Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The evolutionary consequences of interactions between plants in permanent pastures Mehrhoff, Loyal Archie 1989

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1989_A1 M43.pdf [ 4.62MB ]
Metadata
JSON: 831-1.0098261.json
JSON-LD: 831-1.0098261-ld.json
RDF/XML (Pretty): 831-1.0098261-rdf.xml
RDF/JSON: 831-1.0098261-rdf.json
Turtle: 831-1.0098261-turtle.txt
N-Triples: 831-1.0098261-rdf-ntriples.txt
Original Record: 831-1.0098261-source.json
Full Text
831-1.0098261-fulltext.txt
Citation
831-1.0098261.ris

Full Text

THE EVOLUTIONARY CONSEQUENCES OF INTERACTIONS BETWEEN PLANTS IN PERMANENT PASTURES by LOYAL ARCHIE MEHRHOFF, III A THESIS SUBMITTED IN THE REQUIREMENTS DOCTOR OF PARTIAL FULFILMENT OF FOR THE DEGREE OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF BOTANY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August 1989 0 Loyal A. Mehrhoff, 1989 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r a n a d v a n c e d d e g r e e a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l m a k e i t f r e e l y a v a i l a b l e f o r r e f e r e n c e a n d s t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s m a y b e g r a n t e d b y t h e h e a d o f m y d e p a r t m e n t o r b y h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t b e a l l o w e d w i t h o u t m y w r i t t e n p e r m i s s i o n . D e p a r t m e n t o f Botany  T h e U n i v e r s i t y o f B r i t i s h C o l u m b i a V a n c o u v e r , C a n a d a D 16 August 1989 D E - 6 ( 2 / 8 8 ) ABSTRACT The evolutionary consequences of interactions between neighbouring plants of Trifolium repens and three selected grasses were studied in a series of different-aged pastures. Experiments were conducted at three different levels of interaction; between species, between populations, and between individuals. The primary objective was to examine under natural conditions the relevance of two revolutionary theories of species coexistence - niche differentiation and competitive equivalence. Additional studies focused on the differentiation of T. repens subpopulations in response to neighbouring grasses, on the presence of carry-over effects, and the validity of extrapolating the results of greenhouse studies to predictions of natural pasture processes. Plants collected from different pastures showed different patterns of growth in a common garden. These patterns appeared to be related to the age of the pasture from which the plants were originally collected. Occasionally, species which were poorer competitors in 3'ounger pastures became superior competitors in mixtures from older pastures. Plants grown under natural field conditions were affected by both competitive and non-competitive forces. Non-competitive factors such as grazing, disturbance, and/or abiotic conditions, killed more individuals and accounted for more variation in growth than did competitive factors. However, plant interactions were important, including both competitive effects on neighbours and competitive responses to neighbours. These competitive effects and responses were not symmetrical and each appeared to behave independently. The balance between the intensity of intra- and inter-specific competition was critical. Two pairs of species which were more adversely affected by interspecific competition showed evidence of increased niche separation in older compared to younger pastures. A third pair of species was more affected by intraspecific competition and showed a decrease in niche separation in older pastures. These results ii support the theory that competition leads to changes in the pattern of resource use. The ability to predict the growth of species under natural pasture conditions from either greenhouse or controlled garden conditions was poor. This lack of predictability, when coupled with the observed rapid evolution of a species' competitive ability, suggests that caution should be exercised in developing theories based upon greenhouse growth characteristics or on "species" characteristics. Unlike previous studies in a Welsh pasture, no evidence of long-term specialization of Trifolium repens to patches of neighbouring grasses was found. Differences in the age of the pastures, and the average patch size of the neighbouring grasses (environmental grain) may explain these differences. in T A B L E O F C O N T E N T S ABSTRACT i i LIST OF TABLES vi LIST OF FIGURES v i i ACKNOWLEDGEMENTS v i i i OVERVIEW 1 CHAPTER 1: TESTS OF EVOLUTIONARY THEORY Introduction 4 Methods General methods 6 Statistical analyses 8 Results G r o w t h 12 Competitive equivalence 15 Niche partit ioning 16 Competitive effect and response : 16 Discussion Tests of hypotheses 21 Ecological perspectives , 22 Evolut ionary perspectives 24 CHAPTER 2: POPULATION DIFFERENTIATION Introduction 27 Methods General methods 29 Statist ical analyses 30 Results S u r v i v a l and growth 31 Specialization , 35 Discussion Selective pressures ; 36 Microevolutionary trends 38 Specialization 38 iv CHAPTER 3: INDIVIDUAL P E R F O R M A N C E Introduction 40 Methods Genera l methods 41 Stat ist ical analyses 43 Results Over lap i n the performance of species 44 Environment-dependent performance 44 Morphology and performance 50 Discussion Performance reversals 53 Predictions of performance 54 Implications 56 CHAPTER 4: MICROEVOLUTION AND BIOTIC CONDITIONING Introduction 57 Methods Genera l methods 61 Stat is t ical analyses 62 Results 63 Discussion Biotic differentiation 72 Mechanisms of age-related change 72 SUMMARY AND CONCLUSIONS 76 LITERATURE CITED 81 APPENDIX 90 v L I S T O F T A B L E S TABLE 1-1. A N O V A s u m m a r y of net ramet production (NRP) 13 TABLE 2-1. A N O V A s u m m a r y of s u r v i v a l ( S U R V ) and net ramet production (NRP) 32 TABLE 3-1. A N O V A s u m m a r y of Trifolium repens and Lolium perenne transplants 47 TABLE 3-2. Correlations of morphological characters and growth 52 TABLE 4-1. A N O V A s u m m a r y of transplants 66 v i L I S T O F F I G U R E S FIGURE 1-1. Age-related changes i n mixture yield, competitive equivalence, and estimated niche separation 14 FIGURE 1-2. Relat ive balance of yield in monoculture and yield in mix ture . 17 FIGURE 1-3. Age-related changes i n competitive response - the abil i ty to tolerate competition 19 FIGURE 1-4. Age-related changes in competitive effect - the degree to which a species affects its competitor 20 FIGURE 2-1. Percent s u r v i v a l of Trifolium repens transplants 33 FIGURE 2-2. N e t ramet production (NRP) of Trifolium repens 34 FIGURE 3-1. Performance of different-aged populations in mixture 45 FIGURE 3-2. Ranked performance of individuals of different species in mixture 46 FIGURE 3-3. Environment-specif ic performance of Trifolium repens 48 FIGURE 3-4. Environment-specif ic performance of Lolium perenne 49 FIGURE 3-5. Age-related trends in var iabi l i ty 51 FIGURE 4-1. Relative advantage to Trifolium repens from natura l ly coexisting grass '. 64 FIGURE 4-2. G r o w t h of different-aged Trifolium repens 67 FIGURE 4-3. Performance of populations in different treatments 69 FIGURE 4-4. The effects of differential s u r v i v a l of clones during conditioning 71 FIGURE A-l. N e t ramet production of different-aged populations 90 FIGURE A-2. Product iv i ty of mixtures 91 FIGURE A-3. Equivalence of competitive relationships 92 FIGURE A-4. Es t imates of niche separation 93 v i i ACKNOWLEDGEMENTS I would especially like to thank my supervisor, Roy Turkington,, for guidance, support, friendship, and always being there when needed. He has helped to make my experience at the University of British Columbia a good one which I will long remember. My committee, Jack Maze, Gary Bradfield, Fred Ganders, and Bill Neill, deserve special mention for their extensive contributions during project formation, anatysis, and conclusion. While still far from perfect, this work is very much better for their efforts. Elena Klein, Roberta Parish, and John Spence provided a great deal of stimulating discussion, technical advice, and field assistance. Susan Mair, Malcolm Greig, and Frank Ho of the University of British Columbia Computing Centre, spent significant amounts of time addressing both statistical and graphical problems. I am grateful to Bill and Mary Chard for unlimited access to their pastures. Financial support for this work was generously provided by the Killam Foundation through a Killam Pre-doctoral Fellowship and the University of British Columbia through a University Graduate Fellowship. I am most thankful for the encouragement given by my parents and friends, without which this, thesis would probably never have been completed. Peggy Oyama, who also assisted with field work on rainy days and thesis typing, made this entire effort much more enjoyable. vm O V E R V I E W "Species are not static entities with fixed relations to the environment, but plastic elements, changing their genetic constitution under the influence of the physical factors of the environment and of the interactions with other species" (Lewontin 1968, p.3). This statement by L e w o n t i n calls attention to the f luid nature of species and acknowledges the importance of both biotic and abiotic selective factors. M y thesis examines how interactions between neighbouring plants affects the dynamics upon which Lewont in comments. I have chosen to address this topic by focusing on coevolutionary theories of coexistence. M y reason for doing so is because coexistence theory has been one of the dominant theoretical considerations of evolutionary ecology. In addition, this controversial area represents a large, wel l developed, and integrated body of theoretical l i terature (see Diamond and Case 1986). Whi le evolutionary events are k n o w n to occur rapidly (Snaydon 1985), they are often not measurable in the short term. This has forced researchers to infer evolutionary processes from current patterns of morphology, resource use, or competition. It has also led to confusion between processes w h i c h m a y m a i n t a i n coexistence and evolutionary processes which alter species relationships. Several researchers have attempted to avoid both the pitfalls associated w i t h in ferr ing historical events from current patterns and the impract ica l i ty of long term studies by investigating patterns in chronosequences of different-aged ecosystems (Parr i sh and B a z z a z 1982) and populations (Hancock and W i l s o n 1976, M c N e i l l y and Roose 1984, A a r s s e n and Turkington 1985a,b,c,d, 1987, P a r i s h and T u r k i n g t o n , Unpublished). Whi le this technique is potentially powerful , it requires a large number of controls to account for environmental and evolutionary events unique to part icular populations. G i v e n the prevalence of genotype x environment interactions reported in the l i terature, it is important to verify that experimental conditions reflect the actual conditions found i n the chronosequence. F a i l u r e to do so m a y result in an erroneous view of evolutionary events. To date, evolutionary theories of species coexistence have received only l imited empir ical testing on plants under actual field conditions (Berendse 1983, Ke l ley and C l a y 1987). The bulk of such tests have been based upon plants collected f rom natura l situations, but grown under controlled garden or greenhouse conditions (see Turkir igton and A a r s s e n 1984, A a r s s e n and Turk ington 1985d, E v a n s et al. 1985). These latter studies have not been evaluated for applicabi l i ty to natural conditions. Since the p r i m a r y objective of m y thesis is to determine the evolutionary consequences of plant-plant interactions, the thesis must place these interactions in perspective by addressing both competitive and non-competitive aspects of coexistence. Ideally, to do this, one needs to experimentally demonstrate that a plant actual ly affects (directly or indirectly) a neighbouring plant under " f i e l d " conditions and that a genetic change is affected in subsequent generations. Demonstrat ion of such a direct cause and effect is difficult and beyond the capabilities of this project (given restrictions on experimental duration). Consequently, this study w i l l attempt to infer processes from the patterns found in a chronosequence of different-aged pastures (freshly sown to 46 years of age). B u t , unlike previous efforts, this investigation simultaneously examines three theories of coexistence and considers: (1) evolutionary changes in both inter- and intra-specific interactions, as suggested by A l l e y (1982) and Underwood (1986); (2) evolutionary changes w i t h i n and between populations; (3) reciprocal influences between species, both the effect that one species has on a target species and the target species' ability to respond (to competition); (4) the effects of competitive interactions between plants versus the other, non-competitive, selective forces present in pastures; (5) the heritability of observed age-related changes in growth and competitive ability; and (6) the validity of using controlled greenhouse/common garden experiments to determine field processes. The thesis is presented in four chapters. The experiments are designed to answer specific questions pertinent to the above considerations and to address the more important questions multiple times from slightly different perspectives. Consequently, there is some overlap between chapters. In as much as is possible, chapters are treated in isolation and conclusions drawn based upon the chapter's results. The final concluding section attempts to integrate the results into a statement on the evolutionary consequences of competition. C H A P T E R 1: T E S T S O F E V O L U T I O N A R Y T H E O R Y I N T R O D U C T I O N Attempts to expla in the coexistence of s imilar species have occupied a prominent position in the development of ecology. M u c h of this debate has centered on the relative importance of competition. Three major theoretical explanations have emerged: coexistence due to the lack of competition, or the presence of niche separation, or competitive equivalence. A l l three theories deal explici t ly w i t h the occurrence, or non-occurrence, of competitive exclusion. The f irst of these explanations argues that factors such as predation (Paine 1966, Connell 1975), disturbance (Armstrong 1988), or stochastic events (Wiens 1977) prevent populations f rom at ta ining the densities necessary for competitive exclusion to occur. In this model, competition is infrequent and not an important evolutionary force (den Boer 1986, Si lvertown and L a w 1987). A l t e r n a t i v e l y , species m a y coexist if they have separate niches and use different resources. This pattern m a y result f rom coevolutionary changes between competitors which promote the use of different resources, i . e. niche differentiation ( M a c A r t h u r and L e v i n s 1967, L a w l o r and M a y n a r d - S m i t h 1976, Lundberg and Stenseth 1985). It m a y also be due to historic evolutionary events having nothing to do wi th competition (Connell 1980, Bazzaz 1987). A third alternative suggests that m a n y coexisting species are so s imi lar ecologically that they are equivalent species (Peterson 1975, A a r s s e n 1983, Goldberg and Werner 1983, Hubbel l and Foster 1986). T h u s , no species can exert sufficient influence to exclude any other species. Coexistence is facilitated because extinction is now dependent upon other events which m a y operate more or less randomly (Agren and Fagers t rom 1984, Hubbell and Foster 1986). S i m i l a r i t y in niche (T i lman 1977, 1981, A g r e n and Fagerstrom 1984, Ghi larov 1984) as well as competitive abil i ty (Aarssen 1983) have been proposed. A s w i t h coexistence due to niche separation, coexistence due to ecological equivalence could be due to either historic , non-competitive factors (Pacala 1988) or to coevolutionary competitive events (Aarssen 1983). These proposed coexistence mechanisms and their associated evolutionary theories are not mutua l ly exclusive. It is conceivable that they might a l l simultaneously play a role in mainta in ing coexistence in two or multi-species mixtures . This study attempts to test the applicabi l i ty of these theories using a chronosequence of five pastures ranging f rom freshly sown to almost 50 years old. The study investigates coexistence and neighbour interactions f rom both an ecological viewpoint and from a mechanistic , evolutionary perspective. Three specific hypotheses wi l l be tested: (1) that competition between neighbouring plants does not occur and, as a result, exclusion is dependent upon a n y of a number of non-competitive factors such as grazing, m o w i n g , disturbance, and/or abiotic conditions. Thus , changing the density or identity of neighbouring transplants should not affect an individual plant 's growth . If such changes in density or frequency do alter growth then this hypothesis would not be supported; (2) that competitive interactions (both inter- and intra-specific) promote the evolution of competitive equivalence, without altering resource use patterns. T h u s , estimates of equivalence in competitive relationships should be greater among plants f rom older pastures than f rom younger pastures, while estimates of niche separation show no change; (3) that interspecific interactions promote the evolution of differential resource use, while intraspecific interactions promote niche expansion. Thus , for species more affected by interspecific interactions, estimates of niche separation between competitors f r o m old pastures should be greater than those f r o m younger pastures. Species more affected by intraspecific interactions should show either no change in niche relationships or reduced separation in older pastures. M E T H O D S General methods The methodology centers on comparisons of the relative growth of ind iv idua l plants when placed into mixtures wi th neighbours collected f rom populations of different ages. The five populations used in the study were chosen to represent a chronosequence of pastures planted in 1985, 1982, 1977, 1958, and 1939, and representing different-aged populations of 0, 3, 8, 27, and 46 years , respectively. The 0 and 3-year-old populations were located at the U n i v e r s i t y of B r i t i s h Columbia , South Campus field station, while the 8, 27, and 46-year-old populations were pastures on a dairy fa rm near Aldergrove, B r i t i s h Columbia (described in A a r s s e n and Turkington 1985a). Based upon previous results (Turkington and A a r s s e n 1984, A a r s s e n and Turkington 1985d), three species pairs were chosen; one which was thought to eventual^ 7 lead to competitive exclusion (Dactylis glomerata - Lolium perenne), and two which showed pre l iminary indications of convergence in competitive abilities (Holcus lanatus - L. perenne and Trifolium repens • L. perenne). Clones of each pair of species were collected as coexisting neighbours (where both sets of roots were intertwined). S ix of these neighbouring pairs of clones were collected, for each species pair , in M a r c h 1985 from three different-aged populations; the H. lanatus - L. perenne pair was collected f rom 8, 27, and 46-year-old populations, D. glomerata - L. perenne f rom 3, 8, and 46-year-old populations, and the T. repens - L. perenne pair f rom 0, 8, and 46-year-old populations. Time and resource l imitat ions made it impract ica l to conduct experiments on more than three populations at a t ime. A l l clones were mult ipl ied under greenhouse conditions unt i l sufficient mater ia l was available for t ransplant ing. Exper imenta l plantings of al l three species pairs were planted adjacent to the 0 and 3-year-old populations, under common garden field conditions, at the U n i v e r s i t y field stat ion. Two replicates were planted. The T. repens - L. perenne pair was also replicated three times in both the 8 and 46-year-old pastures. E a c h pair of six clones was grown under three sets of conditions: a ful l density monoculture in which each transplant had 12 .25cm 2 of available space (approximate n a t u r a l field density, Mehrhoff , unpublished data); a half density monoculture where t ransplants had 2 4 . 