. The weather throughout was perfect, including, from the organizers’ point of view, the Monday. That was spent, from 7am, mostly quietly in the airport waiting, in vain, for the fog to lift, a welcome respite for them, above all for Mrs Ryan, the Botanical Secretary of La Societe who had put so much work into the preparations and the day before entertained the party to tea. But it was a sad gap for most of the visitors—all they got was a dander in the afternoon to be shown Polypodium vulgare subsp. serrulatum, growing, extraordinarily enough, in deep shade near a path by the southern cliffs.

This is not the report the visitors would have written, most of whom were adding to their diaries and

_ lists many new plants which were not, however interesting, new to the Guernsey tally. But it was a _ pleasure to be able to share so much in such a short time with such a wholly delightful gathering. |

D. MCCLINTOCK

66 REPORTS ENGLAND

TEWKESBURY, GLOUCS. MAY 8TH

On a bright, sunny morning a party of 21 members and friends from eight counties met at Tewkesbury. Mrs S. C. Holland had planned a route that enabled me to demonstrate the variation in habit and appearance of the Black Poplar (Populus nigra). A short walk down Lower Lode Lane introduced the party to the first Black Poplar, a fine female tree, covered with green catkins, growing in a hedgerow in the flood-plain of the River Severn. This tree, the finest female Black Poplar in Gloucestershire, is, I believe, protected by a T.P.O. Another Black Poplar was noticed on the opposite side of the lane, probably an offspring taken as a cutting from the nearly mature tree.

We then moved over the Mythe Bridge and into Worcestershire, where at Lower Marsh Lane, Longdon, two fine male trees were admired. The red male catkins had already fallen, in contrast to the green catkins of the female tree, which it holds till June. At Glenberrow, a few miles further west, we came to a magnificent spreading tree on the village green. It had been pollarded many decades ago and lacked the characteristic arching lower boughs, but the furrowed trunk and leaf characters placed it firmly in P. nigra. We then turned back to our next port of call, Corse Lawn, where the mile-long village green supported a large number of ancient, frequently pollarded, male trees, many growing by ponds. Here a picnic lunch was taken in the shade of some of the poplars.

Back in Gloucestershire the party stopped to inspect four fine male pollards at Tirley, and after crossing the Severn by Haw Bridge, a dying veteran near Apperley. The final port of call was Boddington, where the owners of Boddington Home Farm gave us permission to have a close look at a particularly fine specimen in their orchard, at least 100 feet high and probably more than 150 years old. Other trees of various sizes were growing along an old drainage channel on the farm, one of these can be clearly seen from the M5. All these Boddington trees are characteristic in habit and leaf characters, but they lacked the development of large bosses on thin trunks, so demonstrating that exceptionally this diagnostic character may be lacking.

After the meeting ended, a few enthusiasts went to the Badgeworth Nature Reserve where the smallest P. nigra tree in Gloucestershire was seen. This is a sapling grown from a cutting taken from a tree at Down Hatherley, and planted by Mrs Holland in 1973 to mark ‘Plant a Tree in *73’ year.

I would like to thank Mrs Holland for her admirable planning which made the meeting so profitable for those who attended. It was good to meet so many people interested in my favourite British tree.

E. MILNE-REDHEAD

HAM MEADOW LANDS, GREATER LONDON. JUNE 6TH A party of 23, including members of the B.S.B.I. and the London Natural History Society, met at Ham House car park to study the flora of Ham meadow lands.

A large number of both spring and early summer grasses were seen flowering, one of the dominant being Zerna erecta. Naturalized trees of Robinia pseudacacia and Colutea arborescens, which had been growing here since this area was a rubbish dump, attracted attention. Other naturalized garden remnants included masses of Armoracia rusticana and Galega officinalis.

J. L. GILBERT

SOUTHAMPTON COMMON, S. HANTS. JULY 3RD

This meeting was the first to be held in southern England for some years expressly as an introduction to that most difficult of genera, Rubus. The locality was chosen for its combination of accessibility and richness (in bramble terms it is the Putney Heath of Wessex), and because in a normal year the plants here tend to be exceptionally far advanced, affording the maximum opportunity of seeing and learning some of the species right at the start of the season.

Unfortunately, this was far from being a normal season. The long drought combined with temperatures well in the nineties through much of the preceding week had resulted in all but a few of the brambles being deprived of their petals, thereby drastically reducing the possibilities of demonstrating the species. A day of ferocious heat also induced mass wilting on the part of the 11 members who attended, and it says much for their enthusiasm that the original programme was able to survive in its essentials and all persisted until the end.

