New Phytol. (1988), 109, 369 376 Maintenance of morphological variation in a biotically patchy environment BY R.C.EVANS AND R . T U R K I N G T O N Department of Botany, The University of British Columbia, Vancouver, B.C., V6T 2B1, Canada {Received 19 June 1987 ; accepted 24 February 1988) SUMMARY T h e relationship between morphological variability and biotic environmental heterogeneity was studied in a pasture population of Trifolium repens L. It had been argued that the unexpectedly high levels of \ ariation in T. repens could be maintained by diversifying selection. The mosaic of neighbours (perennial grasses) with which T. repens co-exists constitutes a prominent element of biotic patchiness that may lead to sorting among T. repens genotypes on the basis of neighbour-specific compatibilities. A variation study was conducted on a set of 400 individuals of T. repens collected on a neighbour-specific basis from a 43-year-old pasture and grown for over 2 years under common garden conditions. Variation in a set of 12 morphological characters was assessed after 4 months and again after 27 months. After 4 months' growth, a significant proportion of this variation was accounted for by the neighbour with which the individuals of T. repens had been growing in the pasture. The actual amount of variation accounted for, however, was low (6-19%). W h e n the same characters were assessed after 27 months, none of the neighbour-specific differences in morphology were retained. It is concluded that the original results reflected developmental differences carried over from the pasture, and that diversifying selection is not of importance in the maintenance of morphological variation in this population. Key words: Plasticity, common garden, diversifying selection, pasture, Trifolium repens. demonstrated (Snaydon & Davies, 1976; Turkington & Harper, 1979). One plant species in which high levels of variation T h e study of variation in species and populations has are maintained within populations is white clover, been an important area of emphasis in plant ecology Trifolium repens L. In old permanent pastures, since Turesson (1922) (see reviews by Heslopseedling establishment is rare (Harberd, 1963) and Harrison, 1964; Langlet, 1971; Briggs & Walters, T. repens populations are maintained primarily 1984). Much of this interest has centred on the through clonal propagation. Because of this clonal a m o u n t of genetic variation in populations and its T. repens populations might be expected to habit, maintenance (Lewontin, 1974; Brown, 1979; Ennos, show dominance by a few very large clones and to 1983). Spieth (1979), for example, has referred to the have low overall levels of genetic variability (Harper, explanation of the high levels of variation within 1978 ; Burdon, 1980; Bulow-Olsen, Sackville-Hamilpopulations as 'the central problem in population ton & Hutchings, 1984). Studies on clonal diversity genetics'. In plant populations, part of the solution may involve enx'ironmental heterogeneity and diver- in British pastures, however, have found no evidence that clonal depletion is a dominant trend (Cahn (St sifying selection (Hedrick, Ginevan & Ewing, 1976; Harper, 1976; Burdon, 1980; Gliddon & Trathan, Hamrick, 1982; Ennos, 1983). Diversifying selection 1985). Conversely, studies have consistently shown has received attention because it is clearly involved that pasture populations of T. repens contain genetic in t h e fine-scale differentiation of plant populations \'ariation for a wide array of characters (summaries distributed across sharp enx-ironmental transitions in Burdon, 1983; Turkington & Burdon, 1983). For (Jain & Bradshaw, 1966; Antonovics, Bradshaw & example, Burdon (1980) found that each of 50 clones T u r n e r , 1971). Even finer scale differentiation assocollected from a single pasture was morphologically ciated with mosaic en\ironments has also been INTRODUCTION 37° Fl- (• Evans and R. Turkington distinct. Aarssen & Turkinj^ton (1985 6) also found hi^h levels of variability for morphological characters. Diversifying selection has been specifically invoked as a possible general explanation for the maintenance of variation in populations of T. repens in these permanent pastures (Burdon, 1980). One prominent element of environmental heterogeneity in pastures is provided by the mosaic of patches of the several species