5 c m 2 of available space; and a series of mixtures w i t h three different sets of interspecific neighbours at fu l l density (12 .25cm 2 per plant). Individual transplants were arranged in a manner s imi lar to A a r s s e n and T u r k i n g t o n (1985d). In the full density two-species plots, transplants were placed into a m a t r i x of 6 rows each with 4 plants , where the rows were separated by 2 .5cm, plants along a row separated by 5cm, and the second and fourth rows offset by 2 .5cm. Rows one and three consisted of the six different clones from species " i " and rows two and four the six clones of species " j " . E a c h transplant 's nearest four neighbours were of the other species. The outermost plants of the m a t r i x formed a protective border for the plot and were not harvested. The inner twelve plants were harvested and represented one transplant of each clone of species " i " and each clone of species " j " . F u l l density monocultures used the same matr ix , but two ramets f rom each of the six clones were harvested. H a l f density monocultures used a four x four m a t r i x , w i t h plants separated by 7cm. Rows were offset by 3.5cm and separated by 3.5cm. A g a i n , the border plants were not harvested. The inner six transplants (one f rom each clone) were harvested. This resulted in 21 randomly arranged subplots per replicate for the plantings at the field station. The Aldergrove plantings consisted of only 15 subplots per replicate, because half density monocultures were not planted (due to a l imited amount of propagated material) . A total of 4,896 ramets were transplanted and of these 2,376 were inside the protective border and available for analyses. Aldergrove plots were protected f rom dung deposition and t r a m p l i n g by a l m high fiberglass fence. G r a z i n g was not prevented and appeared to occur at an intensity s imilar to unfenced areas. Plots at the field station were clipped periodically to mimic the mowing and grazing regime present in the Aldergrove pastures. Exper iments on H. lanatus - L. perenne were planted in August 1985, D. glomerata - L. perenne in September 1985, and T. repens - L. perenne i n M a y 1986. Counts of net ramet production (NRP) and a harvest of above ground mater ia l was made on each transplant approximate ly one year after t ransplant ing. Because of the clonal nature of T. repens, it could not be recovered on a per individual basis and was instead collected on a per plot basis (inside the protective border). Statistical analyses The growth of transplants under different treatments was analyzed for both net ramet production (NRP) and dry mass (DM) using A N O V A on transformed data [ l o g 1 0 (x+1)], when appropriate, using: Y = C(O) + O + T + O x T + S + E (1-1) where Y is the yield in N R P or D M , C is C L O N E nested i n O R I G I N , O is the O R I G I N of transplants (population f rom which they were originally collected), T is T R E A T M E N T (density and identity of neighbour), S is S U B P L O T , and E is the remain ing unassigned var iab i l i ty . Due to the unbalanced nature of some contrasts, comparisons of O R I G I N and the O R I G I N X T R E A T M E N T interaction were made using t-tests. A yield suppression ratio (YSR) of component yie ld suppressions (YS) was calculated to estimate the s y m m e t r y , or equivalence, of competitive relationships (Aarssen 1988). If both species compete equally, the relationship is balanced and the result ing ratio is 1.0. If the two species do not compete equally, then Y S R w i l l be less than 1.0. Y S R = Y S i / Y S j = (Yij / Y i ) / (Yji / Y j ) (1-2) where Y i and Y j are the yield of species i and species j , respectively, in monoculture at half density, and Yi j and Y j i are the yield of i in j and j in i , respectively, at full density. The larger of Y S i and Y S j is a lways the denominator. The relative yield total ( R Y T ) , the s u m of relative yields (RY), was used as an estimate of niche separation between two species (de W i t and van den Berg 1965, Berendse 1983, A a r s s e n 1988). The assumption behind this estimate is that when the growth of a species in monoculture is equal to its growth in mixture the species are equal in their impact on each other. The estimate (RYT) w i l l be 1.0 in these cases and i m p l y either that the species completely overlap in resource requirements or that no competition is occurring. Values of R Y T greater than 1.0 indicate "overy ie ld ing" , and some degree of separation in niche. 1 R Y T = R Y i + R Y j = (Yij / Y i i ) + (Yji / Yj j ) • (1-3) where Y i i and Yjj are the yields of species i and species j respectively in monoculture at ful l density, and Yi j - and Y j i are the yields of i i n j and j in i , respectively, at ful l density. Values for R Y , R Y T , Y S , and Y S R were obtained for the matched n a t u r a l mixtures (e. g. 8-year-old H. lanatus and 8-year-old L. perenne). These values were averaged for each clone, transformed as above, and compared across O R I G I N wi th one-way A N O V A and t-tests. The relative importance of inter- versus intra-specific competition was determined by contrasting the performance of a clone in interspecific mix ture w i t h its performance in monoculture plots at the same density (pair-wise t-tests). Populations which had . greater performance in monoculture plots were considered to be relat ively more affected by interspecific competition than by intraspecific competition. L i k e w i s e , populations which had greater performance in two-species mixtures were considered to be more affected by intraspecific competit ion. Competit ive response (CR) is the abil i ty of a plant to wi ths tand the effects • of an interspecific neighbour. Competit ive effect (CE) is the effect that a plant has on a neighbouring plant. Both C R and C E were estimated for each population of each species. Competitive response was estimated by determining the proportional growth of plants in monoculture and plants in mix ture . F o r example , a clone which produced only half as m a n y ramets in mixture as in monoculture would have a C R of .50. A s C R increases, the species' abi l i ty to wi ths tand competition also increases. It is derived by: 11 C R i = Y i j / Y i i ; CRj = Y j i / Yj j (1-4) where, C R i and CRj are the responses of species i to competition from species j , and species j f rom i , respectively; Y i j and Y j i are the yields of species i in mixture wi th species j , and of species j w i t h i , respectively; Y i i and Yj j are the average yields of species i and species j , respectively, in monoculture at ful l density. Competit ive effect (CE) was estimated by determining the proportional effect which a plant has on an "average" neighbouring clone of the competing species. A s C E increases, the effect of the species on another species also increases. It was derived by: where, C E i and C E j are the effects of species i on species j , and species j on species i , respectively, and Y i and Y j values are as in (Eq. 4). Replacement series experiments (as formulated by de W i t and v a n den B e r g 1965), and their associated competitive values (such as R Y T ) , have been criticized on the grounds that competitive outcomes are not consistent over a l l experimental densities (Inouye and Schaffer 1981, F i r b a n k and Watkinson 1985, Connolly 1986, 1987), are sensitive to size inequalities among experimental species (Connolly 1986, 1987), and do not adequately separate the effects of inter- and intra-specific competition (Jolliffe et al. 1984, F i r b a n k and Watkinson 1985). Potential design shortcomings were mitigated through the use of approximately n a t u r a l field density, s imilar-s ized competitors, periodic c l ipping/grazing to further reduce size inequalit ies, and the use of two densities of monocultures (to differentiate the effects of inter- and intra-specific effects). C E i = 1 / (Yji / Yj j ); C E j =• 1 / (Yij / Y i i ) (1-5) RESULTS Growth N e t ramet production ( N R P ) , rather than dry mass ( D M ) , results are presented for two reasons. F i r s t , these two measures of growth showed s imi lar patterns. Second, N R P is probably a more accurate estimate of fitness than is D M for plants w i t h clonal growth because each ramet has the potential to be an independent individual . A n a l y s i s of variance (Table 1-1) indicated that individual plants ( C L O N E S ) accounted for a large proportion of var ia t ion in plant size. The density and identity of neighbouring plants ( T R E A T M E N T ) were also strong determinants of growth. T R E A T M E N T consisted of growing clones f rom different-aged populations (the clone's O R I G I N ) both in monoculture, at two densities, and in mixture wi th each of the three different-aged populations of the competitor species. The widespread significance of T R E A T M E N T in the T. repens - L. perenne plots located in the pastures indicated that plant neighbours did affect clone growth and that grazing or other non-competitive factors did not prevent competition (see Appendix F i g . A - l ) . O R I G I N was significant i n ha l f of the analyses and indicated that for three of the four species (D. glomerata, T. repens, and L. perenne), the populations showed some level of population differentiation. In total, the A N O V A model accounted for 30% to 65% of the var ia t ion in N R P . The productivity of al l three mixtures , as measured by N R P , showed a significant age-related increase (Fig. 1-1). These increases in mixture y ie ld were attributable to increases in the yield of one, but not both, of the two competing species (see Appendix F i g . A-2 ) . In the H. lanatus - L. perenne mix ture , increases in the yield of H. lanatus accounted for the overal l increase in the 13 T A B L E 1-1. A N O V A summary of net ramet production (NRP). Values are proportion of variation accounted for by the model. Species: Holcus . lanatus (H), Dactylis glomerata (D), Trifolium repens (T), and Lolium perenne (LH, LD, LT) with H, D, and T, respectively. S o u r c e H LH D LD T LT C l o n e ( n e s t e d i n 0) .23**** .12** .08 .27**** - .08**** O r i g i n (0) .01 .02 .06* .00 .21**** .06**** T r e a t m e n t (T) .09**** .10**** .06** .05* .17**** .04**** 0 x T .05 .06* .10** .05 .04 .01 S u b p l o t .01 .00 .00 .03** .23**** .11**** T o t a l .39**** .30**** .30**** .40**** .65**** .30**** n o t i n c l u d e d * P < .05 * * P < .01 * * * * p < .0001 >-5 OH 1 r 1 1 1 0 1 0 2 0 3 0 4 - 0 5 0 o "o _. c q VI a> CL a. D V) ~o 0) >-1.0 0.5-0.0 1 1 1 1 1 0 1 0 2 0 3 0 4 0 5 0 0 1 0 2 0 3 0 4 0 Population age (y) FIGURE 1-1. Age-related changes in mixture yield, competitive equivalence, and estimated niche separation, (a) Total yield (net ramet production, NRP), (b) equivalence of competitive abilities as estimated by the yield suppression ratio, and (c) niche separation as estimated by the relative yield total. For natural paired mixtures of: Holcus lanatus - Lolium perenne ( ); Dactylis glomerata - L. perenne (---); and Trifolium repens - L. perenne (—— ). Lines with arrows indicate significant changes from the youngest to the oldest population (P < .05, t-tests and regression). 1 mixture y ie ld . S i m i l a r l y , i n the D. glomerata - L. perenne mixture , increases in D. glomerata accounted for most of the increase in mixture yield. In the T. repens - L. perenne mixtures , L. perenne yield contributed most to the age-related increase. While T. repens showed no change in yield in these paired mixtures , it did show an age-related increase in yield when grown w i t h a s tandard L. perenne competitor (see Appendix F i g . A - l e , " M X " ) . These changes in mix ture productivity could be due to changes in one or both of the species w i t h respect to competitive abil i ty, resource use (niche), or non-competitive factors, such as response to grazing. If these patterns developed in response to factors having nothing to do wi th the influence of interspecific neighbours, then the patterns should be present both when interspecific neighbours are present and when they are absent. Thus , plants grown in monoculture and in interspecific mixtures wi th a s tandard competitor should show s imi lar age-related increases in N R P and paral le l the diagonal in F i g . 1-2. M o s t of the species do not show such a pattern, only T. repens demonstrated a signif icant age-related increase in both situations (P < .05, t-test, and see Appendix F i g . A - l e ) . Thus , it is probable that the changes are due to interactions between neighbouring plants. Competitive equivalence v The degree to w h i c h the addition of neighbours affects individual growth was estimated by the yield suppression ratio (YSR, E q . 1-2). Comparison of Y S R values for mixtures provides an indication of the relative equivalence of the competitive relationship. If two species affect each other in a s imi lar manner , then Y S R = 1.0 and the competitive relationship is balanced or equivalent. This is a m a x i m u m value and as the two species begin to differ in their competitive effect on each other, the value of YSR will decrease from 1.0 to 0, and the relationship will become more asymmetrical. The measure of equivalence (YSR) shows no consistent age-related trend (Fig. 1-lb and see Appendix Fig. A-3). There is no indication of a change from asymmetrical to equivalent relationships as would be hypothesized if plant populations were being selected for competitive equivalence. Niche partitioning An increase or decrease in RYT (Eq. 1-3) implies a corresponding increase or decrease in niche separation between neighbouring species. Relative yield total (RYT) increased with population age in H. lanatus - L. perenne and D. glomerata - L. perenne and showed a decrease in the T. repens - L. perenne pair (Fig. 1-lc and see Appendix Fig. A-4). These observed increases and decreases in niche separation were consistent with the predictions of traditional coexistence and niche differentiation theory (MacArthur and Levins 1967, Lawlor and Maynard-Smith 1976). Species pairs which showed evidence of niche differentiation (Fig. 1-lc) were the same pairs which were more adversely affected by interspecific competition in younger pastures (H. lanatus - L. perenne and D. glomerata - L. perenne) (Fig 1-2 a,b). The species which were more negatively affected by intraspecific competition {T. repens - L. perenne, Fig. l-2c) showed an age-related increase in niche overlap (less separation). Competitive effect and response Changes in the relative competitive ability between two species must distinguish between the "effect" that one species has on its neighbour's performance and the "response" of a species to the "effects" of its neighbours (Goldberg 1987, in press). Thus, an effect of one species on a second species is Monoculture (NRP) FIGURE 1-2. Relative balance of yield in monoculture and yield in mixture. Yield (net ramet production, NRP) for different-aged mixtures and monocultures. The yields of different-aged populations are connected from the youngest to oldest, with an arrow pointing towards the oldest. Populations more affected by interspecific competition are in the shaded area, and by intraspecific competition in the unshaded area. Filled circles indicate that the mixture and monoculture yields are significantly different from one another (P < .05, t-test). the reciprocal of the second species' response to competition. Mixtures of paired populations from the same pastures show how interactions between species are changing between different-aged pastures; it does not show how the individual species themselves are changing. Observed age-related changes in mixture performance could be due to changes in either the competitive response (CR) or the competitive effect (CE) of either species. In order to address how these species are changing, the monoculture performance of each different-aged population was evaluated against its mean performance when in mixture with the three different-aged competitors for CR (Fig. 1-3) and CE (Fig. 1-4). An age-related increase in CR indicates that the species is becoming better able to tolerate competition. An increase in CE means that a species is having an increasingly adverse effect on its neighbouring species. Both H. lanatus and D. glomerata showed an increase in NRP in paired mixtures with L. perenne (Appendix Fig. A-2). However, these increases have different causes. Increases in H. lanatus NRP were due to a reduction in L. perenne's negative effect on H. lanatus, not to improved competitive ability on the part of H. lanatus. Dactylis glomerata, on the other hand, shows an age-related increase in NRP (Appendix Fig. A-2b) due to both a decreased effect of L. perenne (Fig. 1-4) and an improved competitive response by D. glomerata (Fig. 1-3). Trifolium repens increased NRP in both monoculture (Fig. 1-2, Appendix Fig. A-le, "M") and mixture against a standard competitor (Appendix Fig. A-le, "MX"). But, decreases in its ability to tolerate competition (Fig. l-3c) and an increasingly negative effect of L. perenne have offset these gains and resulted in no net change in paired mixtures from the three pastures (Appendix Fig. A-2). a H. lanatus (- -) 2-, V) C ° 1-1 in <D Ql L. perenne ( ) 1 I I I I 10 20 30 40 50 2-i 1_ 0. glomerata (-™* L. perenne (— CO c o C L CO CD 1-I 10 I 1 1 1 20 30 40 50 2-i g± T. repens (— L. perenne (- -) 0 ) io c o CL co (1) 1-0 10 20 30 40 50 Population age (y) — i — 20 FIGURE 1-3. Age-related changes in competitive response - the ability to tolerate competition. An increase in response is an increase in performance when in mixture. For each of the three mixtures, lines with arrows indicate a significant age-related change in response (Eq. 1-4, P < .05, t-test). o v 1 U J |H. lanatus ( ) L. perenne ( I I I I I 0 10 20 30 40 50 2-i o a> i. 1 D. glomerata ( ) " l . perenne ( I I I I I 10 20 30 40 50 2-| T. repens (- -) O <D i. . perenne ( — r — 10 l 40 20 30 Population age (y) i 50 FIGURE 1-4. Age-related changes in competitive effect - the degree to which a species affects its competitor. A decrease in effect is a decrease in a species negative impact on its neighbouring species. For each of the three mixtures, lines with arrows indicate a significant age-related change in effect (Eq. 1-5, P < .05, t-test). DISCUSSION Tests of hypotheses 1. That competition between neighbouring plants does not occur and, as a result, species exclusion is dependent upon any number of non-competitive factors such as grazing, disturbance, and/or abiotic conditions. Not supported. In both the mown plots at the University of British Columbia field station and the grazed Aldergrove plots, neighbouring plants showed evidence of both inter- and intra-specific competition. 