
Evolutionary Ecology of Plant Reproductive Strategies
340
Evolutionary Ecology of Plant Reproductive Strategies
340Paperback
-
SHIP THIS ITEMIn stock. Ships in 1-2 days.PICK UP IN STORE
Your local store may have stock of this item.
Available within 2 business hours
Related collections and offers
Overview
Product Details
ISBN-13: | 9780521528948 |
---|---|
Publisher: | Cambridge University Press |
Publication date: | 10/13/2005 |
Pages: | 340 |
Product dimensions: | 6.89(w) x 9.61(h) x 0.71(d) |
About the Author
Peter Klinkhamer is a Professor in the Institute of Biology, Leiden University. His interests include the evolution of monocarpy, optimal defense against herbivory and selective embryo abortion.
Read an Excerpt
Cambridge University Press
0521821428 - Evolutionary Ecology of Plant Reproductive Strategies - by Tom J. de Jong and Peter G. L. Klinkhamer
Excerpt
1
Optimization models
About thirty years ago there was much talk that geologists ought only to observe and not theorise; and I well remember someone saying that at this rate a man might as well go into a gravel pit and count the pebbles and describe the colours. How odd is it that anyone should not see that all observation must be for or against some view if it is to be of any service.
Charles Darwin, letter to Henry Fawcett, 18 September 1861
1.1 Introduction
The common garden near my house was initially bare soil, flattened by bulldozers. It was planted with fast-growing plants such as periwinkle (Vinca minor) and ivy (Hedera helix) with patches of bamboo. After four years, the ground covering plants had won their battle with the dandelions (Taraxacum officinale) and the bamboos were man-sized, providing places with shade and privacy. It was delightful. In the fifth year, however, disaster struck. One of the bamboos started to produce flowering shoots and, later in the year, other bamboos did the same. By the autumn all the bamboos were completely dead. In the neighbouring garden the bamboos had also flowered and, looking around, you could see that bamboos all over town had flowered, irrespective of the age of the garden or the date they were planted. It took the local garden centre over a year to catch up with the sudden demand for vegetative bamboos, and by that time our garden was replanted with other species. What is the interesting biological question here? One group of biologists will ask a causal question. What causes the synchronous flowering of bamboos? Do all plants respond to a common environmental cue, such as a dry summer or a very cold winter? Are the plants perhaps communicating with each other? Are they releasing volatiles, such as ethylene, that induce others to flower? These are all relevant questions about the underlying mechanism. Yet another group of biologists asks an evolutionary question. Why do the bamboos have synchronized flowering? Starting from ancient unsynchronized bamboo species, what were the selective factors that favoured the synchronous flowering of all plants? Why isn't flowering synchronized in all plant species and what makes these bamboos so special? The motto of this second group of biologists was stated eloquently by the geneticist Dobzhansky: 'Nothing in biology makes sense, except in the light of evolution.'
Dan Janzen is a bamboo expert with a great interest in evolution. In 1976 he wrote a fascinating article 'Why bamboos wait so long to flower'. From historical records, dating back to the year 919, he concluded that the mainland Chinese bamboo Phyllostachys bambusoides flowered once every 120 years, and that flowering of individual plants was synchronized over large areas. Bamboo seeds are at least the size of a rice kernel and have as high a nutritional value. In years of synchronized flowering (so-called mast years) it was not uncommon for bamboo seeds to cover the soil as a thick blanket. Rats and rodents were identified as prime seed-eaters. After masting of bamboo the fertility of rats increased and large litters were produced. However, when the seeding of bamboo had ended and seeds had germinated or disappeared into the soil, a lean period started for the rats and they needed a different food source. Janzen reported that rats destroyed Asian crops after which they died in great numbers, causing famine and infectious diseases such as the plague, which was the reason the mass flowering of bamboos was recorded in the history books in the first place. Janzen hypothesized that mass flowering of bamboos had evolved as a plant strategy to escape from seed predation. In years of flowering, seed production is so abundant that the rats cannot eat everything, even though their population increases rapidly. Some seeds escape and assure survival of the next generation of bamboo plants. Seeds of a bamboo genotype that is out of synchrony and flowers one year too early may all be eaten by the few rats that are around. Seeds of a bamboo genotype that flowers one year too late are certainly eaten by many, hungry rats. Once mast blooming has evolved, there is strong selection for plants to stay in synchrony.