of perennial grasses which dominate the pasture vegetation. The dense and continuous nature of this vegetation ensures that T. repens plants must exist in close association with the grasses. This suggests a potential for selection in the field on T. repens individuals for compatibilities (persistence, growth, and reproduction in association with neighbouring plants) with tbe grasses. Many studies have in fact shown T. repens to be responsive to changes in the identity of neighbouring plants, including the pasture grasses (review in Chestnutt & Lowe, 1970). The clonal habit of T. repens will infiuence its response to a beterogeneous environment. As a consequence of tbe horizontal growth of stolons, difterent parts of a clone will commonly become widely separated. As a clone expands through a spatially heterogeneous environment, its ramets may respond difierently in different microhabitats (Solangaarachchi, 1985; Newton, 1986). If the T. repens population is composed of genotypes which differ in their habitat suitabilities, the distributions of different clones may diverge. In the pasture, genotypes may he selectively sorted among the various grass patches rather tban selectively eliminated. Evidence oi' this neighbour-specific sorting or diversifying selection in T. repens is provided by studies wbich have shown local specialization of T. repens genotypes collected from patches of different grasses (Turkington & Harper, 1979; Aarssen & Turkington, 1985«; Gliddon & Trathan, 1985). The work to be reported here investigated the possibility tbat neigbbour-specific diversifying selection could account for the high levels of morphological variation recorded in T. repens populations. AND iVIF.THODS 7he pasture 71ie material used in this study was collected from a 4()-year-old pasture near Aldergrove in the Fraser Valley of British Columbia (49° 03' 45" N. Iat., 122° 30'45" W. long.). Tbe pasture is grazed intermittently from spring or midsummer until late fall by a herd of 20 30 cows. 7^he pasture receives no fertilizers other than animal excretions, occasionally supplemented with barnyard manure. The pasture is presently composed of 17 dicot species and 15 grasses. The most common of these are Trifolium repens L., Lolium perenne L., Holeus lanatus L., Dactylis glomerata L., and Poa compressa L. These five species constitute approximately 75 "„ of the total vegetation cover. Percentage cover by species \-aries both seasonally and annually (Parisb, 1986). All four of the common grasses are perennial and exbibit varying degrees of patchiness, constituting up to 100"/;, local cover in some patches. Patches dominated by L. perenne or //. lanatus may exceed 1 m^, while those dominated by D. glomerata and P. eompressa are usually smaller. D. glomerata patches are usually less than 0'25 m" but are very dense. The sizes and locations of patcbes vary both annually and seasonally. Collection and propagation of material Material for the study was collected on May 17 and 18, 1982. One hundred patches of each of the grasses (L. perenne, II. lanatus, D. glomerata, and P. eompressa) were identified visually. Ideally, a patch consisted of at least 75",, cover of one of the four grasses over an area of 0 5 nr^. One ramet of T. repens was collected from each of the 400 patches. The T. repens collection thus consisted of four subpopulations, one for each of the most prominent grass neighbours. Each subpopulation contained 100 T. repens genets. Replication was by patches rather than by repeated collections per patcb, in order to minimize the likelihood of genet duplication (Harberd, 1961). I'^ach ramet consisted of a 4 cm section of stolen apex, including the apical bud and the first node with its leaves. Any roots remaining on the node were removed. Only stolons which were actually rooted in the patch were used, and preference was given to stolons with several rooted nodes and/or branches within the patch. However, only a few rooted clover stolons were found within patches of £). glomerata. Where none were available, a stolon growing tbrough the patch and rooted on both sides was chosen. The 400 ramets were transferred to a common garden at the Plant Sciences Field Station on the University of British Columbia campus. Each ramet was treated with rooting hormone and planted separately in a 15 cm pot with 'field station soil'. The pots were