2. That competitive interactions (both intra- and interspecific) promote the evolution of competitive equivalence, without altering resource utilization patterns. Not supported. Estimates of competitive symmetry did not show the required age-related trends to indicate a balancing of competitive abilities. 3. That interspecific competition promotes the evolution of differential resource use, while intraspecific interactions promote niche expansion. Supported. Estimates of overlap in resource use (relative yield total, RYT) for all three species pairs supports the hypothesis! In both species pairs more affected by interspecific competition (H. lanatus - L. perenne and D. glomerata - L. perenne), RYT . values indicated that niche separation increased with population age. The species pair more affected by intraspecific competition, (T. repens - L. perenne) showed a decrease in niche separation. Ecological perspectives The performance of an individual t ransplant was affected by both the i density and identity of its neighbours. The precise manner in which study species affect each other is not k n o w n and was not investigated. Consequently the effects of neighbours m a y be due to direct competition for l imit ing resources, for space, or a number of other factors w h i c h promote "apparent competition" (Holt 1977, Connell in press). Regardless of the mechanism, the detrimental interactions observed in this study are referred to as competition. Disturbance, grazing, and other non-competitive factors did not eliminate competitive interactions in the Aldergrove pastures. However , this does not mean that these forces are unimportant . Other studies have shown that such factors can signif icantly alter competitive outcome (Berendse 1985, C r a w l e y 1988) and be strong selective forces which drive morphological change (Aarssen and Turkington 1985c), changes in s u r v i v a l patterns (Chapter 2), and mainta in pastures in a state of arrested succession (Crawley 1988). Non-competitive factors m a y also reduce competitive dominance to a threshold level at which niche differences or competitive similarit ies are adequate for coexistence or prolong the time necessary for exclusion. This s tudy, and m a n y others have concluded that competition is of widespread occurrence in pasture systems (Donald 1963, Snaydon 1978, Turk ington and A a r s s e n 1984, but see S i lver town and L a w 1987, for a dissenting view). Such m a y not be the case in other systems. Resource competition is more l ikely to occur in pastures than in m a n y systems because pastures are an art i f ic ial communi ty whose component species are planted at high density, are ecologically s imi lar , are commercial ly cultivated for rapid growth under a var iety of environmental conditions, and are artif icial ly maintained in a state of arrested succession for long periods of t ime. In addition, most of the "stock" seed used in the planting mixtures consists of lines grown under monoculture conditions for long periods of time. Thus , the t a x a in these mixes are usual ly not bred to be compatible wi th other species. The result is that species assemblages in pastures are part icular ly prone to demonstrate both competition and its evolutionary consequences. Several studies have suggested that performance of individual plants grown in isolation (Grime and H u n t 1975) or in monoculture (Goldberg and Fleetwood (1987) could be good indicators of competitive performance in interspecific mix ture . M o s t species in this study did not demonstrate this pattern: only T. repens and L. perenne showed any indication of such trends (Fig. 1-2). The dispari ty m a y be due to the use of s imi lar ly sized species as suggested by Gaudet and K e d d y (1988) or to the use of different-aged populations which showed evidence of evolution wi th respect to competitive abi l i ty. This last point is par t icular ly important because mixture yield showed much greater age-related increases than did monoculture yield (Fig. 1-2). In addition, general predictive theories based upon the relative competitive ability of species against a single " s t a n d a r d " species (Gaudet and Keddy 1988) take into account only competitive effect, and not competitive response. Both of these components of competition are important to plant growth in mixtures and both appear to behave and evolve independently. Chapters 2 and 3 suggest that while the pastures compris ing this chronosequence are not identical, at least the p r i m a r y study pastures (8, 27, and 46-year-old) are s imilar environmental ly . This f inding is cri t ical to the below discussion of microevolution in these pastures. Evolutionary perspectives Competit ive interactions not only occur among the study species, but appear to be an important evolutionary force. A l l of the study species, w i t h the exception of H. lanatus, showed a significant age-related change in competitive response, or competitive effect. Results support the hypothesis that interspecific competition leads to niche divergence between two species while intraspecific competition promotes the expansion of each species' niche (Lawlor and M a y n a r d - S m i t h 1976, S la tk in 1980, P i a n k a 1983). The balance between inter-and intra-specific competition m a y be one of the p r i m a r y factors determining evolutionary direction (Al ley . 1982). Species pairs more affected by interspecific competition in younger pastures (H. lanatus - L. perenne and D. glomerata - L. perenne) showed an increase in R Y T (more separation in resource use) in older pastures, while the pair more affected by intraspecific competition (T. repens - L. perenne) showed a decrease in R Y T (less niche separation). Fitness is concerned wi th the absolute number of s u r v i v i n g offspring, not the efficiency of producing those offspring; a decrease in niche separation and an increase in the effect of an interspecific neighbour on growth do not necessarily reflect a decrease in fitness. A l l species showed either an age-related increase in ramet production (NRP) in mix ture , or no change in N R P . None of the species showed a significant decrease in mixture N R P between the young and old pastures. Thus , a l l species either maintained or increased fitness. There is some evidence in animals that both resource use and competitive abil i ty m a y evolve separately (Ar thur 1982). F o r example, Hairs ton et al. (1988) have found that some sympatr ic populations have not evolved niche part i t ioning, but have instead shown changes in behavioural aggressiveness. A s imi lar hypothesis has been put forth for plants (Aarssen 1983, 1984, 1985), but has yet to be conclusively demonstrated. This thesis found no support for Aarssen's (1983) hypothesis that species from older pastures would, over time, become more equal in their competitive abilities, without changing resource use. Aarssen's (1988) recent study on some of the same species from these same pastures also failed to demonstrate that competitive ability evolves independent of resource use. Individual species in his study and this thesis, however, did not behave in the same manner. Differences between studies is presumably due to differences in experimental conditions (greenhouse vs field). Aarssen's (1983) proposed changes in competitive ability, without an accompanying change in the rate or timing of resource use, may be difficult to demonstrate in plants. Changes in morphology, metabolism, or the production of toxic substances will probably also alter the type, amount, rate, or timing of resource use (and thus affect niche). Competitive ability is a relative term and it is dependent upon both environmental conditions and neighbours. The importance of neighbours to competition is well illustrated in this chapter. Differences between populations of the same species were sometimes as dramatic as differences between species. It is, therefore, difficult to determine what constitutes "the" competitive ability of a species. In general, it may be more appropriate to identify both the species and the populations involved. Most communities are composed of unique assemblages of populations, and consequently, the development of competition patterns will be unique to each community, even if evolutionary process are the same. It is not possible to state that interspecific competition should either decrease or increase over time in all communities. Those communities dominated by interspecific competition should tend towards a reduction in interspecific competition, while communities in which intraspecific competition is more important should see a decrease in intraspecific competition. These communities may also see an indirect increase in interspecific competition due to niche expansion resulting from disruptive selection. Prediction of future competitive relationships will be dependent upon the initial balance of inter- and intra-specific competition of the populations forming the community, the frequency of these component taxa, and the nature of external factors which alter competitive relationships. C H A P T E R 2: P O P U L A T I O N D I F F E R E N T I A T I O N I N T R O D U C T I O N M a n y studies have demonstrated that plants display a marked tendencj' to differentiate in response to local environmental conditions (reviewed in Heslop-Harr ison 1964, Antonovics et al. 1971, Langle t 1971, B r a d s h a w 1972, Turkington and A a r s s e n 1984). In contrast, a number of investigations on other species have failed altogether to show population differentiation (McNaughton et al. 1 9 7 4 , Turkington and A a r s s e n 1984, de H u l l u 1985, Roy 1985). One of the most extensively studied species, white clover {Trifolium repens), has shown fine-scale differentiation to a var iety of environmental factors which imply at least some degree of adaptation to geographic location (Evans et al. 1985), soil conditions (Snaydon 1962), neighbouring grass species w i t h i n a pasture (Turkington and H a r p e r 1979, Turkington 1989), neighbouring genets of the same grass species (Aarssen and Turkington 1985b, Gliddon and T r a t h a n 1985), and to the strain of soil bacteria, chiefly Rhizobium leguminosarum biovar trifolii (Turkington et al. 1988, C h a n w a y et al. in press a) and Bacillus polymyxa (Chanway et al. in press b). Development of differentiation patterns over relat ively short time periods (< 100 years) presents an opportunity to pose a number of questions pertaining to microevolutionary processes. (1) Is fine-scale differentiation widespread, or merely an exceptional event? (2) W h e n differentiation occurs, how rapid is it? (3) Is competition the p r i m a r y selective force generating differentiation, or are non-competitive factors, such as grazing, mowing , disturbance, and environmental conditions more important? (4) Do populations show s imi lar differentiation patterns, or does each population display its own unique set of patterns indicating unique evolutionary processes? These questions are addressed through a series of transplantings of Trifolium repens among and wi th in different-aged pastures, among neighbouring grass species, and between neighbouring clones of the same grass species. The first of these questions, the presence or absence of population differentiation is not t r i v i a l . Whi le previous field studies on different systems have shown strong biotic and abiotic differentiation (Turkington and Harper 1979, Turkington and A a r s s e n 1984, E v a n s et al. 1985), studies on the Aldergrove pastures have relied upon greenhouse studies (Aarssen and Turkington 1985b). E v e n then, some have found only transient, phenotypic differences (Evans and Turkington 1988). Ver i f i cat ion of differentiation under field conditions would focus attention on the rate of microevolution, the forces responsible, and the interplay between intra-and inter-populational evolution. The majority of studies on T. repens imply intrapopulat ional differentiation and the fragmentation of the parent , population in response to selection. Based upon these examples of fine-scale differentiation, it appears that each T. repens population develops in a highly unique, independent manner . H o w e v e r , studies conducted on different-aged populations have sometimes shown age-related trends in morphology, growth, and/or competitive abil i ty (Parr i sh and B a z z a z 1982, A a r s s e n and Turkington 1985a,c,d, 1987, A a r s s e n 1988). That at least some populations show a directional, age-related response, indicates that there m a y also be some general processes at work in pasture systems. Simultaneous investigation of both differentiation w i t h i n a population and age-related directional trends of the entire population are necessary to determine their relative importance to microevolution in pastures. METHODS General methods Three clones of T. repens were collected in M a r c h 1985 from- each of three different-aged populations, an experimental pasture planted in ear ly 1985, a young pasture planted in 1977, and an older pasture planted in 1939. These populations represent a chronosequence of 0, 8, and 46 year old populations and were ini t ia l ly planted w i t h a s imi lar seed mixture . The 8 and 46-year-old pastures are located near Aldergrove, B r i t i s h Columbia and are described in A a r s s e n and Turk ington (1985a). The exact location of each collection site was permanent ly marked. The nine clones were mult ipl ied under greenhouse conditions for 16 months to minimize long-term carry-over effects (Evans and Turk ington 1988). Ramets (consisting of a stolon wi th two leaves and associated roots) f rom each clone were marked w i t h colour-coded wire and transplanted in ear ly A u g u s t 1986 into both the young and old pastures. In both pastures three 25m transects were established, one originat ing from each of the three T. repens collection sites. A t irregular ly spaced positions along each transect, three ramets (one from each of the three different-aged T. repens populations) were planted into patches of L. perenne (n=62), Dactylis glomerata (n = 29), and Holcus lanatus (n=24). Ramets were placed into smal l holes, made i n the exist ing s w a r d , approximately 2.5cm apart. N a t u r a l l y occurring T. repens was removed from the general area of transplants. N e a r the base of the transect a 0 .75m x 0 .50m plot was cleared of all vegetation. Two ramets of each of the nine clones were planted 10cm apart into each of the cleared plots. Plots were weeded periodically and protected f rom dung deposition and t rampl ing by a low, fiberglass fence. G r a z i n g was allowed inside these plots and appeared to be s imilar to grazing outside of the plots. Counts of s u r v i v i n g ramets were made at irregular intervals over a one year period, w i t h a final harvest of al l material in late August 1987 and early November 1987 for the old and young pastures respectively. S u r v i v a l , total dry mass , and net ramet production were measured at final harvest. Statistical analyses N e t ramet production (NRP) , dry mass (DM) , and s u r v i v a l ( S U R V ) were transformed [ l o g 1 0 (x+1)] when necessary and analyzed w i t h a single A N O V A model. However , it should be noted that analysis o f S U R V did not include a C L O N E variable because surv iva l values were averaged wi th in a clone. The model used was : Y = C(0) + O + T + P •+ S(P) + O x T + O x P + T x P + O x T x P + E (2-1) where Y is the dependent component ( N R P , D M , or S U R V ) , C is C L O N E nested in O R I G I N , O is the O R I G I N of transplants (the pasture f rom w h i c h transplants were collected), T is the T R E A T M E N T (cleared plot, or w i t h L. perenne, D. glomerata, or H. lanatus), P is the P A S T U R E that transplants were planted into, and S is S U B P L O T nested in P A S T U R E , and E is unaccounted for var ia t ion . Type I V sum of squares were produced (Freund and L i t t e l l 1981) and a l l factors were compared to the experimental error t e r m , wi th the exception of P A S T U R E and O R I G I N which were compared to S I T E ( P A S T U R E ) and C L O N E ( O R I G I N ) respectively. Due to the unbalanced nature of some contrasts, comparison of significant m a i n effects were analyzed w i t h t-tests. Interaction effects were further analyzed wi th A N O V A of a single factor, followed by t-tests on the groups of the other factor. F o r these analyses average clone values were used. RESULTS Survival and growth The presence of neighbouring plants ( T R E A T M E N T ) was the single most important factor affecting the s u r v i v a l and growth of TV repens transplants (Table 2-1). Both measures of growth ( N R P and D M ) produced s imi lar results and, as in Chapter 1, only N R P data are shown. There was a strong O R I G I N effect, indicating that the populations had differentiated. The effects of S U B P L O T , T R E A T M E N T X P A S T U R E , and O R I G I N X T R E A T M E N T X P A S T U R E were all relatively minor . The P A S T U R E , O R I G I N X P A S T U R E , and O R I G I N X T R E A T M E N T effects were in i t ia l ly significant for N R P , but only O R I G I N X T R E A T M E N T was significant at the conclusion of the study. Transp lant ing T. repens directly into patches of different grass species resulted in high mortal i ty and poor growth compared to performance in cleared plots. However , the identity of the grass neighbour was not significant (Fig. 2-1, F i g . 