REPORTS 67

11 species were demonstrated, all of them widespread or at least locally abundant in Britain and including representatives of each of the sections. The prevalence of hybrids in some localities—of which this Common is a particularly marked example—also made a firm, if unwelcome impression.

The more important purpose of the meeting, however, was to provide those proposing to embark on the genus with some useful guidelines. Among the points made under this head were: the desirability of first thoroughly learning a species in the field before submitting a specimen of it for naming; the overriding necessity of collecting suitable material; the need for systematic exploration of whole areas instead of sampling at random; the great difficulty of identifying brambles from descriptions (or from the keys or plates in Watson’s book); the existence of so many, still-unnamed, local forms, which bedevil the task of the national specialists. No one should contemplate taking up Rubus, it was suggested, without an excellent visual memory, the diligence to collect carefully and copiously and the self-discipline to make extensive notes.

D. E. ALLEN

PENNINE FELLS (WARCOP RANGE), CUMBRIA. JULY 17TH—21ST

Since 1972 a small party from the Department of Plant Biology, University of Newcastle has been spending a few days each summer centred on Appleby, and exploring the great range of fells that stretch from Hartside, in the north, south to Brough, and which provide the largest continuous area of land above 2000 ft in England. This scarp and plateau is notable for being largely difficult of access (with the exception of the radar road to the summit of Great Dun Fell, which, at 2780ft, is the highest metalled-road in Britain), and for being previously little worked, apart from the National Nature Reserve at Moorhouse. There are very large exposures of limestones, with at least some outcrops in most 1km squares, whilst the Whinsill has impressive exposures in most of the main dissections, as at High Cup, Knock Ore, Middle Tongue, Crowdundle and Black Door. Other notable features include sinkholes in plenty, very good mesotrophic spring-fed mires, heavy-metal spoil, and high-level sandstone block-scree. All have their share of interesting plants which include Saxifraga hirculus (ten sites), Alopecurus alpinus (six sites), Phleum alpinum, Poa alpina, Myosotis alpestris (three sites), Dryopteris assimilis (locally common) and Dryopteris villarii. In common with the more widespread species Sedum rosea, Vaccinium uliginosum, Carex bigelowii, Epilobium anagallidifolium, Alchemilla glaucescens and Euphrasia scottica, these rarities are not found in neighbouring Teesdale. However, the area also shares with Teesdale such species as Orthilia secunda, Gentiana verna, Primula farinosa, Potentilla crantzii, Polygonum viviparum, Carex capillaris, Thalictrum alpinum, Thlaspi alpestre, Draba incana, Asplenium viride, Juncus triglumis, Saxifraga hypnoides, S. aizoides, S. stellaris, Myosotis stolonifera, Epilobium alsinifolium, Minuartia verna, Lycopodium alpinum, Plantago maritima and Viola rupestris. Thus, although little known, its importance can scarcely be over-emphasized. In order to describe this area, to record the frequency and distribution of the less common plants, and to document sites of high conservation priority, higher plants and Bryophytes have been mapped on | km squares.93 of these have now been completed, and some 20 more remain to be covered.

In order to help with recording, a B.S.B.I. party met at Appleby. Altogether 18 members took part on one or more days, including several local members. It had been hoped to concentrate on the M.o.D. range at Warcop, but a late change in firing dates only allowed access on the first day, which was distinguished chiefly by the discovery of the very rare necrophilous moss Haplodon wormskjoldii. Plantago maritima was also found, by a path, and thus possibly introduced; this constituted the first record for the area. Several species mostly restricted to the southern part of the range such as Saxifraga tridactylites, Carduus nutans, Pimpinella saxifraga, Scabiosa columbaria and Geranium lucidum were also recorded.

The Tuesday brought a drive up the radar road, and members saw well-known sites for Saxifraga hirculus (in flower), Myosotis alpestris, Juncus triglumis and Potentilla crantzii. The rediscovery of Phleum alpinum (three plants), in what may be its sole remaining English station, brought pleasure only equalled by the sight of the Howitts’ 50 year-old Rolls negotiating the road with contemptuous ease. Equally remarkable in this annus mirabilis was some very heavy rain. Unworked squares on the back of Little Dun Fell provided more sites for Alopecurus alpinus and Juncus triglumis.

On the Wednesday we worked an area further south than any previously covered, to the north-east of Brough. Here we were guided by Lancelot Henderson who has worked this previously little-known area thoroughly. A flourishing colony of Draba incana was