arranged in groups corresponding to the four neighbour-specific sub-populations. Experiinent 1 In July, 1982, 6 weeks after establisbment, a cutting was taken from each pot and replanted in a second pot in the same manner as before. The second set of pots was given one teaspoon of fertilizer (20",, N, 20",, P,,O,, 20",, K.,O) per pot. After a further 10 weeks (September 1982) the second generation of ramets was harvested by removing the entire plant Irom the pot and washing soil from the roots. Each plant was assessed for the following set of characters : Morphologieal differentiation in Trifolium repens Numbers of parts. A count was made ot the number of primary stolons rising directly from the main taproot, and of the total number of stolons including secondary and tertiary branches from the primaries. T h e count did not include dormant or unelongated b u d s (less than 10 mm). The number of internodes on t h e longest primary stolon was also counted. Sizes of parts. The lengths of the longest primary stolon and the longest secondary stolon derived from it were measured. Three internodes were measured on the longest stolon. The choice of internodes excluded the actively elongating stolon tip and unelongated internodes at the stolon base. Five leaves were selected from each plant. These were taken from the third to iilth nodes below the apex and from as many different stolons as possible. The length of each petiole from the top of the stipule attachment to the base of the leaflets, and the length and width of each terminal leaflet were measured. T h e leaves were collected and pressed just prior to harvesting of the plants. Measurements were made on t h e pressed leaves. Weights. Root and shoot material was dried separately for at least 3 d at 100 °C. Root and shoot dry weights were measured directly and total dry weights derived from them. Leaf tnarks. Kach plant was scored for two types ot leaf markings, red flecks and white che\ rons, which are genetically independent (Carnahan et al., 1955; Corkill, 1971). The red flecks were recorded as present or absent, and the white chevrons classified into types which are known to be genetically distinct (Carnahan et al., 1955). Recording was done independently by three obser\ers on two occasions. 371 Experiment 2 The clones originalh- planted in May 1982 were repropagated in March 1983, August 1983, and July 1984, gi\'ing a total of four generations (27 months) since the original collection ; there were 376 sur\'i\'ing clones of the original 400. Following the final propagation, the set of ramets was given the same period of growth (10 weeks) as the 1982 set. The plants were harvested in October 1984 and a revised set of characters measured. The original character 'primary stolon number' was eliminated because of uncertainty in classifying stolons as primary or secondary. Only the total number of stolons was recorded, l l i e character 'internode number' was also eliminated because of the difficulty of counting the first few internodes which are commonly unelongated. One additional measurement, total stolon length, was recorded. \N.M.Y.SK .• S AND RKSIILTS Statistical analyses Analyses of \ ariance were performed on the data to test for the presence of variance cotnponents representing differences among the four neighbourspecific subpopulations of Trifotium repens genets. The proportion of the overall variance accounted for by differences among the subpopulations was estimated tor those characters in which a significant among-subpopulations variance component was present. The two-leaf mark characters were analysed using the non-parametric Kruskal-Wallis test. The \ariances of all characters in expt 1 were highly heterogeneous. Logarithmic transformations (Sokol & Rohit, 1981) reduced heterogeneity in most cases and retnoved it for fi\-e characters. Analvses of Table 1. Summary of analyses of variance for 12 morphological characters 0/Trifolium repens {data from expt I) MS, Root weight Shoot weight Total weight Primarv stolon number Total stolon number Internode number Primarv stolon length Secondarv stolon length Internode length Petiole lengtli Leaf width Leaf length 0-559 0-686 0-645 0-245 1610 14-8 0-042 0-119 0-078 0-461 0-062 0-145 MS,,. 