2-2). Th is pattern became established soon after t ransplant ing and remained throughout the study (Treatment effects, Table 2-1). Turnover of grass species during the study was high, w i t h approximately 20% of the grass clumps being replaced by a different species. Re-analyz ing data without those patches which changed identity did not alter results. Trifolium repens showed a clear age-related increase in s u r v i v a l (Fig. 2-lb) and N R P (Fig. 2-2b) wi th the 46-year-old pasture > 8-year-old pasture > 0-aged pasture. A s with treatment, the pattern became established wi th in one month of 32 T A B L E 2-1. ANOVA summary of survival (SURV) and net ramet production (NRP). Transplants surveyed after one month (initial) and after 15 months (final). Values are proportion of variation accounted for by the ANOVA model. I n i t i a l ! Final Source SURV NRP SURV NRP Clone (nested in 0) .04**** .04** Origin (0) .10*** .07* .10*** .07* Treatment (T) . 27**** .29**** 22**** . 15**** Pasture (P) .0 .01 .0 .0 Subplot (nested in P) .05 .01 .04 .02** 0 x T .02 .03*** .05 .06**** 0 x P .02 .01* .01 .0 T x P .01 .0 .01 .0 0 x T x P .0 .0 .01 .0 Total . . 4 7 * * * * .46**** . 4 4 * * * * . 3 4 * * * * not included i n analysis * P < .05 ** P < .01 *** P < .001 * * * * P < .0001 Treatment Origin Origin X Treatment FIGURE 2-1. Percent survival of Trifolium repens transplants, (a) Transplants placed into patches (Treatments) of Dactylis glomerata (D), Holcus lanatus (H), Lolium perenne (L), or cleared plots (CI), (b) Age of the pasture (Origin) from which clones were collected, and; (c) Origin X Treatment interaction. Origin of T. repens: 0-aged ( 0 ), 8-year-old ( EH ), 46-year-old ( _£j ). Within a group, bars sharing the same letter are not significantly different (P > .05, t-test on average clone values). 34 Treatment Origin Origin X Treatment FIGURE 2-2. Net ramet production (NRP) of Trifolium repens. (a) Transplants placed into patches (Treatments) of Dactylis glomerata (D), Holcus lanatus (H), Lolium perenne (L), or cleared plots (Cl). (b) Age of the pasture (Origin) from which clones were collected, and; (c) Origin X Treatment interaction. Origin of T. repens: 0-aged ( £3 ), 8-year-old ( El ), 46-year-old ( H ). Within a group, bars sharing the same letter are not significantly different (P > .05, t-test on average clone values). t ransplant ing and remained to the end of the study (Origin effect, Table 2-1). Specialization Evidence of an advantage to clones when in their "home" pastures or w i t h their "home" grass species was in i t ia l ly indicated by significant O R I G I N X P A S T U R E and O R I G I N X T R E A T M E N T interactions (Table 2-1). The O R I G I N X P A S T U R E interaction showed that clones f rom the 8-year-old and 46-year-old pastures each outperformed the other in their respective home pastures. Th is effect was temporary and by final harvest the clones from the 46-year-old pasture outperformed those from the 8-year-old pasture, in both pastures. Presence of a "home" advantage to the natura l ly neighbouring grass species should have resulted in a significant O R I G I N X T R E A T M E N T interaction, w i t h T. repens showing greatest performance in L. perenne (the "home" grass) compared to D. glomerata or H. lanatus (F ig . 2 - l c , 2r2c). While the O R I G I N X T R E A T M E N T effect was significant throughout the study, T. repens did not perform best in L. perenne. Trifolium repens does not appear to differentiate to neighbouring grass species. Fine-scale specialization to neighbouring clones could not be assessed through the A N O V A model. Instead, a comparison was made in each pasture wi th T.repens grown in its "home" S U B P L O T versus the two " a l i e n " S U B P L O T s . Clones f rom the 46-year-old pasture showed an ini t ia l "home" advantage (P = .04, n = 3, t-test) which disappeared by the end of the s tudy. Clones from the 8-year-old pasture showed no significant "home" advantage at any time. D I S C U S S I O N Selective pressures Neighbouring grasses have a major impact on T. repens; reducing s u r v i v a l by 75% and growth by 90%. The precise nature of this neighbour "interference" (sensu H a r p e r 1977) m a y or m a y not be due to competitive interactions, since alternative explanations involving "apparent" competition were not tested. It is possible that other, non-competitive, factors such as predation (Holt 1977) or mutualistic interactions with a third species (Connell in press) m a y indirectly cause these patterns. Regardless of the specific mechanism, neighbouring plants exert a strong selective force on the adult phase of the life cycle. It is also apparent that factors such as grazing, disturbance, and environmental conditions do not preclude neighbours f rom influencing one another. The lack of species-specific effects of different grass neighbours is consistent wi th studies by Fowler (1986) and Goldberg (1987) and m a y indicate that T. repens is exposed to "di f fuse" (Moen 1989), neighbourhood competition among competitive equivalents (Goldberg and Werner 1983, M i t c h l e y 1987, P a c a l a 1988). However , these findings contrast wi th previous field studies on T. repens which have shown neighbouring grasses to exert strong species-specific effects (Turkington and Harper 1979, Turk ington and M e h r h o f f in press, Turk ington 1989, Chapter 1). Reconciliation of these findings m a y be l inked to fundamental differences in experimental sites. A N o r t h Wales sheep pasture which is approximately 100 years old and composed of extensive grass patches dominated by single species of grass, shows significant species-specific effects of neighbouring grass species (Turkington and H a r p e r 1979, Gl iddon and T r a t h a n 1985). The B r i t i s h Columbia pastures on the other hand, lack species-specific effects, are younger (< 50 years), and, while composed of the same or similar species, do not have large, extensive grass patches (> 100 m2), but a mosaic of very small grass patches (< 1 m2). Populations in the Welsh pasture are probably exposed to biotic selection which is both more intense and more predictable. In addition, the greater age of this pasture means that these selective pressures have probably been operating for a much greater length of time. The high degree of spatial turnover and small size of natural grass patches in the British Columbia pastures may result in a system with inconsistent and ephemeral biotic selection. While selective forces exerted by neighbours may be less intense they may still be important, since studies using experimental plantings in the British Columbia pastures have shown that different-aged competitors of the same species can significantly affect growth. Comparison of age-related patterns indicates that selection due to non-competitive factors may be greater than that due to competition. The similarity of age-related patterns under both cleared plot (non-competitive) conditions and sward (competitive and non-competitive) conditions indicates that competition alone is not generating this pattern. Non-competitive factors may, therefore, be the more important selective force. Mortality patterns also provide information on the relative importance of competitive and non-competitive selective forces. In cleared plots, without the effects of competitors, transplants suffered approximately 60% mortality. Numerous biotic and abiotic factors could play a role in this mortality. Over 90% of the transplants inserted into the surrounding sward died. This mortality is most likely due to a combination of non-competitive factors present throughout the pasture and additional factors associated with the presence of neighbouring plants. The difference in mortality between the cleared and sward treatments was 30% and presumably reflects the net impact of adding grass neighbours. These patterns show that non-competitive forces are responsible for twice the number of deaths attributed to competitive effects, another indication of the relative importance of non-competitive factors. Microevolutionary trends Trifolium repens shows a strong age-related increase in s u r v i v a l , N R P , and dry mass production under both competitive and non-competitive field conditions. Average population fitness increased rapidly w i t h pasture age, showing significant differences wi th in eight years of sowing. This pattern is s imi lar to the age-related increase i i i T. repens dry mass and N R P found under inter- and intra-specific competition in the same pastures (Turkington and Mehrhof f i n press, Chapter 1). However , it contrasts wi th studies conducted in greenhouse or ungrazed garden plots which show either no age-related trends (Aarssen and Turkington 1985d, Chapters 3) or an age-related decrease in dry mass (Aarssen and Turkington 1985c, 1987, A a r s s e n 1988, Chapter 4). This inconsistency is presumably due to an interaction between growth and environmental conditions. Caution should, therefore, be exercised in applying results f rom greenhouse studies to field situations. Specialization Trifolium repens shows strong evidence of microevolution resul t ing in the differentiation of populations based upon pasture of origin, but not intrapopulational differentiation, the specialization of subpopulations to fine-scale local conditions. Greenhouse studies on plants f rom the B r i t i s h Columbian pastures do indicate a "home" advantage to neighbouring grass clones (Aarssen and Turkington 1985b), but this phenomenon is possibly due to either a Rhizobium -grass interaction or a Bacillus - grass interaction, rather than genetic differentiation of T. repens (Turkington et al. 1988, C h a n w a y et al. in press). Addi t iona l ly , some form of long-term conditioning s imi lar to that described by E v a n s and Turkington (1988) m a y result in transient specialization phenomena. The lack of persistent O R I G I N X E N V I R O N M E N T interactions and the presence of only a transient "home" advantage for neighbouring grass clones, implies a corresponding lack of fine-scale specialization. Thus , the selective factors associated wi th different pastures, neighbouring grass species, and neighbouring genotypes appear to be relat ively homogeneous in their effect on T. repens. The lack of identifiable "home" advantages contrasts wi th a number of other studies on T. repens where there were clear "home" advantages. (Turkington and H a r p e r 1979, A a r s s e n and Turk ington 1985b, E v a n s et al. 1985, Gliddon and T r a t h a n 1985, Turk ington 1989). Resolution of these apparently conflicting results m a y , again, be attributed to differences in the frequency and duration of encountering specific neighbours. Systems, such as the Welsh pasture, w i t h larger patch size and longer evolutionary periods m a y promote a more specialized system w i t h fine-scale biotic differentiation to both neighbouring grass species and neighbouring grass clones. In contrast, systems w i t h smal l , transient grass patches m a y show only a general plastic response to the presence of neighbouring plants (Turkington 1989, T u r k i n g t o n and Mehrhof f in press). C H A P T E R 3: I N D I V I D U A L P E R F O R M A N C E I N T R O D U C T I O N Plant species are not uni form entities. Intraspecific var iabi l i ty in m a n y species is quite high, w i t h populations differentiated spat ia l ly , in the form of ecotypes (Heslop-Harrison 1964, Langlet 1971, Turk ington and A a r s s e n 1984), and temporally due to rapid microevolution (Antonovics 1978, Turk ington and A a r s s e n 1984. Snaydon 1985, Chapters 1, 2). In addition, indiv idual performance and competitive abil i ty are highly dependent upon the competitive environment (Kelley and C l a y 1987, Turkington et al. 1988) and abiotic conditions (Aust in et al. 1985, Rice and Menke 1985, C l a y and L e v i n 1986, T i l m a n 1988). This combination of environment-dependent performance and rapid microevolution paints a picture of the species as a dynamic unpredictable entity whose performance is unique to every situation. This has lead to emphasis on studies at the level of populations or individuals , rather than the species, and to caut ionary notes against inferr ing species-wide performance f rom one or a few "representative" populations (Grubb 1985, Thompson 1988). Whi le it m a y appear that interspecific interactions are an unpredictable assemblage of special cases (Gaudet and Keddy 1988), other studies have shown predictabil i ty. Some species have demonstrated relatively constant competitive abil i ty against different species (Welbank 1963, Rouse and Radosevitch 1985, Fowler 1986, M i l l e r and Werner 1987), show no evidence of ecotype format ion (Turkington and A a r s s e n 1984, de H u l l u 1985, Roy 1985), have constant relative performance across a range of environments (Schmid 1985, Goldberg 1988), and have " in-the-f ield" performances which are predictable f rom monoculture or "greenhouse" performances A u s t i n et al. 1985, Rouse and Radosevitch 1985, Mitchley and Grubb 1986, Goldberg and Fleetwood 1987). These f indings have led to attempts to identify general ecological principles such as Gr ime ' s (1977) proposed relationship between life history strategy, plant growth, stress, and competition or Gaudet and K e d d y ' s (1988) efforts to develop a model of competition based upon competitive abi l i ty against a s tandard competitor. Such attempts have often been challenged as inappropriate simplif ications (Grubb 1985, Loehle .1988, Bai ley 1989, F i r b a n k and W a t k i n s o n 1989, S i lver town 1989). This chapter investigates the degree to which the growth of pasture species is dependent upon the abiotic and competitive environment. Specifically the study attempts to: (1) document the extent to which -pasture species and populations overlap in terms of individual growth, (2) determine the degree to which microevolution and environmental conditions can alter growth, and (3) determine whether growth "in-the-field" is predictable f rom growth in monoculture, growth under greenhouse conditions, or from general morphological characteristics. METHODS General methods This study uses data obtained from a previous field study (Chapter 2) and supplemental data collected under greenhouse conditions. In M a y 1985 three pairs of species were collected from three different-aged pastures; Holcus lanatus -Lolium perenne f r o m an 8, 27, and 46-year-old pasture, Dactylis glomerata - L. perenne f rom a 3, 8 and 46-year-old pasture, and Trifolium repens - L. perenne from a freshly sown (0 aged), 8, and 46-year-old pasture. S ix clones of each species were collected from each population. These clones were propagated under greenhouse conditions for five to twelve months, then planted in monoculture and in two-species mixtures . E a c h group of six different-aged clones f rom each of the three species pairs was grown in monoculture and wi th a l l three different-aged competitors f rom the same species pair . Exper imenta l treatments were replicated twice at the U n i v e r s i t y of B r i t i s h Columbia South C a m p u s field station. The T. repens - L. perenne pair was also replicated three times in both the 8 and 46-year-old pastures. Thus , each clone can be evaluated for its performance in monoculture and w i t h the same competitor species f rom three different-aged populations. In addition, the L. perenne clones wi th T. repens can also be compared across three different environmental conditions present in the freshly sown, 8, and 46-year-old pastures. The clonal growth form of T. repens precluded harvest on a per clone basis and this species was instead harvested on a per plot basis. A more detailed account of experimental conditions is found in Chapter 1. The L. perenne clones used in the experiments wi th T. repens were also grown under greenhouse conditions to obtain measurements on morphological characteristics. One ramet of each of the eighteen L. perenne clones was placed into a single 5cm x 5cm plastic pot w i t h a 2-1-1 m i x of peat, sand, and perlite, replicated five t imes, t r immed to 2cm high, and allowed to grow for five • weeks under w a r m greenhouse conditions. M a x i m u m clone height was measured weekly to obtain a weekly growth rate. A t the conclusion of the growth period, eight morphological measurements were taken and two derived ratios obtained: number of ramets , number of leaves per ramet , m a x i m u m leaf length, m a x i m u m leaf width , m a x i m u m ramet height, shoot mass , root mass , root length, root/shoot mass ratio, and root/shoot length ratio. Statistical analyses In order to address the predictability of growth in different environments , i performance ( N R P and D M ) in the greenhouse, experimental common garden, and actual pastures were correlated wi th each other and w i t h the greenhouse morphological values. The overal l performance of T. repens and L. perenne was evaluated w i t h A N O V A using the model: Y = C(0) + O + T + L + S(L) + O x T + O x L + T x L + O x T x L + E (3-1) where Y is either N R P or D M , C is C L O N E nested in O R I G I N , O is O R I G I N (the pasture the clone was init ial ly collected from), T is T R E A T M E N T (the density and identity of neighbours), L is L O C A T I O N (the pasture into which the clones were planted), S is the S U B P L O T (replicate plot) nested in L O C A T I O N , and E is unaccounted variat ion. Analyses on L. perenne were conducted on the basis of measurements of individual ramets . Trifolium repens was analyzed only on an aggregated plot basis. S i m i l a r i t y of multiple rank orders was determined by concordance analysis in M I D A S (Fox and Guire 1976). The skewness coefficient and coefficients of var iat ion were used to assess size inequities and hierarchy in different populations (Bendal et al. 1989). Values for these coefficients were generated by the U N I V A R I A T E