Patric Nilsson

INDUCTION OF OVERCOMPENSATION IN THE FIELD GENTIAN, GENTIANELLA CAMPESTRIS

We present field evidence for the induction of overcompensation, or increased fruit and seed yield as a consequence of damage, in the grassland biennial field gentian, Gentianella campestris (Gentianaceae). We compared equally sized clipped and unclipped plants in two populations in central Sweden during three years, 1992-1994, and plants clipped at different occasions, from 20 June to 2 August. Clipping once, by removing half of the biomass, significantly increased fruit production without affecting the number of seeds per fruit or seed mass. The degree of compensation was sensitive to the timing of clipping. Damage induced overcompensation only during a restricted inductive time period (ITP) in July. Plants clipped before about 1 July or after about 22 July achieved no overcompensation. The early limit of ITP was presumably determined by the availability of resources that could be mobilized for regrowth after damage. The late limit, on the other hand, depended primarily on the differentiation of meristems close to flowering in early August. The effects of clipping varied between years, presumably due to drought in 1994. During 1992-1993, plants consistently overcompensated for clipping on 1-20 July, whereas in 1994 only early clipping from 1 to 12 July induced overcompensation. In 1994, plants clipped in late July compensated less well, due to delayed fruit maturation leading to a high proportion of immature fruits at the end of the season. Because of this between-year variation, we used geometric mean fitness to calculate the expected long-term effects of damage over generations. The analysis suggests that the long-term effects can vary from positive to negative, depending on the frequency of bad fruiting years. The time limits of ITP fit well the hypothesis that predictable damage in July may have selected for a capacity of overcompensation in the field gentian. Because the ultimate limits of ITP are set by the length of the vegetation period, we expect overcompensation in this species to be more common in regions with a longer growing season.

EVIDENCE FOR AN EVOLUTIONARY HISTORY OF OVERCOMPENSATION IN THE GRASSLAND BIENNIAL GENTIANELLA CAMPESTRIS (GENTIANACEAE)

Swedish University of Agricultural Sciences, Department of Conservation Biology, Section of Conservation Botany, Box 7072, 750 07 Uppsala, Sweden; University of Oulu, Department of Biology, Linnanmaa, 90 570 Oulu, Finland, and University of Lund, Department for Theoretical Ecology, Ecology Building, 223 62 Lund, Sweden; University of Lund, Department for Theoretical Ecology, Ecology Building, 223 62 Lund, Sweden

Plant Compensatory Responses: Bud Dormancy as an Adaptation to Herbivory

Some plants can compensate, and even overcompensate, for the loss of productivity caused by herbivory. The presence of latent meristems, or dormant buds, is one of the basic prerequisites of such compensation mechanisms. We present a mathematical model in order to analyze compensation responses in relation to the intensity of herbivory. The model generates a number of qualitatively different kinds of compensation curves when seed production is plotted against the proportion of active meristems lost per grazed plant. The shape of the curves depends on the proportion of dormant buds and their activation sensitivity in relation to meristem loss. Overcompensation is most probable when dormant buds are easily activated. When plants are grazed only once, as assumed in our model, selection favors high bud sensitivity. However, we expect that repeated damage may select for a more restrained pattern of bud activation. When relatively few buds remain dormant, plants can overcompensate for low levels of damage only. On the other hand, when most buds remain dormant, they can overcompensate even for high levels of damage. We consider compensation capacity a potential benefit of bud dormancy when plants are subject to damage. However, bud dormancy may also imply costs on plant productivity and fecundity in the absence of herbivory. Still, intense herbivory may favor bud dormancy in spite of the potential costs. Selection for bud dormancy requires both that the risk of herbivory is high and that herbivores remove a large fraction of active meristems per plant. Consequently, overcompensation is a theoretically plausible possibility, and intense herbivory is a potential selective force that favors bud dormancy. None of these results, however, imply that herbivory is beneficial to plants. In our case, plants with bud dormancy never have higher seed production than plants that have no dormant buds and that are not grazed.