0-096 0-117 0-101 0-036 442 9-07 0-026 0-041 0-023 0-018 0-005 0-006 Significanee \ ariation ** 4-80 4-81 5-33 5-73 2-65 0-65 0-66 2-02 2-42 20-19 11-75 19-61 «* *# #* * n.s. n.s. * ** ** ("0) MSj^, mean square of differences among neighbour-specific subpopulations of the T. repens collection ; MS,,, error mean square; Variation is the percentage of total variation attributable to differences among subpopulations. **, P < 0-01 ; *, P < 0-05 ; n.s., not signilkant. 372 R. C. Evans and R. Turkington variance were carried out on the raw or transformed data as a p p r o p r i a t e . Because the data from expt 2 did not show significant levels of heterogeneity a m o n g variances, t h e analyses were carried out on raw data. Experiment 1 For the data from expt 1, significant (P < 0-0'5) among subpopulations variance components were detected for 10 out of 12 characters, all except internode numbers and primary stolon lengths (Table 1). The proportion of the overall variation accounted for by differences among the subpopulations ranged from 2 '/„ for secondary stolon number to 20",, for petiole length. Neither of the leaf-mark characters showed any significant differences among the subpopulations. Because most of the characters in the study involved the sizes and numbers of parts, it is likely that they vary in parallel. In fact, of the 66 possible character pairs, all except four were significantly correlated ( f <0'01), indicating that the data set might be best represented by a single character, probably reflecting plant size. A principal com- ponents analysis (PCA) was performed to reduce the 12 morphological characters in the original data set to a new set of five uncorrelated multivariate characters (Table 2). The first principal component was, as expected, primarily infiuenced by variation in plant size, weights and stolon numbers contributing most strongly to the pattern. It accounted for .S4",, of the total variation in the data set. Analyses of variance were performed on the principal component scores for the T. repens genets. Significant (P < 0-05) among-subpopulations variance components were present for four out ofthe first five principal components, all except component four (Table 3). Experiment 2 [n contrast to the results after 4 m o n t h s in the c o m m o n garden, the m e a s u r e m e n t s taken after 27 m o n t h s showed no evidence of differences among the s u b p o p u l a t i o n s for any of the characters (Table 4). In no case was a significant a m o n g - s u b p o p u l a t i o n s variance c o m p o n e n t detected. As was the case for the earlier data, all but three of the character pairs were significantly correlated Table 2. Principal components analysis of 12 niorpholofiical characters q/Trifolium repens {data from expt 1) Component PC 1 PC 2 Variation ("c) CumuliUive .S4-36 .S4-36 14-32 68-68 Root weight Shoot weight Total weight Primary stolon number Total stolon number Internode numher Primary stolon length Secondary stolon length Internode length Petiole length Leaf width Leaf length 0-322 0-362 0-363 ()-2LS 0-300 0-150 0-303 0-32.S 0-281 0-288 0-234 0-248 -0-287 -0-178 -0-220 PC 3 10-97 79-65 0-106 0-022 0-050 0-253 -0-426 -0-361 -0-0.S3 0-244 0-095 0-292 0-225 0-404 0-397 0-063 -0-598 -0-403 -0-277 0-188 0-168 0-373 0-344 PC 4 PC 5 6-18 85-83 4-32 90-15 -0-068 0-041 0-007 -0-232 -0-224 -0-235 0-079 -0-093 - 0-668 0-121 0-190 0-627 0-479 0-156 -0-363 -0-310 0-082 0-030 0-161 0-064 0-276 -0-524 0-123 0-208 Tahle entries are coefficients for each character for the first five principal components. The per cent variation is the ainount of vari-<uion in the multiv-ariate data set which is explained by each component. Table 3. Summary of analyses of variance of principal component scores for four neighbour-specific subpopulations of TrifoUum repens {data from expt I) Component PCA 1 PCA 2 PCA 3 PCA 4 PCA 5 MS,,. 