command in S A S ( S A S 1985). RESULTS Overlap in the performance of species Plants of the same species, but collected f rom different-aged pastures showed large differences in their ability to grow in mixtures . Some of these differences appeared to be related to the age of the pasture from which they were collected (Fig. 3-1). F o r example, Holcus lanatus (Fig. 3-la,b) has greater net ramet production (NRP) in mixtures collected f rom the oldest (46-year-old) pasture than in mixtures from younger pastures. E a c h species pair also demonstrated age-related reversals in its "dominance" or f inal abundance in mixture . F o r example, the dry mass (DM) production of L. perenne in mixtures from young pastures is greater than the D M production of H. lanatus, but in older pastures H. lanatus has greater D M production. Because the populations differed in their average performance ( N R P and D M ) and demonstrated age-related reversals , indiv idual plants from the different species were not clearly separable in terms of performance (Fig. 3-2). The degree of overlap between populations var ied w i t h the species pair and the measure of growth ( N R P or D M ) . Environment-dependent performance The performance of T. repens and L. perenne clones were dependent (Table 3-1) not only on the identity (ORIGIN) of the populations and the measure of growth, but also on the identity and density of competitors ( T R E A T M E N T ) and the L O C A T I O N (the pasture) where these mixtures were planted. In addition, the relative performance of populations differed between the study pastures. Patterns of N R P were less affected by changing environmental conditions than were D M patterns (Figs. 3-3,3-4). O f part icular note was the negative correlation between 0_ Q_ 15-, 10 5-H. lanatus ( L. perenne ( ) 1 1 1 1 1 0 10 20 30 4-0 50 4 - i H. lanatus ( L. perenne ( C D 1 1 1 1 1 0 10 20 30 40 50 15-10-0. glomerata ( L. perenne ( ) 0_ Q_ 0 10 20 30 40 50 0. glomerata ( " L. perenne ( ) CO Q_ Q_ 15-, 10-5 -e L T. repens ( ) . perenne (-f 1 1 1 1 0 10 20 30 40 50 Population age (y) T. repens ( ) L. perenne (— ) CD 3 -2 -1 1 1 1 1 0 10 20 30 40 50 Population age (y) FIGURE 3-1. Performance of different-aged populations in mixture. Net ramet production (NRP, a,c,e) and dry mass (DM, b,d,f) for paired mixtures. Lines with arrows indicate a significant age-related change in performance (P .05, ANOVA). Rank Rank FIGURE 3-2. Ranked performance of individuals of different species in mixture. Net ramet production (NRP, a,c,e) and dry mass (DM, b,d,f) of individual plants in all three different-aged, paired mixtures. Individual values of T. repens were not available, so the three population values are shown. Open bars indicate L. perenne individuals and filled bars indicate individuals of the competing species. T A B L E 3-1. A N O V A summary of Trifolium repens and Lolium perenne transplants. Values are the proportion of variation accounted for by the ANOVA model for net ramet production (NRP) and dry mass (DM). Lolium perenne was analyzed by individual plant performance and Trifolium repens on a per plot basis. Trifolium repens Lolium perenne Source NRP DM NRP DM Clone (nested in 0) - — .09**** .06**** Origin (0) .08**** .07* .05* Treatment (T) .16**** . i g * * * * .05**** .02**** Location (L) .01 .14 .10** .26*** Subplot (nested in L) ,22**** .08**** .01* .01* 0 x T . OA* .04 .0 .01* 0 x L .03* .07*** .02** .07**** T x L . i o * * * * .10**** .0 .01 0 x T x L .07* .09* .01 .02* Total .80**** .79**** .35 .4g**** not included * P < . 0 5 ** P < . 0 1 *** P < . 0 0 1 Location Treatment Origin Garden Pasture F I G U R E 3-3. E n v i r o n m e n t - s p e c i f i c p e r f o r m a n c e of Trifolium repens. Net ramet production (NRP) and dry mass (DM) for (a,e) Location, (b,f) Treatment, (c,g) Origin, and (d,h) Origin X Location X Treatment interaction. Treatment is performance in monoculture (M), and with competitors of the same species, but of different ages (0, 8, and 46 years old). Origin effects are shown for a freshly sown (0-aged, £3 ), 8-year-old ( g§ ), and 46-year-old ( S3 ) population. The interaction is illustrated for clones grown in monoculture (M) and mixture (MX) in an experimental common garden (Garden) and in the two permanent pastures (Pasture). Within a group, bars which share the same letter are not significantly different (P > .05, t-test). Location Treatment Origin Garden Pasture FIGURE 3-4. Environment-specific performance of Lolium perenne. Net ramet production (NRP) and dry mass (DM) for (a,e) Locat ion, (b,f) Treatment, (c.g) Or ig in , and (d,h) Orig in X Locat ion X Treatment interaction. Treatment is performance in monoculture (M), and wi th competitors of the same species, but of different ages (0, 8, and 46 years old). O r i g i n effects are shown for a freshly sown (0-aged, El ), 8-year-old ( ), and 46-year-old ( • ) population. The interaction is i l lustrated for clones grown in monoculture (M) and mixture ( M X ) in an experimental common garden (Garden) and in the two permanent pastures (Pasture). The D M of L. perenne under " G a r d e n " conditions were reduced by a factor of ten to facilitate graphing. W i t h i n a group, bars which share the same letter are not signif icantly different (P > .05, t-test). DM of T. repens in monoculture in the experimental common garden and in the two permanent pastures (Fig. 3-3h); the non-significant age-related decrease in DM in common garden contrasts with the age-related increase in DM under pasture conditions for monocultures (P < 0.05, n = 3, df=l, Pearson r =-0.9985) and for mixtures with L. perenne (P < 0.05, n = 3, df=l, Pearson r = -0.9987). The constancy of rank orders, skewness of size distributions, and population variability were also affected by the age of the pasture into which they were planted (Fig. 3-5). Skewness and the coefficient of variation declined with pasture age, while the rank constancy increased with age. The mixtures planted into older pastures appeared to be less variable than the same mixtures planted into the youngest pasture. Again, NRP and DM showed similar age-related trends, but the NRP measures were more constant across the three different pastures. Morphology and performance The morphological and growth variables collected on plants grown in the greenhouse were compared to growth in the experimental common garden and in the pastures. The two permanent pastures showed similar patterns for NRP and DM production. Consequently, they were combined to simplify comparisons. Greenhouse growth variables were more frequently correlated with growth in the experimental common garden than with growth "in-the-field" (Table 3-2). Performance of L. perenne in the common garden at the experimental station was correlated with several of the greenhouse growth/morphology measurements. Most of these correlated measures were associated with measures of length or height. Only one of the greenhouse growth variables (number of leaves per ramet) was correlated with growth in pastures (DM in mixtures). Monoculture NRP was 51 0 + — i — 20 10 30 Pasture age (y) ~r— 40 50 10 20 30 I 40 50 Pasture age (y) FIGURE 3-5. Age-related trends in variability. Population skewness ( • ), coefficient of variation ( o ) as a proportion (not percent), and concordance of rank orders ( • ) for measures of (a) net ramet production (NRP) and (b) dry mass (DM) for Lolium perenne grown with Trifolium repens in different-aged pastures. T A B L E 3-2. Correlations of morphological characters and growth. Lolium perenne under permanent pasture (PASTURE), experimental common garden (GARDEN;, and greenhouse (GHOUSE) conditions. Plants in the PASTURE and GARDEN were transplanted into monospecific plots (mono) and mixture plots with Trifolium repens (mix). GHOUSE plants were grown in isolation (iso). Significant correlation coefficients are shown (Pearson r > .05, n = 18). a. Net ramet production (NRP) PASTURE GARDEN GHOUSE mono mix mono mix iso PASTURE - monoculture:NRP — .677 NS NS NS - mixture :NRP .677 - NS NS NS GARDEN - monoculture:NRP NS NS - .728 NS - mixture :NRP NS NS .728 - NS GHOUSE - NRP NS NS NS NS -(iso) - shoot DM NS NS NS NS NS " - root DM NS NS NS NS NS - number of leaves per ramet NS NS NS NS -.588 - leaf length NS NS NS NS -.819 - leaf width NS NS .542 NS NS - ramet height NS NS NS NS -.795 - root length NS NS -.510 .493 NS - ramet mass NS NS NS NS -.808 - root/shoot mass NS NS NS NS NS " - root/shoot length NS NS NS -.479 .791 - growth rate NS NS NS NS -.506 b. Shoot dry mass (DM) PASTURE GARDEN GHOUSE mono mix mono mix iso PASTURE - monoculture:DM — NS NS NS NS - mixture :DM NS - NS NS NS GARDEN - monoculture:DM NS NS - .904 NS - mixture :DM NS NS .904 — NS GHOUSE - shoot DM NS NS NS NS — (iso) - root DM NS NS NS NS NS " - number of leaves per ramet NS .672 NS NS NS - leaf length NS NS .577 .588 NS " - leaf width NS NS NS NS NS " - ramet height NS NS .518 .529 NS - root length NS NS NS NS NS - ramet mass NS NS NS NS NS " - root/shoot mass NS NS NS NS NS - root/shoot length NS NS -.569 -.566 NS " - growth rate NS NS NS NS NS correlated wi th mixture N R P when both were planted in the pastures and again when both were planted into the common garden near the 0-aged pasture. However , performance in pastures was not correlated w i t h performance in the common garden or the greenhouse. DISCUSSION Performance reversals Species, populations, and individuals did not m a i n t a i n a constant rank order of performance under different conditions. The relative growth of one species compared to its competitor was dependent upon both the identity of the competitor (competitor age) and the location (the pasture) into which the mixture was planted. The physical environment was also cr i t ical , w i t h a particular species or population of T. repens or L. perenne dominant under some, but not al l conditions. Changing the competitor and/or exper imental location resulted in reversals of dominance. This lack of constancy contrasts w i t h some reports of stable competitive relationships or uni form performance under different experimental conditions (Welbank 1963, Rouse and Radosevitch 1985, and Goldberg 1987). E n v i r o n m e n t a l conditions in the present study spanned a greater diversity of environmental conditions and this probably accounts for the discrepancy because other studies which have examined performance in widely differing situations have also shown performance reversals (Aust in et al. 1985. Rice and Menke 1985). Microevolution in these pastures has apparently occurred rapidly and resulted in dramatic age-related differences in the outcome of experiments. Therefore, microevolution is probably also at least par t ia l ly responsible for the observed reversals in abundance and the large overlap between species in terms of individual performance. S imi lar examples of microevolutionary changes in competitive performance are known only from this same pasture system (Aarssen and Turk ington 1985d, Aarssen 1988, Chapter 1) or f rom studies on animals (Seaton and Antonovics 1967, Antonovics 1976, 1978). Age-related changes in competitive performance were correlated w i t h the relative importance of inter- and intra-specific competition (Chapter 1). Some pairs of species collected f rom young pastures were re lat ively more affected by intraspecific competition than by interspecific competition. Other pairs showed the reverse: they were more affected by interspecific competition. These relationships changed in the different-aged pastures. Two pairs of species which in younger pastures were more affected by interspecific competition showed evidence of niche divergence and a reduction in interspecific competition in older pastures (H. lanatus - L. perenne and D. glomerata - L. perenne). The pair of species more affected by intraspecific competition showed an age-related increase in niche overlap (T. repens - L. perenne). This was presumably due to disruptive selection and a broadening of niches. Predictions of performance A number of studies have shown correlations between performance in mixtures and performance in monocultures, or correlations between mixture performance and growth attributes such as relative growth rate (Grime 1977, A u s t i n et al. 1985, Roush and Radosevitch 1985, M i t c h l e y and Grubb 1986, Goldberg and Fleetwood 1987, Gaudet and Keddy 1988). Previous experiments (in Chapter 1) indicated little correlation between a species' or population's monoculture performance and its performance in mixture . O n l y L. perenne and T. repens displayed any similarity between mixture and monoculture performance. This chapter focused on the predictability of individual L. perenne performance and found predictability to be dependent upon the experimental conditions and the specific estimate of growth used (NRP or DM). Monoculture NRP of individual L. perenne was sometimes, but not always, positively correlated with mixture NRP. The correlation was strongest and significant only when both mixtures and monocultures were planted into the experimental garden and, again, when both were planted into the actual pastures However. NRP (and DM) in the experimental garden were not correlated with NRP (or DM^ in the pastures. Thus, predictions about the performance of these species are most valid within similar environments. While individual growth in one environment was not significantly correlated with growth in another environment, some inferences can be made from the presence or absence of trends in population performance. Age-related increases in NRP of L. perenne and T. repens in pastures were generally maintained under experimental garden conditions, though usually these trends lost statistical significance due to a decrease in mean population differences (Fig. 3-3, 3-4). The corresponding trends in pasture DM were markedly different from the common garden trends, and negatively correlated. Measuring NRP, rather than DM, in situations where experimental conditions differ from natural conditions more accurately estimated growth under those natural conditions. Studies which attempt to evaluate microevolution or competitive relationships by using only DM in experimental gardens or the greenhouse may not accurately estimate natural competitive relationships. Implications Studies on plant populations have generally found that experimental conditions affect wi th in population var iabi l i ty , measurements of the inequality of size distributions such as skewness and the G i n i coefficient (Benjamin and H a r d w i c k 1986). The observed differences in skewness and the coefficient of var iat ion and corresponding increase in concordance indicate conditions in the older pastures promoted a reduction in var iab i l i ty . A s w i t h other systems (Weiner 1985, Weiner and Thomas 1986), experimental conditions wi th higher product ivi ty (the youngest pasture) also had greater var iab i l i ty . Increases in product ivi ty are thought to promote an increase in competition and, hence, an increase in growth differences between individual plants. While there were large differences in both growth and var iabi l i ty between the 0-aged and older (8 and 46-year old) pastures, there was very little difference between the 8 and 46-year-old pastures. Extensive overlap between species and the large within-species var iab i l i ty in individual growth make generalizations about "species" performance difficult , i f not impossible. Species, in the context used in this s tudy, are populations of individuals . Performance and competitive abil i ty m a y be more appropriately studied at the level of individuals or populations. The inabil i ty of this study to predict " in-the-f ield" performance from performance f rom non-field conditions emphasizes the difficulty in developing general predictive techniques derived f rom greenhouse studies (Grime 1977, Gaudet and Keddy 1988). Such attempts at generalization may adequately characterize natural phenomena in certain situations. However , for species such as those in this study, they m a y produce predictions seldom realized in nature. Generalizations m a y provide their greatest contribution by defining an "expected" condition or an init ial s tart ing point from which to study departures from expectations. CHAPTER 4: MICROEVOLUTION AND BIOTIC CONDITIONING INTRODUCTION Pasture communities and their component populations are dynamic systems, showing both seasonal patterns and a number of age-related directional trends (Sarukhan and H a r p e r 1971, Snaydon 1985). A s pastures become older, individual plants on average become shorter, more prostrate, and have reduced rates of flower production (Aarssen and Turkington 1985c). In addition, the var iabi l i ty associated wi th these morphological measures decreases (Aarssen and Turkington 1985c). In older pastures species distributions and associations become more consistent wi th fewer fluctuations in abundance or changes in associated species (Aarssen et al. 1984, A a r s s e n and Turkington 1985a, O'Connor and A a r s s e n 1988, Par i sh and T u r k i n g t o n unpublished). Interactions between species also change, wi th at least some species showing evidence of niche part i t ioning (Aarssen 1988, Chapter 2). In some old pastures, populations of Trifolium repens show evidence of fine-scale differentiation to both neighbours and abiotic conditions (Snaydon 1962, Turk ington and H a r p e r 1979, Gliddon and T r a t h a n 1985, A a r s s e n and Turk ington 1985b). This differentiation often results in populations or individuals matched to their specific local environment. Consequently, transplants from these environments show their greatest relative performance when replanted into the natura l or "home" environment and reduced performance under " a l i e n " conditions. Intra- and inter-specific competition are thought to have opposing effects on population var iab i l i ty , w i t h intraspecific competition acting in a disruptive manner and interspecific in a stabil izing or directional manner (P ianka 1983). Species such as Trifolium repens which are usual ly more affected by intraspecific competition (Harper 1977, C h i r w a 1985, A a r s s e n 1988, Chapter 1) would be predicted to show age an age-related increase in wi th in population var iab i l i ty and an increase in differentiation to specific grass neighbours. The rate at which these patterns develop is not k n o w n , but several studies indicate that such patterns can develop w i t h i n 10 years (Snaydon and Davies 1976, 1982, Snaydon 1978, Turkington 1979, Chapter 2). The mechanisms promoting age-related trends i n growth , competitive