Of course plants do not have a nervous system and certainly do not make conscious decisions about what to do next. Nevertheless, in the case of Phyllostachys bambusoides one cannot help thinking that the plant has outsmarted its seed predators. In the past, different bamboo genotypes have played a game against each other. The winner was the genotype with the rather bizarre strategy of flowering once every 120 years, in total synchrony with the others, and devoting all its resources to reproduction. Plant strategies are the subject of this book. Life-history strategies encompass characters such as generation time, the fraction of resources spent on reproduction, and the resources allocated per offspring (seed). Complementary to these are sex allocation strategies which include the allocation to male and female function, attraction of pollinators and so on. The focus of this book will be on general models and data collection that allow testing of hypotheses concerning these strategies. Before going from natural history to the mathematics of these models it is useful to examine the arguments of people who have been sceptical about testing evolutionary hypotheses.
1.2 The spandrels of San Marco or the spaniels of St Marx?
For any student with an interest in the history of science, Gould and Lewontin's (1979) article is a must. The ingenious title is: 'The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme'. This title certainly needs some explanation. Spandrels are the curved, triangular areas between adjoining arches. Those in St Mark's cathedral in Venice are embellished with mosaics, each depicting one of the four evangelists. When you study these spandrels in isolation, you might come up with the silly idea that they are the reason for the system of arches and domes that surrounds them. Of course this is not true, they are just a logical consequence of the way churches were built in those days. Spandrels were necessary for holding up the dome of the church, they are a by-product of the construction. To give a simpler example; the function of the nose is in smelling and breathing. Someone might come up with the idea that the nose evolved to hold up spectacles, but of course this is nonsense. This function of noses is just an accidental by-product. The Panglossian paradigm is the other element of Gould and Lewontin's (1979) title. A paradigm is an example. Dr Pangloss was a character from Voltaire's book Candide, which was written in a satiric style. The doctor's views are perhaps best illustrated by a quote: 'It is demonstrable that things cannot be otherwise than as they are; for as all thing have been created for some end, they must necessarily be created for the best end.' The Gould and Lewontin paper is not about architecture or literature though. It is a critique of evolutionary studies without a sound methodological basis. First, Gould and Lewontin used the spandrels metaphor to argue that characters should not be studied in isolation but as a whole. Second, they used Dr Pangloss to ridicule the biologist who can always dream up an adaptive explanation of some character afterwards, not questioning his basic premise that each character in each organism is fully adapted to its environment. 'Adaptive storytelling' is a term also used in this context. Basically, Gould and Lewontin argued that many evolutionary studies are just-so stories without rigorous hypothesis testing.
Queller was one of the authors who took on Gould and Lewontin's challenge. His 1995 paper was written in the same spirit and was titled: 'The spaniels of St Marx and the Panglossian paradox: a critique of a rhetorical programme' (Queller 1995). A short explanation is again required. One of the sections of Gould and Lewontin's paper was named 'The master's voice re-examined'. While 'the master' referred to Darwin in Gould and Lewontin's paper, Queller intentionally misinterpreted the allusion and substituted Marx. The spaniel refers to the logo of the record label His Master's Voice, which shows a dog listening to the voice of his dead master on a creaky 78. Putting clever metaphors aside, Queller countered the two points made by Gould and Lewontin as follows. With regard to the first point (characters should always be studied in combination), Queller emphasized the success of reductionism in science. He argued that to solve a problem, starting with one factor A and then, once you know its effect, moving on to factor B and the interaction between A and B, is usually better than starting with all possible factors and all interactions at the same time. The second point (no hypothesis testing) is not so easily refuted. Studies with sloppy methodology ending in a long list of possible explanations for an observed phenomenon are found in all times. However, Queller pointed out that many of the hypotheses are, in fact, testable and that many modern evolutionary studies do just that.
Let us look at an example. Janzen's paper on bamboos was a great story. But it was more than that. It had a clear hypothesis that synchronous flowering evolved to satiate predators. Under this hypothesis one would expect that if you induced some bamboo plants to flower outside their mast years, perhaps by supplying plant hormones, their seeds would suffer higher predation than in mast years. An experiment to test this hypothesis could be set up fairly easily. As far as we know Janzen did not do this for Phyllostachys bambusoides, nor were systematic data collected for the more than 1200 bamboo species that exist, some with masting over shorter periods and some with ordinary perennial life cycles. But this does not mean we cannot test the hypothesis and other authors have done this in different systems. In New Zealand, Kelly and co-workers (1997, 2001) examined two hypotheses exploring benefits from the highly synchronized flowering in the grass Chionochloa pallens (mid-ribbed snow tussock). First, they hypothesized that wind pollination of the grass was better if individuals flowered in synchrony. Second, they hypothesized that plants flowering in synchrony could escape seed predation, as Janzen proposed. By manipulating density they found that the pollination effect was small. Seed predation was high (up to 94%) and masting satiated the insect seed predators but not the birds foraging on seeds.