66-16 5-67 13-02 0-04 2-13 6-03 1-69 1-22 0-75 0-50 Ahhreviations are as in Tahle 1. Significance Variation (%) * *# 9-73 2-49 9-46 n.s. ()-(){) *# 3-36 Morpholoi(ieal differentiation in Trifolium repens 373 Table 4. Summarv of atialyses of varianee for 10 niorphologieal characters o/Triioliun-i repens (data from expt 2) 0-26.S 1-11 2-13 50-8 455 000 6 140 26-2 96-4 1-08 1-61 Root weight Shoot weight Total weight Total stolon number Total stolon length Primary stolon length Internode length Petiole lengtli Leal width Leaf length MSj. Significance 0-460 1-37 3-23 80-4 474000 5 300 31-9 97-1 2-89 2-04 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 1-1.S. Abbreviations are as in Table (/^<()-01). Frincip;il components analysis showed that the main trend in the data (first principal component accounted for 53 "„ of the total variation) again reflected plant weights and stolon numbers (Table 5). Analyses of variance of the principal components scores detected no significant differences among the T. repens subpopulations. DISCUSSION T h e results after 4 months in the common garden demonstrate that some niorphologieal variation in this Trifolium repens population is influenced by the species of grass dominating the immediate vicinities of T. repens genets. For 10 out of 12 morphological characters and four out of five principal components, a significant proportion of the variation \u the data was accounted for by the species of grass with which the T. repens individuals had been growing in the field (Tables 1 and 2). These results are similar to those of Turkington & Harper (1979) and of Aarssen & Turkington (198.Sa) relating \-ariation in dry weight production of T. repens genets to the identity of a genet's neighbour. This type of pattern has been interpreted as exidence that variation ii-i these T. repens populations is maintained by neighbourspecific diversifying selection (Turkington & Harper, 1979; Ikirdon, 1980; .Aarssen & Turkington, 1985 6). Such an interpretation, howexer, carries with it the implication that the \-ariation has a genetic basis. The results of the 27 month measurements (e.xpt 2), how-e\-er, suggest that for these morphological characters, such is not the case. Where there was a 6-7",, differentiation among the T. repens subpopulations after four months, after two adciitional y'ears none remained (Table 5). 'Inhere is thus no evidence for genetically-based morphological differences, nor is there ex-idence that diversifying selection is involved in the maintenanee of \-ariation in this population. The differences detected in expt 1 apparently refleeted the carry-over of de\-elopmental adjustments made by the plants to the \-arious environments presented by the different grasses. Oxer the 27 months of the study, the plants T a b l e 5. Prineipai eomponeiits analysis of 12 morphological characters rj/Trifolium repens (data from expt 2) Component Variation (",,) Cumulati\-e Root weight Shoot xveight Total weight T^otal stolon number Primarv stolon length Secondary stolon length Internode length Petiole length Leaf xvidth Leaf length PC 1 PC 2 PC 3 PC 4 PC 5 53-42 53-42 18-30 71-72 13-72 85-44 5-84 91-28 3-75 94-93 0-192 0-135 0-161 0-273 0-172 -0-144 -0-238 -0-262 -0-574 -0-583 0-228 0-062 0-126 0-262 -0-009 -0-522 -0-534 -0-236 0-347 0-356 -0-059 -0-084 -0-077 0-097 0-208 0-303 0-245 -0-861 0-171 0-090 -0-514 -0-168 -0-304 0-638 0-357 -0-170 0-087 0-207 0-027 -0-002 0-366 0-412 0-40S 0-331 0-392 0-291 0-262 0-255 0-159 0-150 T a b l e entries are coefficients for each character for the first Hve principal con-iponents. P e r cent variation is tlie amount of xarialion in the multivariate data sel which is explained by each component. 374 ^ - Evans and R. Turkington converged morphologically as they adjusted developmentally to the common garden environment. The extreme plasticity of T. repens is well known (Hill, 1977; Brougham, Ball & Williams, 1978). In fact, Bradshaw (1965) referred to the petioles of T. repens as one of the most plastic plant organs known. There is also ample evidence that 7\ repens responds plastically to changes in the surrounding vegetation, e.g. as a result of cutting frequency or fertilizer applications (Chestnutt & Lowe, 1970; Wilson 1978). It should not be surprising, then, that T. repens would also respond plastically to changes in its neighbours. It should also not be surprising that some of the plastic effects were retained for several months under common garden conditions. What was unexpected, however, was that all of the neighbourrelated differences would turn out to be no more than carry-over effects. The common-garden technique has been a fundamental tool of genecology since its inception (HeslopHarrison, 1964; Bradshaw, 1984). The method suffers, however, from uncertainty over the amount and persistence of variation that is carried over from the field environment to the garden. Studies on differentiation in perennial plants have commonly demonstrated