abi l i ty , and differentiation to local environmental conditions are also not precisely known. The majority of studies have implied that these trends are the product of various selective pressures associated wi th grazing, competit ion, and nutrient depletion, that the patterns have a genetic basis, and that the patterns are, thus, examples of rapid microevolution (Turkington and H a r p e r 1979, Turk ington and A a r s s e n 1984, Aarssen and Turkington 1985a,b,c,d, 1987, A a r s s e n 1988, Turkington and Mehrhof f in press, Chapters 1,2,3). Other studies, such as that by Evans and Turkington (1988), imply that some trends previously thought to be genetic may instead be due to phenotypic plasticity and carry-over effects. Microevolutionary changes occur if the attributes in question are heritable and actually passed to subsequent generations. H o w e v e r , m a n y attributes associated with competitive abil i ty, are thought to have low heri tabi l i ty (Venable 1984, Bazzaz and . Sul tan 1987, Sul tan 1987). Pasture plants are most ly long-lived, clonal species, which only infrequently reestablish by seed in mature pastures (Harberd 1963, Par i sh and Turkington unpublished). A s a result , observed changes in the structure of pasture populations and plant morphology may be due not to selection on multiple generations of populations, but to long-term depletion of the original generation sown into the pasture. If the attributes in question are heritable, then sexual reproduction results in microevolution, if not, then it m a y reset the evolutionary clock and negate observed phenotypic changes. Several studies support the view that pastures undergo a rapid depletion of individuals and genetic var iat ion. M o r t a l i t y in young pastures is very high, wi th approximately 90% of grass seedlings present at pasture init iat ion dying wi th in the f irst year (Charles 1961) and morta l i ty continues to signif icantly reduce the number of individuals present even in 10-year-old pastures (McNei l ly and Roose 1984). A a r s s e n and Turkington (1985c) found that even though old pastures had a great deal of morphological var iat ion between individuals , plants from younger pastures showed even greater between plant var iab i l i ty . Envi ronmenta l ly induced genetic changes or long-term conditioning m a y also conceivably p lay a role in such trends (Cullis 1987, Schaal 1988). Carry -over effects from a clone's previous environment have been shown to signif icantly affect clone morphology for weeks (Libby and J u n d 1962), months (Evans and Turkington 1988), or even years (Durrant 1962). E v a n s and Turkington (1988) have shown that Trifolium repens collected from different grass neighbourhoods were morphologically distinct from one another even after four months of common garden growth and one tissue turnover. However , these differences were transient and disappeared after two years of common garden growth and four tissue turnovers. The morphology and total nuclear D N A content of "plast ic" lines of Linum usitatissimum were altered for at least seven sexual generations by manipulat ing the level of nutrients applied to the plants, even when subsequent conditions are reversed (Durrant 1962, 1971, E v a n s et al. 1966). Cul l i s (1987) has suggested that these heritable genomic changes are caused by environmental "stresses" which cause an increase in the amount of nuclear D N A . L o n g - t e r m conditioning effects or environmental ly induced changes to the genome could potentially account for the patterns of neighbour specialization found in T. repens (Turkington and H a r p e r 1979, Gliddon and T r a t h a n 1985, A a r s s e n and Turkington 1985b) i f (1) conditioning effects were mainta ined through successive vegetative turnovers and (2) conditioning either improved performance in the "home" environment or restricted performance in the " a l i e n " environment . This chapter examines a number of related topics deal ing w i t h how T. repens populations change over time. The p r i m a r y objective is to determine i f populations show age-related trends in "home-site" advantage, population var iabi l i ty , and plant size. Secondary questions are concerned w i t h the speed w i t h which these trends might develop and the mechanisms which produce them. To address these questions, Trifolium repens was collected f rom a number of different-aged pastures ranging f rom freshly sown to almost 50-years-old. E a c h population was evaluated for its growth in mixture wi th grasses, the wi th in population var iabi l i ty , the presence of a "home-site" advantage, and response to long-term conditioning. Clones derived from commercial seed were compared to clones collected as mature plants from the oldest pasture and to clones derived from seed natural ly produced by these old clones. These comparisons aid in distinguishing between genetic and phenotypic patterns and can provide prel iminary indications of heritable changes. 6] METHODS General methods In A u g u s t 1985, eighteen clones of T. repens were collected f rom Lolium perenne and Dactylis glomerata neighbourhoods in pastures of various ages (freshly sown, 3, 8, 27, and 46-year-old pastures). The collections thus represent a chronosequence of T. repens f r o m two different grass neighbourhoods; L. perenne and D. glomerata. Comparisons of the relative performance of these two types of T. repens enables the determination of the rate w i t h which a population progresses f rom a relat ively homogeneous mixture to a population differentiated into subpopulations specialized to specific grass neighbours. E a c h clone was divided into two cuttings. One cutt ing was placed into a two inch plastic pot wi th a 2-1-1 m i x of peat, sand, and perlite (the "unconditioned" treatment). The other cutting was placed into a two inch pot having the same soil m i x , but which also had a dense growth of grass (the "conditioned" treatment). Clones or iginal ly from L. perenne neighbourhoods were grown i n pots w i t h L . perenne. Clones from D. glomerata neighbourhoods were grown wi th D. glomerata. A l l clones were mainta ined in these treatments for 20 months. A l l clones collected f rom the 46-year-old pasture also had mature, natural ly pollinated seed in the inflorescences. These seeds were germinated and one seedling f rom each "parent" clone was randomly chosen to represent the "of fspring" of each "parent " clone. These offspring clones were handled as before, wi th both conditioned and unconditioned treatments. A l l plants were clipped and watered regular ly , but not ferti l ized. Monoculture plots (2m x 10m) of both L. perenne and D. glomerata were established in A u g u s t 1985; these were mown and weeded regular ly . In A p r i l 1987, each surv iv ing clone was divided into two cuttings (with two leaves and associated roots). One cutt ing from each clone was planted into each of the two different species of grass. The cuttings were placed 10 cm apart in rows separated by 10 c m . Thus, T. repens clones from a series of young to old pastures were propagated in the greenhouse for 20 months. D u r i n g this propagation period each clone was grown both by itself and with a mixture of grass clones of the same grass species wi th which it had co-occurred in the pasture. A t the end of this propagation period, al l of the surv iv ing plants were divided into two s imi lar ly sized units. One unit was planted into a bed of L. perenne and the other unit into D. glomerata. E a c h clone was represented by four t ransplants ; a conditioned transplant placed into each of L. perenne and D. glomerata, and an unconditioned transplant placed into each of L. perenne and D. glomerata. Some clones died during conditioning and propagation resulting in a final total of 576 transplants representing 192 clones. A n unharvested border of T. repens cuttings was planted around the transplants to minimize edge effects. Transplants were allowed to grow for 6 months, then harvested. The net number of ramets was recorded and a l l above ground biomass collected, dried, and weighed. Statistical analyses Net ramet production ( N R P j and clone dry mass (DM) were transformed [ l o g 1 0 (x+lYj to homogenize variances. A single unbalanced A N O V A model (Freund and Li t te l l 1981) was used to determine the importance of exper imental factors to T. repens growth: Y = 0 + T + G + O x T + O x G + T x G + O x T x G + E (4-1) where Y is the dependent component ( N R P or D M ) , 0 is the O R I G I N of transplants (the pasture from which transplants were collected), T is the T R E A T M E N T (conditioned or unconditioned), G is the G R A S S plot which transplants were planted into, and E is unaccounted var ia t ion . Type I V s u m of squares were produced (Freund and L i t t e l l 1981) and al l factors were compared to the experimental error term. Due to the unbalanced nature of some contrasts, comparison of significant m a i n effects were analyzed wi th t-tests. Interaction effects were further analyzed w i t h A N O V A on a single factor, followed by t-tests on the groups of the other factor. Transformed N R P and D M values were also subjected to pair-wise t-tests of conditioned and unconditioned treatments. The relative advantage of each biotype was determined by dividing the performance of the biotype in its "home" grass by the performance of the other biotype in the same grass. Analyses were conducted separately for both conditioned and unconditioned t ransplants . Coefficients of var iat ion ( C V . ) were calculated for a l l populations to determine how var iab i l i ty changes with pasture age and the conditioning treatment. R E S U L T S Clones of T. repens collected from patches of D. glomerata and L. perenne did not show an advantage when grown wi th their respective " h o m e " grass (Fig. 4-1). Nei ther transplants which had been previously conditioned by greenhouse propagation w i t h grasses nor those unconditioned by growth in isolation showed FIGURE 4-1. Relative advantage to Trifolium repens from naturally coexisting grass, (a) Unconditioned, transplants propagated in isolation for 20 months prior to transplanting, (b) Conditioned, transplants propagated with grass prior to transplanting Values are proportional advantage or disadvantage in NRP of T. repens clones collected from the native grass compared to clones collected from the alien grass. For: Lolium perenne ( o ) and Dactylis glomerata neighbourhoods ( • ), filled symbols indicau th;1* the value is significantly different from the line of no advantage I ). Clones derived from seed produced by the 46-year-old clones are indicated by the unconnected symbols. an advantage to home grasses. Transplants derived from cuttings from exist ing clones showed an age-related decline in N R P and D M , though only the D M values were significant in the A N O V A (the Origin effect in Table 4-1, F i g . 4-2a) and regression (P = 0.001, n = 1 5 8 , r *=0 .07) Within-populat ion variabi l i ty as measured by both N R P and D M decreased between the 0-aged (young seed) and 8-year-old populations (Fig . 4-2b), but showed no subsequent change in variabi l i ty . The offspring clones derived f r o m the 46-year-old clones were the least variable. Clones from the 0-aged pasture which were derived f rom commercial seed (young seed), were compared with clones collected f r o m the oldest pasture (old parent) and clones derived from seed produced by the oldest clones (offspring or old seed) (Fig. 4-2a). The D M of "old seed" clones was signif icantly less than the D M of the "young seed", but not different f r o m "old parent" D M . This indicates that sexual reproduction does not el iminate age-related differences in D M . Differences in N R P , however, were not signif icant indicat ing a lack of age-related change. The species of grass into which transplants were placed was the single most important factor in transplant growth (Table 4-1). Trifolium repens grew better when planted into Lolium perenne than into Dactylis glomerata. This resul t was highly significant (P < .0001, Table 4-1) and consistent for both N R P (x = 47.7 vs 17.1) and D M (x = 2.23g vs 0.82g). Propagat ing T. repens clones with a grass prior to t ransplant ing them into a bed of grass ( T R E A T M E N T ) was not significant for either N R P or D M (Table T A B L E 4-1. ANOVA summary of transplants. Values are proportion of variation accounted for by the ANOVA model using net ramet production (NRP) and dry mass (DM). Source NRP DM Orig i n (0) .01 (#) 03*** Treatment (T) .00 .00 Grass (G) . 13**** .20**** 0 x T .01 (//) .02** 0 x G .01 .01 T x G .00 .00 0 x T x G .01 (#) .02*** Total .20**** .30**** (//) P < .09 P < .01 67 FIGURE 4-2. Growth of different-aged Trifolium repens. Performance in mixture with Lolium perenne and Dactylis glomerata. (a) Net ramet production (NRP) and dry mass (DM) of different-aged clones and (b) Coefficiences of variation (C.V.), for different-aged populations based upon NRP ( • ) and DM ( • ). The (a) DM line demonstrated significant variation (P < 0.05, ANOVA) among population values. Populations which share the same letter are not significantly different (P > 0.05, t-test). Values along the NRP line are not significantly different (P > 0.05, ANOVA). Lines in (b) were not evaluated for significance. D_ Q-Population age (y) FIGURE 4-3. Performance of populations in different treatments, (a) Net ramet production (NRP) and (b) dry mass (DM) of unconditioned ( O ) a n d conditioned ( # ) transplants from different-aged populations. Offspring clones derived from the 46-year-old clones are shown with unconnected symbols. Only the 0-aged transplants differ significantly between the conditioned and unconditioned treatments for NRP and DM (P < .05, t-test). Lines with letters are significant in ANOVA (P < .05). Populations which share the same letter are not significantly different (P > 0.05, t-test). s u r v i v a l in the unconditioned treatment, but in the conditioned treatment, clones from younger pastures had higher mortal i ty than those from older pastures (Fig . 4-4a). Clones w h i c h survived the unconditioned treatment, but did not survive the conditioning treatment w i t h a grass neighbour, performed more poorly than clones which survived both pre-treatments (Fig. 4-4b). 71 FIGURE 4-4. The effects of differential survival of clones during conditioning, (a) Percent survival of clones in conditioned ( # ) and unconditioned ( O ) treatments. Young populations (< 8 years old) have significantly greater mortality than older populations (P = 0.013, Fisher's Exact test, n=189). Cb) Average net ramet production (NRP) and (c) dry mass (DM) of clones in the naturally coexisting "home" grass and the "alien" grass for clones which survived both conditioned and unconditioned treatments ( • ) and those which survived only the unconditioned treatment ( 0 ). Within a group, bars sharing the same letter are not significantly different (P > .05, t-test). D I S C U S S I O N Biotic differentiation Trifolium repens shows neither an age-related increase in population variability nor evidence of within population biotic differentiation to grass neighbours. This conclusion concurs with some of the other studies conducted in these pastures (Evans and Turkington 1988), but it disagrees with studies conducted in an old Welsh pasture (Turkington and Harper 1979, Turkington 1989). This discrepancy may be due to differences in the relative strength, constancy, and duration of biotic selective forces resulting from the structure or grain of the environment, as perceived by T. repens (Turkington and Mehrhoff in press, Turkington 1989, Chapter 3). The Welsh pasture is twice as old as the oldest Aldergrove pasture and has large patches of dominant grasses (up to 100 m2). In contrast, the British Columbian pastures are younger (< 50 years) and composed of much smaller grass patches (< 1 m2). Given these differences in grain one would expect that the two pasture systems would differ with respect to the duration and intensity of interspecific interactions they impose on T. repens. Situations such as are found in the Welsh pasture should have greater potential for intense and consistent interactions between neighbours and would be expected to show greater biotic differentiation (Hedrick et al. 1976, Ennos 1983) than in the British Columbian pastures, and they do so. Mechanisms of age-related change This study was originally designed to examine the genetic and phenotypic components of intrapopulational biotic differentiation. While no differentiation within populations was found, the data do provide some information on differentiation between populations of T. repens. Conditioning affects the performance of T. repens in a complex manner . In experimental common gardens, both conditioned and unconditioned transplants demonstrated an age-related decrease in D M . The patterns found among conditioned and unconditioned transplants differ only in the relative performance of the 0-aged clones derived f rom commercial seed. Nei ther of these patterns has a strong age-related component, wi th pasture age accounting for 10% or less of total var iat ion. These patterns, though weak, were mainta ined after sexual reproduction. A s a result , it appears that even though biotic selective pressures appear to be lower in these pastures, age-related changes in indiv idual plant D M found by this and previous studies (Aarssen and T u r k i n g t o n 1985c, 1987, Aarssen 1988) m a y reflect microevolutionary changes. H o w e v e r , this age-related decrease in D M is not representative of actual in-the-field D M performance. It has been shown (Chapter 3) that clones of T. repens which produce an age-related decrease in D M under controlled, field station conditions show an age-related increase in D M when transplanted into actual pasture conditions. This negative correlation m a y be the result of microevolutionary changes producing plants of smal l stature which are potentially better at avoiding or tolerating grazing, but poorer competit ively in ungrazed (but mown) situations such as the experimental common garden. Propagation of clones w i t h grasses prior to t ransplant ing into a competitive situation apparently promotes the growth of clones derived f rom commercia l seed. This pre-treatment, however, has no effect on clones f rom older, long-established clones or on clones derived f rom the seed of these older clones. A n explanation of this pattern is problematic. Other studies have found that the effects of a previous environment m a y have only a temporary impact on clonal g