BUD DORMANCY AS A BET-HEDGING STRATEGY

The presence of dormant buds allows grazed plants to compensate for destroyed active meristems. In this article we present a theoretical analysis of the adaptive significance of bud dormancy when the risk of herbivory varies from year to year. Under constant herbivore pressure, selection tends to favor either plants that have no dormant buds and hence no capacity of compensatory growth due to low risk of herbivory, or those that leave most of their buds in dormancy because of high risk of herbivory. We show that when the risk of herbivory varies from year to year, selection will favor intermediate phenotypes having both dormant and active meristems. The intermediate phenotypes represent bet-hedging strategies that have lower variance of seed production than either of the extreme strategies. We also show that intensive herbivory can favor meristem allocation strategies that allow plants to overcompensate for herbivore damage. Our results suggest that two kinds of herbivores might cause selection pressures favoring compensatory growth: large ungulate herbivores that remove a large proportion of the aboveground parts of the attacked plants and that predictably attack fairly many plants (>50%) each year, and invertebrate herbivores that periodically cause extensive damage during years of mass attacks.

Plant compensatory growth : herbivory or competition?

The presence of an overcompensatory response to damage in some plant species has recently created a debate concerning whether this trait is an adaptation to herbivory, or simply a physiological consequence of adaptations to competition for light. According to the latter hypothesis, competition for light favors fast vertical growth and strong apical dominance. The removal of apical dominance by damaging the primary shoot allows the growth of secondary shoots and hence increases productivity. We compare predictions of these two hypotheses in a model-system where plants are exposed to both a risk of damage and a risk of competition. Compensatory seed production is assumed to depend on the number of dormant buds that can be activated by damage, and on the seed production of surviving shoots. In accordance with earlier theoretical analyses, we expect that intensive herbivory can favor overcompensatory seed production. In contrast, competition for light should at best lead to exact compensation when the competitive environment remains unchanged. Competition acts against overcompensation for two reasons. First, competitive plants should have poor resource reserves to support compensatory growth. Second, competition for light is assumed to favor unbranched architecture and thus. activation of many secondary shoots should not increase the seed yield. However, we cannot exclude the possibility that plants adapted to competition may overcompensate when grown singly. In spite of this caveat, it is likely that overcompensation requires damage related adaptations that may evolve only under intensive and relatively predictable risk of damage.

Even repeated grazing may select for overcompensation

0 0 0 EWif0 0 E 0 00 00 g 0 0 00 00 0 0 0 00 00 0 0 CELEELEiLE iLLEREEDiEREDRELREDEEDiERELD!iDiSiESlliELSiSS0iE i! : i... .. ......... . . .. . ....... ... .. ... . . . ..:ESESiEiiEdiEiEf 00 E 0 :: t0j04:40 i:EA:4:40 WA040 WAW 0 0 iA: iAg E 0 0 E i :0 WAg 0 0 0 g 0 0 itE it! 0 itE E ititiSESESEtESE ititi;EtELitESELESESititiSESi !titStiSESESit0000000000000050402040004040 i004000 g 0 i 0 g E 0 0 i i 0 i i 0 0 0 0 0 iSigE i..... . .. .. ...... . . . . . ... ......... . iEiEEEd: ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ .... .....ESEEDd ......... .E ..:LE ...00 ..ddEEEEEELE0:yE . ...... . .. ........|l:g:S0 ........ EEEEEEESESEE::iiRSNLEL!f::ECid00 . .EEiEi ..SS i E EE E! L 20 00t h CESifiSi:SiLiiSiSSif t Li0|:EEdELLiEELLiS:EEEESEtE ....: ......... .. .. . .E EEEX 40 *0LLiStgig :fi L:EE i k i : i: :E ESE iEE i S S4c ; g0 g g 0 w-EiS sESE SES ii i;i; g iES EEEEXi0iEE iddES g F iE ....... .. ..... SSESEELSiLSEE fiS;00000000000000000000000000000004 00S00SAit0ititi:f:E:it):iSE1i:CCEdEL t L400 :SSEEE ........ .. . .: ....... .. ..... ......... ..... ... ... .... .... . .. . .... .. ........ ........ ...... 404000000000000000000000000000000000000000::SitirESELiS tit::00 ......... ... .: . .. . ........ ....0000000004020 iE :iiE 0 0g;i 00 00 E i! 00 0i E0 ; 0| gE ii E;LLiX E0 SEERDiESiEEDSliSREiLiiXiiEXiSiER ;f4E i4fE i . ........ . ..... .... . .. . ... . ...EXiE SE Ed Ei ii PEi} giE iED ti04EEiXf :::.....:.:...: ......:: :: ........::::::.::.:.:::..:: ::::::::7:::::::::: :::::::::::::::: . .