carry-over effects among individuals transplanted to a common environment (Watson, 1969; Warwick & Briggs, 1979; Akeroyd & Briggs, 1983; Seliskar, 1985). In these cases, morphological convergence was noted over periods ranging from 6 months to 2 years. Clearly, carry-over effects must be anticipated if measurements are made within the first season after collection even if replanting has taken place in that period. In the present case, the pattern of variation after 4 months in the common garden was confounded by the persistent carry-over effects. The data from expt 1, taken alone, would have supported the misleading conclusion that diversifying selection does play a role in the maintenance oi morphological variation in this T. repens population. The usual method of addressing carry-over effects in transplanted material is to give the plants an establishment or preconditioning period during which their physiological states ostensibly become adjusted to the transplant environment and fieldderived resources are used up. A range of preconditioning periods have been used in previous studies of variation in T. repens. These include Snaydon (1962), at least 3 months; (4 18 months, Snaydon personal communication); Turkington & Harper (1979), 3 months; Gliddon & Trathan (1985), 5 months; Aarssen & Turkington (1985a), no preconditioning; and Aarssen & Turkington (1985 6), 4 months. Burdon (1980) did not report any preconditioning. Although these mostly exceed the period used in this study (6 weeks), tbe magnitude of the carry-over effects detected here stress the need for caution in interpreting short term transplant studies in T. repens. It should be noted that an extended period under experimental conditions may compensate for a short preconditioning period. Again, the 10-week experimental period used in this study was brief. For example, the two studies which demonstrated neighbour-specific differentiation in T. repens (Turkington & Harper, 1979; .Aarssen & Turkington, 1985/;) used experimental periods of 1 year. These studies, then, may have been less influenced by carry-over effects than the present one. In addition, reciprocal transplant/replant studies, as were used by Turkington & Harper (1979), Aarssen & Turkington (1985/J), and Gliddon & Trathan (1985) may be less affected by carry-over cffc'cts than common garden studies. The existence of significant variation among the subpopulations of T. repens after 4 months (expt 1) confirms that the different grass patches clo, in tact, represent an ein-ironmental heterogeneity. The lack of a genetic component to the variation, however, suggests that the biotic heterogeneity has not been an effective source of diversifying selection. Tbis could be a consequence of the high plasticity of T. repens. The range of morphologies available to a 7'. repens genet may be broad enough that most genets can persist across the range ol habitats represented by the various grasses in this pasture. This would be consistent with the pattern of survival found by Turkington & Harper (1979). If so, genetic variation for morphology would not be strongly infiuenced by selection. The high levels of genetic variability for morphology present in this and other T. repens populations thus remains unaccounted for. Several nonselective mechanisms for producing and maintaining variation could be applicable to this situation including somatic variation (Antolin & Strobeck. 1985; Gill, 1986) and the effects of soil microorganisms (Turkington, et al., 1988). Alternatively, Soane & Watkinson (1979) suggest that the expectation of genotypic depletion in pasture populations of clonal herbs may be unrealistic. Their modelling study found that even the low rates of seedling recruitment typical of these populations would be sufficient to maintain genotypic diversity. Parish (1986) documented an episodic recruitment of five established T. repens seedlings per m" on the site ofthe present study; Chapman (1987) reported similar values of T. repens seedling recruitment. There is clearly more than one way in which plants respond to the challenge of a patchy environment. Some species may develop genetic specialization to local conditions, a response that depends on environmental grain and predictability (Hedrick et al.. 1976; Ennos, 1983). Alternatively, a phenotypically flexible species may surv-ive under a range of