r o w t h in the "new" environment (Libby and J u n d 1962. Watson 1969, W a r w i c k and Briggs 1979. A k e r o y d and Br iggs 1983, Evans and Turkington 1988). Some studies have shown carry-over effects into subsequent generations through a maternal effect (Durrant 1962. H i l l 1964, Schaal 1984). Such effects can impact on the performance of offspring, part icular ly in situations where in i t ia l seedling size is an important determinant of future fitness (Harper 1977, W i l s o n 1988). S t i l l other studies have indicated that the environment can induce genetic changes which are then passed on to future generations (Evans et al. 1966, Cul l is 1987). There are several potential explanations for this chapter's f indings: (1) plants which grew best in the experimental grass beds did so because they completed the greenhouse " t reatment" phase in a healthier condition and had an ini t ia l advantage over other transplants , (2) natural selection has selected against plasticity or the abi l i ty to respond to conditioning, and (3) environmental s t imul i have induced either phenot3 rpic or genetic changes on "plas t ic " young populations, but have no effect on established populations. O f these three explanations. the first is probably not applicable. Exper imenta l conditions did provide some artificial selection g iv ing an advantage to "conditioned" 0-aged clones, but this accounted for only a smal l portion of their superior performance. In addition, those populations w h i c h had the highest morta l i ty in the conditioned treatment also had the highest growth in the experimental grass beds, a result counter to the explanat ion. Dist inguishing between the second and third explanations is not possible and either explanation m a y be va l id . The abil i ty to respond plastically is thought to be a heritable trai t (Bradshaw 1965, W i l k e n 1977, Sultan 1987) and as such m a y be influenced by selection. A n adaptive explanation for a decrease i n plast ic i ty w i t h age is not obvious given our current knowledge of pasture systems. If this explanation were va l id , then it w i l l have to be reconciled with other studies which have demonstrated long-term carry-over effects in plants f rom older pastures. E v a n s and Turkington (1988) could not determine i f the observed conditioning effects were due to the influence of neighbouring grasses, or to abiotic microsite conditions associated w i t h grass distributions. The current study examined biotic conditioning in isolation from abiotic conditioning and found no evidence of such conditioning in older populations. Thus , reconciliation m a y be possible if their (Evans and Turkington 1988) carry-over effects were due to abiotic field conditions rather than to the effects of plant neighbours. Snaydon (1978) has previously suggested, but not demonstrated, that s imi lar patterns showing biotic specialization (Turkington and H a r p e r 1979) m a y actually be due to specialization to soil conditions. Since conditioned and unconditioned clones were not evaluated for genetic differentiation, it is possible that these patterns could also be due to environmental ly induced genomic changes. E n v i r o n m e n t a l "stresses" have been shown to induce a number of genomic changes, such as gene amplif ication, in some, but not a l l , lines of Linum usitatissimum (Cullis 1987). In part icular , both environmental induction in Linum and conditioning in T. repens show treatment effects only in previously unselected populations derived f rom "stock" seed sources. In Linum, these previously unselected, plastic populations became " f ixed" after induction and were no longer capable of responding to further inductions. The rapid loss of conditioning response in T. repens between the 0-aged and 3-year-old population m a y have resulted from a s imi lar phenomena. E v a l u a t i o n of offspring from conditioned and unconditioned treatments w i l l be necessary to resolve this question. Regardless of the precise mechanism, populations of T. repens demonstrate a heritable, age-related decrease in individual plant D M and its associated variabi l i ty , indicat ing that microevolution occurs in these pasture systems. SUMMARY AND CONCLUSIONS Microevolut ion in these pastures is rapid . E a c h of the four chapters in this thesis documented changes wi th in 8 years of pasture establishment (Figs. 1-1, 2-2, 3-1, 4-2). A l l species, except Holcus lanatus, showed age-related changes in growth and/or competitive abil ity. Microevolution occasionally resulted in the reversal of dominance in mixtures which blurred distinctions between the performance of different species. Competitive outcome was dependent upon both the species used and the age of the population used. These findings, and those of others (Aarssen and Turkington 1985c, Kel ley and C l a y 1987, A a r s s e n 1988) argue against the use of al l encompassing "species" performances to predict community dynamics . In many instances it w i l l be necessary to determine performances on a population-by-population basis. A s a result , ecological theories based upon the performance of species, such as those proposed by G r i m e (1977) and Gaudet and K e d d y (1988) must take into account the var iabi l i ty of species and their dynamic nature. If these factors are not incorporated, the result ing predictions m a y only be applicable to a very narrow set of conditions. Such theories must also address both competitive response, the abil i ty to tolerate competition, and competitive effect, the ability to affect neighbouring plants . Fa i lure to dist inguish between these two aspects m a y lead to confusion, since i n this thesis, these components were not correlated and each appeared to change independently of one another (Fig. 1-3, 1-4). It w i l l also be necessary to consider the va l id i ty of us ing performance under one set of conditions (e. g. greenhouses) to predict performance under different conditions (e. g. pastures). Predictions of performance in mixture which are based upon performance in isolation or in monoculture were generally not val id . However , both T. repens and L. perenne showed some predictability wi th in s imilar environments (Table 3-2). Extrapolat ion of growth performance to different environments was dependent upon the measure of growth used. Net ramet production in experimental conditions reflected growth under na tura l conditions more frequently than did D M . Consequently, N R P is more appropriate when experimental conditions differ f rom natural conditions. Selective pressures generated by neighbouring plants are probably not so strong as the summat ion of pressures generated by non-competitive factors. Most of the experiments reported have shown that competit ion-driven evolution potentially accounts for only a fraction of the total var iat ion and that it generates some, but not a l l , of the observed age-related patterns. Non-competitive factors, such as grazing, disturbance, and abiotic conditions apparently k i l l more transplants and account for more var iat ion than do competitive factors alone. Non-competit ive factors m a y also facilitate coexistence by reducing the intensity or frequency of competition (Harper 1969, Pickett 1980) to threshold levels at which differences in niche allow coexistence. This does not mean that competition is unimportant , just that it operates wi th in the parameters established by non-competitive forces. The patterns present among different-aged pastures indicate that competitive interactions do have evolutionary consequences, and that those consequences conform to the expectations of tradit ional coexistence and niche divergence theory (Lawlor and M a y n a r d - S m i t h 1976, P i a n k a 1983, Lundberg and Stenseth 1985). Interspecific competition appears to promote differential resource use leading to niche divergence and a reduction in interspecific competition (Fig . l - l a , b ; l -2a,b) . Intraspecific competition broadens resource use and m a y indirectly increase both niche overlap and interspecific competition (Fig 1-lc, l-2c). These competitive effects were not eliminated by extraneous, non-competitive factors such as grazing, mowing, disturbance, or abiotic conditions. No evidence was found to suggest that coexistence resulted from coevolutionary processes leading to a balancing of competitive abilities, as suggeted by Aarssen (1983). Intra- and inter-specific competition are thought to have opposing effects on within-population variability, with intraspecific competition acting in a disruptive manner and interspecific competition in a stabilizing or directional manner (Pianka 1983). Species such as Trifolium repens which are usually more adversely affected by intra- versus inter-specific competition (Harper 1977, Chirwa 1985, Chapter 1) would be expected to show extensive within-population differentiation and an age-related increase in this variability. Trifolium repens is well known for examples of differentiation to abiotic (Snaydon 1962, Turkington and Harper 1979) a n d biotic conditions (Turkington and Harper 1979, Gliddon and Trathan 1985, Aarssen a n d Turkington 1985). The work represented in this thesis found extensive population variability, but no evidence of genetic differentiation to specific grass neighbours (Fig. 4-1). What "home-site" advantage was found may be due to phenotypic carry-over effects rather than genetic changes (Evans and Turkington 1988, Chapter 2, 4). The presence of biotic differentiation in Welsh pastures (Turkington and Harper 1979, Turkington 1989) and its absence in the British Columbian pastures may be due to differences in the intensity, constancj^ , and duration of biotic selective forces. The Welsh pasture is older and has relatively large areas (> 100m2) where a single grass species is dominant. In this pasture, transplants showed biotic differentiation and species-specific effects due to different grass neighbours. The British Columbian pastures, on the other hand, are younger, and composed of small (lm2), transient grass patches. Transplants placed into these patches show evidence of transient phenotypic differences, but not genetic differences. These transplants also fai l to show significant differences in growth or s u r v i v a l when planted into different grasses (Chapters 2,4). Increasing the constancy of patches in these pastures through art i f ic ial plantings (Chapter 1) did, however, demonstrate that T. repens responds significantly to different-aged populations of L. perenne. This fine-scale response suggests that T. repens m a y be affected in a species-specific manner in larger, more persisent na tura l patches. Age-related changes in growth and performance appear to result f rom heritable genetic events rather than transient phenotypic responses (Fig 4-2). Transplanted T. repens d id , however, show indications of carry-over effects due to previous environmental conditions. Clones derived f rom commercial seed displayed significantly better growth in L. perenne if they were propagated wi th this grass prior to re-planting into the same grass. Clones derived from older, established populations showed no response to this pre-treatment. These patterns could be due to a number of factors such as transient phenotypic carry-over effects or more permanent environmental induction such as gene amplif ication (Cullis 1987). The pr imary conclusion of this thesis is that populations, and hence communiti t ies , are unique in both space and time. This uniqueness is probably due to both chance events and selective factors. F r o m a selection standpoint, species appear to continual ly respond to biotic and abiotic forces (Lewontin 1968). A s physical conditions and biological neighbours change across the landscape, so do the selective pressures exerted on the inhabitants . This may result in locally (spatially) differentiated populations. Microevolution can occur rapidly , and can dramatica l ly affect competitive interactions in just a few years. These examples of rapid microevolution emphasize that populations and species are not fixed, but instead are f luid , dynamic entities unique to a specific point in t ime. Recognition of this uniqueness requires that caution be exercised in proposing universal theories which are dependent upon the performance of "static" species or "uniform" responses to the environment. Emphasis should instead be placed upon generalized models which address processes or concepts which do not rely upon such unrealistic constancy. Interactions between neighbouring plants will not be, and should not be expected to be, the most important force in all systems or even in all pastures. However, in systems where competition is important community development and evolution will depend upon the populations involved, the relative balance of inter-and intra-specific competitive relationships, the frequency of component taxa, and the presence of factors which disrupt or complement competitive relationships. LITERATURE CITED Aarssen, L.W. 1983. Ecological combining ability and competitive combining abil i ty in plants : towards a general evolutionary theory of coexistence in systems of competition. A m e r i c a n N a t u r a l i s t 122:707-731. Aarssen, L.W. 1984. O n the distinction between niche and competitive abi l i ty : implications for coexistence theory. A c t a Biotheoretica 33:67-83. Aarssen, L.W'. 1985. Interpretations of the evolutionary consequences of competition in plants: an experimental approach. Oikos 45:99-109. Aarssen, L.W. 1988. 'Pecking order' of four plant species from pastures of different ages. Oikos 51:3-12. Aarssen, L.W. and Turkington, R. 1985a. Vegetation dynamics and neighbour associations in pasture-community evolution. J o u r n a l of Ecology 73:585-603. Aarssen, L.W. and Turkington, R. 1985b. Biotic specialization between neighbouring genotypes in Lolium perenne in Trifolium repens from a permanent pasture. Journa l of Ecology 73:605-614. Aarssen, L.W. and Turkington, R. 1985c. Within-species diversity in n a t u r a l populations of Holcus lanatus, Lolium perenne, and Trifolium repens f rom four different-aged pastures. J o u r n a l of Ecology 73:869-886. Aarssen, L.W. and Turkington, R. 1985d. Competitive relations among species f rom pastures of different ages. C a n a d i a n J o u r n a l of Botany 63:2319-2325. Aarssen, L.W. and Turkington, R. 1987. Responses to defoliation in Holcus lanatus, Lolium perenne, a n d Trifolium repens f rom three different-aged pastures. C a n a d i a n J o u r n a l of Botany 65:1364-1370. Agren, G.I. and Fagerstrom, T. 1984. L i m i t i n g diss imi lar i ty in plants : randomness prevents exclusion of species with s imilar competitive abilities. Oikos 43:369-375. Akeroyd, J.R. and Briggs, D. 1983. Genecological studies in Rumex crispus L . I . G a r d e n experiments using transplanted mater ia l . N e w Phytologist 94:309-323. Alley, T.R. 1982. Competit ion theory, evolution, and the concept of an ecological niche. A c t a Biotheoretica 31:165-179. Antonovics, J., Bradshaw, A.J . , and Turner, R.G. 1971. H e a v y meta l tolerance in Plants . Advances in Ecological Research 7:1-85. Antonovics, J . 1976. The input from population genetics: "the new ecological genetics". Systematic Botany 1:233-245. Antonovics, J . 1978. The population genetics of mixtures . In P l a n t Relations in Pastures (Edited by J . R . Wilson) , pp. 233-252. C S I R O , Melbourne, A u s t r a l i a . Armstrong, R.A. 1988. Effects of disturbance patch size on species coexistence. J o u r n a l of Theoretical Biology 133:169-184. Arthur, W. 1982. The evolutionary consequences of interspecific competit ion. Advances in Ecological Research 12: 127-187. Austin, M.P., Groves, R.H., Fresco, L.M.F., and Kaye, P.E. 1985. Relat ive growth of six thistle species along a nutrient gradient w i t h multispecies competition. J o u r n a l of Ecology 73:667-684. Bailey, R . C . 1989. P lant Competi t ion. Nature 337:122. Bazzaz, F. 1987. Exper imenta l studies on the evolut ionary niche in successional plant populations. In Colonization, Succession, and Stabi l i ty . (Edited by J . A . G r a y . M . J . C r a w l e y , and P . J . Edwards) , pp. 245-271, B l a c k w e l l Scientific, London, U K . Bazzaz, F. and Sultan, S.E. 1987. Ecological var iat ion and the maintenance of diversity in plant populations. In Differentiation Pat terns in H i g h e r P l a n t s . (Edited by K . M . Urbanska) . Academic Press , London. Bendal, R.B., Higgins, S.S., Teberg, J.E., and Pyke, D . A . 1989. Compar ison of skewness coefficient, coefficient of var ia t ion , and G i n i coefficient as inequality measures wi th in populations. Oecologia 78:394-400. Benjamin, R.B. and Hardwick, R . C . 1985. Sources of var ia t ion and measures of variabi l i ty in even-aged stands of plants. A n n a l s of Botany 58:757-778. Berendse, F. 1983. Interspecific competition and niche differentiation between Plantago lanceolata and Anthoxanthum odoratum in a n a t u r a l hayf ie ld . Journal of Ecology 71:379-390. Berendse, F. 1985. The effect of grazing on the outcome of competition between plant species wi th different nutrient requirements. Oikos . 44:35-39. Bradshaw, A . D . 1965. Evolut ionary significance of phenotypic plast ic i ty in plants . Advances in Genetics 13:115-155. Bradshaw, A.D. 1972. The evolutionary consequences of being a plant . Evolut ionary Biology 51:25-47. Chanway, C P . , Holl, F.B., and Turkington, R. In press a. Effect of Rhizobium leguminosarum biovar trifolii on specificity between Trifolium repens and Lolium perenne . J o u r n a l of Ecology. Chanway, C P . , Holl, F.B., and Turkington, R. In press b. Rhizobium specificity of Bacillus polymyxa isolates wi th genotypes of Lolium perenne L . and Trifolium repens in a grass/legume pasture. J o u r n a l of Ecology. Charles, A .H. 1961. Dif ferent ia l s u r v i v a l of cult ivars of Lolium, Dactylis, and Phleum. Journal of the B r i t i s h Grass land Society 16:69-75. Chirwa, R . M . 1985. Inter- and intra-specific competition in perennial ryegrass-white clover mixtures . M . S c . thesis, U n i v e r s i t y College of W a l e s , A b e r y s t w y t h , U . K . Clay, K. and Levin, D.A. 1986. Environment-dependent intraspecific competition in Phlox drummondii. Ecology 67:37-45. Connell, J .H . 