Does seed dormancy benefit the mother plant by reducing sib competition

SummarySeed dormancy has been considered, almost without exception, as a bet-hedging strategy in a temporally varying environment. However, in this paper we show that seed dormancy can improve the reproductive success of the mother plant when competition between sibling seedlings and adult plants is intense even if the environment is temporally invariable. We allow a cohort of sibling seeds to germinate simultaneously in the same patch and assume a density dependent survival and fecundity of seedlings. In the model, the mother plant is assumed to control the germination behaviour of the seeds, e.g. by enclosing the seeds in coats of different hardiness. When sib competition is intense, a postponed germination of her seeds can increase the reproductive success of the mother plant up to four times, measured in terms of the number of grandchildren. Consequently, our results suggest that postponed germination may function as a mechanism that alters local interactions in viscous plant populations with limited dispersal.

Herbivory, inducible defence and population oscillations : a preliminary theoretical analysis

The regular cyclic population dynamics of herbivores has frustrated naturalists since the middle of the 16th century. Olaus Magnus, archbishop of Uppsala, Sweden, concluded already 1555 that the lemming populations showed oscillatory population changes (Olaus Magnus 1555) Furthermore, he suggested that the cyclic outbreaks of stoats and weasels could be caused by the 3-4-yr lemming cycle. Since then, several ecologists have speculated about the causes of these spectacular phenomena. For cyclic microtine populations, the hypotheses presented, may, according to Stenseth and Ims (1993) be divided into four main categories: (1) abiotic hypotheses (e.g. Elton 1924, Moran 1953a, b), (2) intrinsic factor hypotheses (e.g. Chitty 1967, Charnov and Finerty 1980), (3) prey-predator/pathogen interactions (e.g. Stenseth 1980, Ydenberg 1987, Anderson and May 1991), and (4) vegetation-herbivore interactions (e.g. Schultz 1969, Haukioja and Hakala 1975, Haukioja 1980). Obviously, this is a general classification that could be valid for all kinds of cyclic herbivore populations. Among the vegetation-herbivore hypotheses, we find the ones concerning food quality especially interesting. Green and Ryan (1972) suggested that plants could respond to herbivory by the production of proteinase inhibitors with adverse effects on herbivore performance (growth, survival and reproduction). These defence responses may, according to Haukioja and Hakala (1975), be a driving force of fluctuating herbivore populations. In a recent contribution, Seldal et al. (1994) propose that an inducible defence mechanism can be the cause of the population cycles exhibited by the lemming in the tundras of America, Fennoscandia and Russia. Their main findings are: (1) Physical damage to those plants that constitute the bulk of the lemmings diet will induce the synthesis of large amounts of proteinase inhibitors, proteins that form irreversible complexes with trypsin and other proteolytic enzymes. This does not only effectively inhibit the protein digestion in the gut of most mammals, birds and insects, but also drain their reserves of essential amino acids that are excreted into the faeces (reviewed by e.g. Gallaher and Schneeman 1986, Ryan 1990). (2) Experiments have shown that proteinase inhibitors, when introduced with artificial food, have adverse effects on the growth and can eventually cause the death of mammals, birds and insects (reviewed by e.g. Gallaher and Schneeman 1986, Ryan 1990). Similar effects could be observed when young lemmings were given food collected in their natural habitat, if the samples were taken during the decline and trough phases of the lemming population cycle. The phytochemical induction by herbivory has been studied extensively by other workers. In the case of proteinase inhibitors, the general conclusion that can be drawn from these studies is that a plant can exhibit different levels of induction, depending on the size and number of damages (Green and Ryan 1972). A single damage event will trigger the production of inhibitors, but without further damages, the synthesis will cease and the concentration of these substances will decline in the plant tissue (Seldal et al. 1994). Repeated damages do, however, cause high levels of proteinase inhibitors that will last for extended periods (Green and Ryan 1972, Gustafson and Ryan 1976). Seldal et al. (1994) argue that this effect might cause a delay in the lemming population regulation, and that this could be sufficient for causing cyclic population dynamics. The more a plant has been damaged through grazing, the higher will its defence level be, and the longer will it remain in this induced state. In this note we propose a fairly simple mathematical model that encapsulate these properties, in order to investigate the hypothesis that inducible defences can cause oscillations in herbivore populations.