conditions, its individual genotypes buffered from the effects of local selection and changing conditions (Sultan, 1987). Other studies with 7\ repens (Turk- Morphological differentiation in Trifolium repens ington &: Harper, 1979; Aarssen & Turkington, 1985a; Gliddon & Trathan, 1985) have provided evidence of local specialization in response to biotic heterogeneity. The present study, however, provides a contrasting picture of T. repens as a generalist with respect to the same conditions. .A C K N O VV L K n G E M I-; N T S T h i s research wiis funded by the Natural Sciences and Engineering Research Council of Canada. We are grateful to F . Ganders, J. Maze, T. McNeilly, J. Meyers, and R. Snaydon for comments on the manuscript, and to Janet E \ a n s tor encouragcnifnt throughout the project. We are also grateful lo Bill ami Mary Chard for access to their pastures. 375 KNNO.S, R. .A. (1983). Maintenance of genetic \ ariatioti in plant populations. F'coiutionary Bioiogy 16, 129-155. (JlI.L, U. E. (1986). Indixidual plants as genetic mosaics: ecological organisms versus exolutionary indixiduals. In: Piiinl Ecology (Ed. by M. J. Crawley), pp. 321-343. Blackxvell, Oxford. (;i.ll)l)ON, C. & TRArilAN, p. (1985). Interactions betxveen xvhite clover and perennial ryegrass in an old pertnanent pasture. In: Structure and Funetioning of Piant Popuiations, 7'oi 2 {VA. bx' J. Haeck & J. W. Woldendorp), pp. 161 169. North Holland Publishing Company, .Amsterdam. HAMRICK, J . L . (1982). Plant population genetics and exolution. American Journai of Botany b^, 1685 1693. HAHHI'RII, D . J . (1961). The ease for extensive rather than intensixe sampling in geneeology. Nezt: Ptivtoiogist 60, 325338. HAI!HI:RI), D . J . (1963). Obserxatioiis on natural clones of Trifoiium repens. Nezc Phytoiogist 62, 198 2t)4. HARI'FR, J . L . (1978). Plant relatiotis in pastures: .A keynote address. In: Plant Reiations in Pastures (VA. by J. R. Wilson), pp. 3-16. CSIRO, Melbourne. Hi:nRicK, P. W., (JINEVAN, M . E . & EWING, E . P. (1976). Genetic R F. F E R R N C Ii S .A.M<.SSI:N, I,. W . &|'^HKl^• • (;• ^)^J l• , R. (lyS.Sn). Bintic specialization b e t w e e n n e i g h b o u r i n g genotype.s in Loiiitm perenne and Tri- Jcilium repens from a permanent pasture. J(nirnai of Eeoiogy 73, 605-614. .A.ARSS!;N. I,. W . S; T I U K I N C ION, 1-t. (l'^cS.'i/)). Within-spccies diversity in natural populations of lioictis ianalus, Loiium perenne and I'rifoiiinn repens from four different aj,'ed pa.stures. Journai of Feoiogy 73, 869-886. .AKr-Hoii), J. R. & Unices, D. (I'J.S.^). (K'liecoloHical studies in Rumex erispus L. 1. (Jarden experiments usin^ transplanted material. Neiv Piivtologist 94, 309 .12."!, .ANTOI.IN, M . F . & STHOIU:CK, C . (1<),S.S). 'I'he population ^reneties f)f .somatic mutations in plants. .-Imerican Natiindist 126, 52 62. .ANTONOVICS, I., HR.\I).SIIA\V, A . D . & TrRM:!;, R. C (1971). I Ieil\ >' metal toleraiu e in plants. .Ad^'diices in Feoiogieai Resetircit 7, 1 S.S. BK.ADSII.^VV, A . D . (1965). Evolutionary signilicanee ol phenotypie phisticit\' in plants. Advances in Genetics 13, 115 155. BK.AIJ.SHAVV, .\. D . (1984). Eeological sJKnilicanee ol" ^enetie variation hetwfen jiopulations. h i : Perspeelires on Piant Fopiiiation Feoiogy (Ed. by R. Dirzo & J. Sarukhan), pp. 21.^ 228. Sinauer .Assoeiates. Sunderland, Mass. BHIGGS, I). &• \V.\i.Ti;ns, S . M . (1984), Piant I'ariaiion and Ez-olntion. Cambridge University Press, Cambridge. Bnot_-on.AM, R. \V., B..\i.i., V. R. & Wii.l.i.AMS, W. M. (1978). The ecology and management of white elover based pasture's. In: Plant Reiations in Pastures (VA. by J. R. Williatiis), pp. .iO9 324. CSIRO, Melboutne. B R O W N , A. H. D. (1979). ICnzyme polymorphism in plant populations. Theoretieai Popiiiation Bioiogy 15, 1 42. BtM.ow-()i.si:N, .A., SACKVII.I.I:-I I.^Mii.roN, N. R. & I U iciiiNcs, M . J. (1984). .A study of growth form in genets of Trifoiilim repens L. a.s alTected by intra- anil interphitit contacts. Oeeoiogia {Berlin) 61, 3X3 .^87. Bi-'RDON, J . J . (1980). lntra-s|5ecilie diversity in a natural popul a t i o n of Trifoiium repens. Journai of Eeoiogy 6 8 , 7 1 7 - 7 3 5 . B i - H i K x v , J . J . ( 1 9 8 3 ) . I ! i o l o « i e a l flora o f t h e B r i t i s h I s l e s : Trifoiium repens. Journai of luoiogy 71, 307 330. C.