1975. Some mechanisms producing structure in n a t u r a l communit ies : a model and evidence from field experiments. In Ecology and Evolut ion of Communit ies . (Edited by M . C . Cody and J . M . Diamond) , pp. 460-490. Be lknap Press , Cambridge, Massachusetts , U S A . Connell, J .H. 1980. D i v e r s i t y and the coevolution of competitors, or the ghost of competition past. Oikos 35:131-138. Connell, J .H. In press. Apparent versus " r e a l " competition in plants . In Perspectives on Plant Competit ion. (Edited by J . Grace and D . T i lman) . Academic Press , N e w Y o r k , U S A . Connolly, J . 1986. O n difficulties wi th replacement-series methodology in mixture experiments . J o u r n a l of Appl ied Ecology 23:125-137. Connolly, J . 1987. W h a t is wrong with replacement series? Trends in Ecology and Evolut ion 3:24-26. Crawley, M.J . 1988. Herbivores and plant population dynamics . In P l a n t Populat ion Ecology. (Edited by A . J . D a v y , M . J . Hutchings , and A . R . Watkinson) , pp. 367-392. B lackwel l Scientific, London, U K . Cullis, C A . 1987. The generation of somatic and heritable var ia t ion in response to stress. A m e r i c a n N a t u r a l i s t 130:562-573. den Boer, P.J. 1986. The present status of the competitive exclusion principle. Trends in Ecology and Evolut ion 1:25-28. Diamond, J . and Case, T.J . 1986. Communi ty Ecology. H a r p e r and Row, N e w Y o r k . U S A . Donald, C M . 1963. Competit ion among crop and pasture plants. Advances in A g r o n o m y 15:1-118. Durrant, A. 1962. The environmental induction of heritable change i n Linum. Heredity 17:27-61. Durrant, A. 1971. Induction and the growth of flax genotrophs. Heredi ty 27:277-298. Ennos, R.A. 1983. Maintenance of genetic var iat ion in plant populations. Evolut ionary Biology 16:129-155. Evans, D.R., Hill, J., Williams, T.A., and Rhodes, I. 1985. Effects of coexistence on the performance of white clover - perennial ryegrass mixtures . Oecologia 66:536-539. Evans, G.M., Durrant, A., and Rees, H. 1966. Associated nuclear changes i n the induction of f lax genotrophs. Nature 212:687-699. Evans, R.C. and Turkington, R. 1988. Maintenance of morphological var ia t ion in a biotically patchy environment. N e w Phytologist 109:369-377. Firbank, L .G. and Watkinson, A.R. 1985. On the analysis of competition wi th in two-species mixtures of plants. J o u r n a l of Appl ied Ecology 22:503-517. Firbank, L.G. and Watkinson, A.R. 1989. P lant competit ion. N a t u r e 337:122. Fowler, N.L. 1986. Density-dependent population regulat ion in a Texas grass land. Ecology 67:545-554. Fox, D . J . and Guire, K . E . 1976. Documentat ion for M I D A S . T h i r d edition. Statistical Research Laboratory , U n i v e r s i t y of M i c h i g a n , A n n A r b o r , Michigan , U S A . Freund, R.J. and Littell, R . C . 1981. S A S for L i n e a r Models . S A S Institute C a r y , Nor th C a r o l i n a , U S A . Gaudet, C L . and Keddy, P . A . 1988. A comparat ive approach to predicting competitive abil i ty from plant traits. N a t u r e 334:242-243. Ghilarov, A . M . 1984. The paradox of the plankton reconsidered; or, w h y do species coexist. Oikos 43:46-52. Gliddon, C . and Trathan, P. 1985. Interactions between white clover and perennial ryegrass in an old permanent pasture. In Structure and Funct ioning of P l a n t Populations/2. (Edited by J . Haeck and J . W . Woldendorp), pp. 161-169, Nor th-Hol land Publ i sh ing , A m s t e r d a m , Nether lands . Goldberg, D . E . 1987. Neighborhood competition in an old-field plant communitv . Ecology 68:1211-1223. Goldberg, D . E . 1988. Response of Solidago canadensis clones to competition. Oecologia 77:357-364. Goldberg, D . E . In press. Components of resource competition in plant communities. In Perspectives on Plant Compet i t ion. (Edited by J . Grace and D . Ti lman) . Academic Press, N e w Y o r k , U S A . Goldberg, D . E . and Fleetwood, L. 1987. Comparison of competitive effects and responses among annual plants. Journal of Ecology 75:1131-1144. Goldberg, D . E . and Werner, P . A . 1983. Equivalence of competitors in plant communities: a nul l hypothesis and a field exper imental approach. A m e r i c a n Journal of Botany 70:1098-1104. Grime, J.P. 1977. Evidence for the existence of three p r i m a r y strategies in plants and its relevance to ecological and evolutionary theory. A m e r i c a n Natura l i s t 111:1169-1194. Grime, J.P. and Hunt, R. 1975. Relative growth rate: its range and adaptive significance in a local f lora. Journal of Ecology 63:393-422. Grubb, P. 1985. P l a n t populations and vegetation i n relation to habitat , disturbance, and competition: problems of generalization. In The Populat ion Structure of Vegetation. (Edited by J . White) , pp. 595-621. D r W . J u n k , Dordrecht, Nether lands . Hairston, N.G., Nishikawa, K . C , and Stenhouse, S.L. 1988. The evolution of competing species of terrestrial sa lamanders : niche part i t ioning or interference? Evolut ionary Ecology 1:247-262. Hancock, J.F. and Wilson, R.E. 1976. Biotype selection in Erigeron annus during old field succession. Bulletin of the Torrey Botanical Club 103:122-125. Harberd, D.J . 1963. Observations on natural clones of Trifolium repens. New Phytologist 62:198:204. Harper, J.L. 1969. The role of predation in vegetation diversity. In Diversity and Stability in Ecological Systems, Brookhaven Symposiums in Biology 22:48-62. Harper, J.L. 1977. Population Biologv of Plants. Academic Press, New York, USA. Hedrick, P.W., Ginevan, M.E., and Ewing, E.P. 1976. Genetic polymorphism in heterogeneous environments. Annual Review of Ecology and Systematics 7:1-32. Heslop-Harrison, J . 1964. Forty vears of genecologv. Advances in Ecological Research 2:1959-247. Hill, J . 1964. Environmental induction of heritable changes in Nicotiana rustica. Nature 207:732-734. Holt, R.D. 1977. Predation, apparent competition, and the structure of prey communities. Theoretical Population Biology 12:197-229. Hubbell, S.P. and Foster, R.B. 1986. Biology, chance, and history and the structure of tropical rain forest tree communities. In Community Ecology. (Edited bv J. Diamond and T.J. Case), pp. 314-329. Harper and Row, New York, USA. Hullu, E., de. 1985. Phenotypic variation of Rhinauthus augustifolius in a successional series. In Structure and Functioning of Plant Populations/2. (Edited by J. Haeck and J.W. Woldendorp), pp 115-125, North-Holland Publishing, Amsterdam. Netherlands. Inouye, R.S. and Schaffer, W.M. 1981. On the ecological meaning of ratio (de Wit) diagrams in plant ecology. Ecology 62:1679-1681. Jolliffe, P.A., Minjas, A.N. , and Runeckles, V.C. 1984. A reinterpretation of yield relationships in replacement series experiments. Journal of Applied Ecology 21:227-243. Kelley, S.E. and Clay, K. 1987. Interspecific competitive interactions and the maintenance of genotypic variation within two perennial grasses. Evolution 41:92-103. Langlet, O. 1971. Two hundred years genecology. Taxon 20:653-722. Lawlor, L.R. and Maynard-Smith, J . 1976. The coevolution and stability of competing species. American Naturalist 110:79-99. Lewontin, R.C. 1968. Population Biology and Evolution. Syracuse University Press, Syracuse, New York, USA. Libby, W.J. and Jund, E . 1962. Var iance associated w i t h cloning. Heredi ty 17:535-540. Loehle, C . 1988. Problems with the rectangular model for representing plant strategies. Ecology 69:284-286. Lundberg, S. and Stenseth, N . C . 1985. Coevolution of competing species: ecological character displacement. Theoretical Populat ion Biology 27:105-119. Mac-Arthur, R. and Levins, R. 1967. The l imi t ing s i m i l a r i t y , convergence, and divergence of coexisting species. A m e r i c a n N a t u r a l i s t 101:377-385. McNaughton, S .J . , Folsom, T .C . , Lee, T . , Park, F., Price, C , Roeder, D., Schmitz, J., and Stockwell, C. 1974. H e a v y meta l tolerance in Typha latifolia without the evolution of tolerant races. Ecology 55:1163-1165. McNeilly, T . and Roose, M . L . 1984. The distr ibution of perennial ryegrass genotypes in swards . N e w Phytologist 98:503-513. Miller, T . E . and Werner, P.A. 1987. Competit ive effects and responses between plant species in a f irst-year old-field community . Ecology 68:1201-1210. Mitchley, J . 1987. Diffuse competition in plant communit ies . Trends in Ecology and Evolut ion 2:104-106. M i t c h l e y , J . and Grubb, P.J. 1986. Control of relative abundance of perennials in chalk grasslands in southern E n g l a n d . I. Constancy of rank order and results of pot- and field-experiments on the role of interference. J o u r n a l of Ecology 74:1139-1166. Moen, J . 1989. Dif fuse competition - a diffuse concept. Oikos 54:260-263. O'Connor, I. and Aarssen, L.W. 1988. Species association patterns in abandoned sand quarries. Vegetatio 73:101-109. Pacala, S. 1988. Competit ive equivalence: the coevolutionary consequences of sedentary habit. A m e r i c a n N a t u r a l i s t 132:576-593. Paine, R .T. 1966. Food web complexity and species d ivers i ty . A m e r i c a n N a t u r a l i s t 100:65-75. Parrish, J . A . D . and Bazzaz, F.A. 1982. Competit ive interactions in plant communities of different successional ages. Ecology 63:314-320. Peterson, R. 1975. The paradox of the plankton: an equi l ibr ium hypothesis. A m e r i c a n N a t u r a l i s t 109:35-49. Pianka, E .R. 1983. Evolut ionary Ecology. T h i r d edition. H a r p e r and Row, N e w Y o r k , U S A . Pickett, S .T .A. 1980. Non-equi l ibr ium coexistence in plants . Bul le t in of the Torrey Botanical Club 107:238-248. Rice, K.J . and Menke, J.W. 1985. Competi t ive reversals and environment-dependent resource part i t ioning in Erodium. Oecologia 67:430-434. Rouse, M.L. and Radosevitch, S.R. 1985. Relationships between growth and competitiveness for four annual weeds. J o u r n a l of A p p l i e d Ecology 22:895-905. Roy, J . 1985. Comparisons of Dactylis glomerata and Bromus erectus populations from contrasted successional stages. In Structure and Funct ioning of P lant Populations/2. (Edited by J . Haeck and J . W . Woldendorp), pp. 51-64, N o r t h - H o l l a n d Publ ishing, A m s t e r d a m , Netherlands. Sarukhan, J . and Harper, J .L. 1971. Studies on plant demography: Ranunculus repens L . , R. bulbosus L . , and R. acris L . I. Population f lux and survivorship . J o u r n a l of Ecology 61:675-716. SAS. 1985. S A S U s e r ' s Guide: Statistics. S A S Institute, C a r y , N o r t h C a r o l i n a , U S A . Schaal, B.A. 1984. L i fe history variat ion, natura l selection, and maternal effects in plant populations. In Perspectives on Plant Populat ion Ecology. (Edited by R. Dirzo and J . Sarukhan) , pp. 188-206. S inauer Associates , Sunderland, Massachuset ts , U S A . Schaal, B.A. 1988. Somatic variat ion and genetic structure in plant populations. In P l a n t Populat ion Ecology. (Edited by A . J . D a v y , M . J . Hutchings , and A . R . Watkinson) , pp. 47-58. Blackwel l Scientific, L o n d o n , U K . Schmid, B. 1985. C lonal growth in grassland perennials . III. Genetic var iat ion and plasticity between and within populations of Bellis perennis and Prunella vulgaris. J o u r n a l of Ecology 73:819-830. Seaton, A.P.C. and Antonovics, J . 1967. Population interrelat ions. I. Evolut ion i n mixtures of Drosophila mutants . Heredi ty 22:19-33. Silvertown, J . 1989. P l a n t competition. Nature 337:123. Silvertown, J . and Law, R. 1987. Do plants need niches? Some recent developments in plant community ecology. Trends i n Ecology and Evolut ion 2:24-26. Slatkin, M. 1980. Ecological character displacement. Ecology 61:163-177. Snaydon, R.W. 1962. The growth and competitive abil i ty of contrasting populations of Trifolium repens L . on calcareous and acid soils. J o u r n a l of Ecology 50:439-447. Snaydon, R.W. 1978. Genetic changes in pasture populations. In P l a n t Relations in Pastures . (Edited by J .R . Wilson), pp. 253-269. C S I R O , Melbourne, A u s t r a l i a . Snaydon, R.W. 1985. Aspects of the ecological genetics of pasture species. In Structure and Funct ioning of Plant Populations/2. (Edited by J . Haeck and J . Woldendorp), pp. 131-152. Nor th-Hol land , A m s t e r d a m , Nether lands . Snaydon, R.W. and Davies, M.S. 1976. Rapid population differentiation i n a mosaic environment. I V . Populations of Anthoxanthum odoratum at sharp boundaries. Heredi ty 37:9-25. 8 8 Snaydon, R.W. and Davies, M.S. 1982. Rapid divergence of plant populations in response to recent changes in soil conditions. Evolut ion 36:289-297. Sultan, S.E. 1987. Evolut ionary implications of phenotypic plasticity in plants. Evolut ionary Biology 21:127-178. Thompson, J .N. 1988. V a r i a t i o n in interspecific interactions. A n n u a l Review of Ecology and Systematics 19:65-87. Tilman, D. 1977. Resource competition between planktonic algae: an experimental and theoretical approach. Ecology 58:338-348. Tilman, D. 1981. Tests of resource competition theory us ing four species of L a k e M i c h i g a n algae. Ecology 62:802-815. Tilman, D. 1988. P l a n t strategies and the dynamics and structure of plant communities. Princeton U n i v e r s i t y Press , Pr inceton, N e w Jersey , U S A . Turkington, R. 1979. Neighbour relations in grass-legume communities . I V . Fine scale biotic differentiation. C a n a d i a n J o u r n a l of Botany 57:2711-2716. Turkington, R. 1989. The growth, distr ibution, and neighbour relationships of Trifolium repens in a permanent pasture. V . The coevolution of competitors. J o u r n a l of Ecology (in press). Turkington, R. and Aarssen, L .W. 1984. Local-scale differentiation as a result of competitive interactions. In Perspectives on P l a n t Population Ecology. (Edited by R. Dirzo and J . Sarukhan) , pp. 107-127. Sinauer Associates, Sunderland, Massachusetts , U S A . Turkington, R. and Harper, J . L . 1979. The growth, dis tr ibut ion, and neighbour relationships of Trifolium repens in a permanent pasture. I V . Fine-scale biotic differentiation. J o u r n a l of Ecology 67:245-254. Turkington, R., Holl, F.B., Chanway, C P . , and Thompson, J.D. 1988. The influence of microorganisms, par t i cular ly Rhizobium, on plant competition in grass-legume communities. In P l a n t Populat ion Ecology. (Edited by A . J . D a v y , M . J . Hutchings , and A . R . Watkinson) , pp. 343-366. B lackwel l Scientific Publications, London , U K . Turkington, R. and Mehrhoff, L . A . In press. The role of competition in s tructur ing pasture communities. In Perspectives on P l a n t Competition (Edited by J . Grace and D . T i lman) . Academic Press , N e w Y o r k , U S A . Underwood, T. 1986. The analysis of competition by field experiments. In C o m m u n i t y Ecology: Pat tern and Process. (Edited by J . K i k k a w a and D . J . Anderson), pp. 240-268. B lackwel l Scientific, London, U K . Venable, D . L . 1984. U s i n g intraspecific var ia t ion to study the ecological significance and evolution of plant life-histories. In Perspectives on Plant Populat ion Ecology. (Edited by R. Dirzo and J . Sarukhan) , pp. 166-187. Sinauer Associates, Sunderland, Massachusetts , U S A . Watson, P. 1969. Evolut ion in closely adjacent plant populations. V I . A n entomophilous species, Potentilla erecta, in two contrast ing habitats. Heredity 24:407-422. Warwick, S.I. and Briggs, D. 1979. The genecology of weeds. III. C u l t i v a t i o n experiments w i t h Achillea millefolium L . , Bellis perennis L . , Plantago lanceolata L . , Plantago major L . , and Prunella vulgaris L . collected f rom lawns and contrasting grassland habitats. N e w Phytologist 83:509-536. Weiner, J . 1985. Size hierarchies in experimental populations of annual plants . Ecology 66:743-752. Weiner, J . and Thomas, S .C . 1986. Size var iab i l i ty and competition in plant monocultures. Oikos 47:211-222. Welbank, P.J. 1963. A comparison of competitive effects of some common weed species. A n n a l s of Applied Biology 51:107-125. Wiens, J . A . 1977. O n competition and variable environments . A m e r i c a n Scientist 65:590-597. Wilken, D.H. 1977. Local differentiation for phenotypic plast ici ty in the annual Collomia linearis (Polemoniaceae). Systematic B o t a n y 2:99-108. Wilson, J.B. 1988. The effect of init ial advantage on the course of plant competition. Oikos 51:19-24. Wit, C . T . , de, and van den Berg, J.P. 1965. Competit ion between herbage plants. Netherlands Journal of A g r i c u l t u r a l Research 13:212-221. FIGURE A - l . Net ramet production of different-aged populations. The average ramet production (NRP) of al l clones of a species when placed into different treatments (T): monoculture ( • ) and in mixtures w i t h competitors collected from young ( 0 ), intermediate ( [g] ), and old ( • ) pastures. The performance of each different-aged populations is shown when in monoculture (M) and when in mixture against an averaged competitor ( M X ) . W i t h i n a group, bars which share the same letter are not significantly different (P > .05, t-test). 91 5-0-\ 1 1 1 1 1 0 10 20 30 40 50 Population age (y) FIGURE A-2. Productivity of mixtures. Total yield (net ramet production, NRP) of natural paired mixtures. The bold line is total NRP mixture yield, dashed line is NRP of L. perenne, and solid line is NRP of the other species. Lines with arrows indicate significant changes from the youngest to the oldest population (P < .05, t-tests). 0_| ( , ! ! , 0 10 20 30 40 50 P o p u l a t i o n a g e (y) FIGURE A-3. Equivalence of competitive relationships. Yield suppression (YS) and yield suppression ratios (YSR) of mixtures. The bold line is YSR, an estimate of competitive equivalence of the two speices in mixture, the dashed line is YS of L. perenne, and the solid line is YS of the competing species. An increase in YSR indicates that the two species are becoming competitively equivalent. An increase in YS indicates that interspecific competition is becoming less important to that species. Lines with filled arrows indicate significant changes from the youngest to the oldest population (P < .05, t-tests and regression). 9 3 H. lanatus (— L. perenne ( ) 1 I I 1 1 1 0 10 20 30 40 50 L D . glomerata ( — ^ L . perenne ( ) H I I 1 1 1 0 10 20 30 40 50 T. repens (— L. perenne ( ) n i i 1 1 1 0 10 20 30 40 50 Population age (y) FIGURE A-4. Estimates of niche separation. Relative yield (RY) and relative yield total (RYT) of paired mixtures. The bold line is RYT, an estimate of niche separation between the two species in mixture, the dashed line is RY of L. perenne, and the solid line is RY of the competing species. An increase in RYT indicates an increase in niche separation. An increase in RY indicates that the species is becoming less effected by interspecific competition. Lines with arrows indicate significant changes from the youngest to the oldest population (P < .05, t-tests and regression). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0098261/manifest

Comment

Related Items