Plant Adaptations to Herbivory: Mutualistic versus Antagonistic Coevolution

The discovery of overcompensatory responses to damage in some plant species has inspired attempts to classify some plant-herbivore interactions as mutualism. Since the debates over plant-herbivore mutualism and overcompensation have been intense, we attempt to outline three conceptual models of plant-animal interactions and evaluate the status of interactions with help of three different fitness criteria: relative fitness, absolute fitness, and mean absolute fitness. Our three plant-animal interactions are assumed to represent plant-pollinator mutualism, plant-herbivore antagonism and the evolution of overcompensation, respectively. Each case describes how absolute fitness, and consequently also the other two fitness criteria, is assumed to change with animal encounters for plants with special adaptations to cope with those encounters and for plants with no such adaptations. As a result, all these types of interactions may be considered as mutualism when taken relative fitness only into account. Obviously, this criterion is too weak because any trait, evolving under natural selection, should improve fitness relative to other, alternative traits. Absolute fitness increases with animal encounters for the adapted phenotype in the first and in the last case and thus, in the context of absolute fitness, overcompensation in plants would indicate plant-herbivore mutualism. However, absolute fitness as such may not be sufficient when discussing the evolutionary history of plant-animal mutualism. In the light of mean fitness, overcompensation as presented in earlier studies does not represent mutualism between plants and herbivores. Mean absolute fitness in plant populations decrease with the risk of herbivore attack in our model of overcompensation, while the reverse trend characterises our plant-pollinator model. We, therefore, suggest that mean absolute fitness may well provide an appropriate criterion for distinguishing mutualistic and antagonistic plant-animal interactions in coevolutionary contexts. In order to evaluate our model-system, we compare the predicted patterns with empirical data on the grassland biennial Gentianella campestris.

Parent-offspring and sexual conflicts in the evolution of angiosperm seeds

In angiosperm plants the parental investment consists mainly of the endosperm, a nutritive tissue formed by genetic elements from both parents and used upon germination. We analyze a one-locus model of the evolution of the endosperm assuming that the alleles expressed in the endosperm determine the seed provisioning. In the model, we assume that large endosperm increases the probability of survival when young, but decreases seed set when the plant has reached the reproductive stage. We show that there is an evolutionarily stable strategy (ESS) of endosperm amount which is influenced by mating system, ploidy of the endosperm, paternity and genomically imprinted genes with parent-specific expression. The ESS value is higher than the value which maximizes the reproductive success of the plant. When the maternal genetic contribution is higher than the paternal (e.g. Polygonum plants), the ESS is almost always lower than when the parents have equal influence over the endosperm as in Oenothera plants. In both types, increasing the number of pollen donors to a seed crop selects for higher levels of endosperm. Accordingly, genes expressed only when inherited from the father are selected for higher endosperm amounts than genes expressed only when inherited from the mother, except when all seed have the same pollen parent. ESS values for imprinted genes do not differ between Polygonum and Oenothera types of plants. The ESS values are shown to be both locally and globally stable. The results are discussed in relation to evolutionary conflicts between the sexes.