-\J1N, M. C;. & HAHPI:U, J. L. (1976). The bloloKy of tlie leaf m a r k polymorphism in Trifoiium repens L. I. Distribution ol phenotypes at a local scale. Heredity 37, 309 325. C.^R.MAHAN, H. L., H I M . , II. D., HANSON, .A. .A. i- BHOUN, K . <.'• . (1955). Inheritance and frequencies of leaf markings in white cUncr. Journai of Heredity 46, 109 114. CH.APMAN, I). 1'', (1987). Natural re-seeding aiul Trifoiium repens demography in H'azctl hill pastures. 11. Seedling appearance a n d aurvivii]. J<iurnai of Appiied Feoiogy 24, 1037 1043. CHESTNL^TT, D . M . B. CV Lowi:, J. (1970). .Agronomy of white clt)\ er-j{rass swards. Rex iew. Oec/isionai Symposium of tiie British Grassland Society 6, 1 9 1 - 2 1 3 . CoRKii.i., L. (1971). Leaf markings in white clover. Journai of Heredity 62, 307 310. polymorphism in heterogeneous environments. Annual Revietc of Feoiogy and Systemalics 7, 1-32. HESI.OP-HARRISON, J . (1964). Forty years of genecology. .4draiiees in Feoiogieai Researeh 2. 159-247. HIM., J . (1977). Plasticity of white clover groxvn in competition xvith ryegrass. Report of tiie W'eisii I'iant Breeding Station for 1976, Aberystxvyth, pp. 24-25. JAIN, S . K . & BRADSHAW, A. D. (1966). l-lvokitionary divergence among ailjacent plant populations. I. The ex idence and its theoretical analvsis. Heredity 21, 407 41. LANGI.I'T, O . (1971). Txvo hundred years getiecology. Taxon 20 653 722. LEWONTIN, R. C . (1974). 'Tiie Genelie Basis of Fvoiutionary Ciiange. CoUimbia I. nixersity Press, Nexv York. Ni:xvi<)N, P. V. D. (1986). Tiie estabiishment, growth, and fate of white eiover piants u'itii speeiai reference to the piiysioiogy of sioion grotciii. Ph.D. Thesis, L'niversity of Wales. PARISH, R. (1986). The roie of disturiianee in permanent pasture. Ph.D. Thesis. Unixersity of British Columbia, N'ancouver. ScM.SKAU, 1). .M. (1985). Effect of reciprocal transplanting between extremes ol plant zones on morphotiictric plasticity of five plant species in an Oregon salt marsh. Canadian Journai of Botany 63, 2254-2262. SNAVDON, R. W . (1962). Tlie groxvth and competitive ability of contrasting populations of 'Trifoiium repens L. on calcareous and acid soils. Journai of Feoiogy 50, 439 447. SNAVDON, R.W. ^^ DAVIHS, M . S . (1976). Rapid population difierentiation in a mosaic en\ironment. I\'. Populations of Antiioxantiiuin oiloratum at sharp boundaries. Hereditv 37 9 25. SdANi:, 1. 1-;. & WATKIN.SON, A. R. (1979). Clonal xariati<in in populations of Raniiniuiiis repens. Nejc Piiytoiogist 82, 557 573. SOKOI., R. R. & Roiii.F, 1". J. (1981). Biometry, 2nd Edn. I-'reetiian, San I'rancisco. Soi.ANGAARACiiciii, S. .M. (1985). 'The nature and amtioi of itranciling pattern in white eiorer (Trifolium repens /,.). Ph.D. 'Phesis, diixersity of Wales. Sl'linH, P. 'P. (1979). Environmental heterogeneity: a problem of contradictory selective pressures, gene How, and local polymorphistn. American Naturaiist 113, 247 260. Sn.TAN, S.I-;. (1987). Evolutionary implications of phenotypic plasticity in plants. Evoiutionarv Bioiogv 21, 127 178. P• t RESSON, Ci. (1922). 'Phe genecological response of the plant speeies to the habitat. Hereditas 3, 211 350. PiRKiNcTON • , R. & Hi RiioN, J. J. (1983). The liiology of Canadian Weeds. 57. Trifoiium repens L. Canadian Journai of Piant Seienee 63, 243 266. TtRKiNCTON, R. &: HARIM:I(, J. L . (1979). The gtowth, distribution, and neighbour relationships of Trifolium repens in permanent pastures. \\. Fine scale biotic differentiation. Jfnirnai of Eeoiogv 67, 244-254. 'Pi RKINCTON, R., iloi.I., F. B., CHANXVAV, C . P . & TllOMP.SON, J. D. (1988). 'Phe influence of micro-organisms, particularly Riiizobium, on platit competition in grass-legume communities. In: Piant Population Dynamies (Ed. by .A.J. Davey, M . J . Hutchings &: .A. R. Watkinson). Symposium of tiie British Eeoiogiiai .Society (in t h e Press). 376 R. C. Evans and R. Turkington WARWICK, S. I. & RRIGCS, D . (1979). The genecology of lawn vveeds. III. Cultivation experiments with Achittca mittifotium L., Bettis perennis L., Ptantago tanccotalci L., Plantano major L., and Prunetta vtdgaris L. collected from lawns and contrasting grassland habitats. New Ptiytotogist 83, 509-.'J36. WATSON, P. (1969). Kvolutiiin in clcsely adjacent plant populations. VI. An enthomophilous species, Potentitia erecta, in two contrasting habitats. Heredity 24, 407 422. WlI.SON, J. R. (Ed.) (1978). Ptant Rotations in Pastures. CSIRO. Melbourne.