Kelly and Sork (2002) reviewed the literature and suggested that similar factors played a role in the evolution of mass flowering in unrelated species from different continents. In 16 studies they found support for the cornerstone of Janzen's hypothesis, that there is lower seed predation and therefore higher seed survival in years of high seeding, while three contradictory examples existed. All plants with masting were dominant community species and it is indeed plausible that common species can saturate generalist seed predators more efficiently than rare species. We do not assert that we have answered all questions regarding masting. On the contrary. Kelly and Sork (2002) list many aspects that require additional study. Also we have not answered the question how, starting with an ancient, unsynchronized bamboo species, masting is selected for. Mathematical models that address this question are still quite recent (see Satake and Iwasa 2000). In general, the history of evolution may be difficult to reconstruct. We can, however, go out to the field and collect data that can potentially refute a hypothesis. If seed predation is similar in masting and non-masting years this would refute the idea that seed predation is the driving force behind masting. In this sense evolutionary hypotheses are testable.
1.3 Are evolutionary theories scientific?
The criticism of Gould and Lewontin that hypotheses are not rigorously tested has not disappeared and should be taken seriously. For instance, Murray (2001) wrote a review flavoured with philosophy called: 'Are ecological and evolutionary theories scientific?' In his review he criticized the explanations that biologists employed afterwards, when all data have been collected. One example he mentioned concerned the size differences between the sexes in birds. In hawks and owls, females are bigger than males and apparently no fewer than 14 hypotheses have been proposed for this. Some of these hypotheses were based on a single, specific observation. While all these hypotheses may in principle be rejected or accepted in a study on a new owl or hawk species, chances are that, overwhelmed by the many possible choices, this new study only states which hypothesis best fits the data or, if none of them do, comes up with a fifteenth hypothesis. Murray argued for fewer hypotheses, more general hypotheses and rigorous testing of the predictions of each hypothesis before moving on to the next one.
Although much of traditional biology is descriptive, that does not mean that biology cannot have predictive theories. Towards the end of the article Murray (2001) mentions life-history theory as an area for which this is possible and other authors (Morgan and Schoen 1997) have made the same argument for
Image not available in HTML version |
Fig. 1.1 Sex systems in higher plants. Male and female function is either present within the same flower, in different flowers on the same plant or on different individuals. Cosexuality refers to both sexual functions being present on the same individual and therefore includes monoecy and hermaphroditism.
sex allocation theory. Sex systems in plants may be an example, in contrast to animals, which are hermaphrodite or have separate sexes, in the plant kingdom many sex systems are found (Fig. 1.1). Plants with hermaphrodite flowers are most common (about 72% of all species) while dioecious species, i.e. those with separate male and female individuals, are rare (c.4%). Gynodioecy, females and hermaphrodites, is relatively common (c.7%). On the other hand androdioecy, males and hermaphrodites, is extremely rare and has only been shown for a few plant species (Pannell 2002). The minority sex systems occur at different places in the plant kingdom. Helped by the recent molecular phylogeny of the flowering plants (Soltis et al. 1999) we can be confident that the evolution of specific sex systems occurred several times in a number of unrelated events. The theory used to explain the different sex systems is based on the idea of natural selection. This idea implies that there is genetic variation, that such variation leads to differences in success and that this leads to changes in allele frequencies. We will detail this in the coming chapters. As we will see, the theory is not perfect or complete, but it does go a long way in meeting Murray's requirements. The theory is general, it is based on general principles rather than on the available data. The theory applies to all plant species instead of just a few, and it can be tested in a number of specific cases. If our model, for instance, correctly predicts which plant species have become dioecious in one plant family, we could always turn to another plant family with different ecological conditions and a different evolutionary history to see if the predictions are correct for this new situation also. In this sense the theory is testable and scientific.
1.4 100% adaptation?
After reading Gould and Lewontin's critique of the easy acceptance that organisms are in all respects fully adapted to their environment, one might expect that the degree of adaptation is a big issue. Well, it isn't really. No one expects organisms to be 100% adapted. Consider adaptation as climbing a mountain without a map and with limited visual range, say only 100 m. You are determined to make it to the top and the best thing to do would be to go to the highest point you can see and from there look again if you can climb any higher. No doubt, this will get you started up the mountain but you may get stuck at a local peak which is higher than anything within a 100 m radius, but lower than the top. Similarly, organisms may change genetically by small steps. Selection will favour the steps that result in 'better' plants until the point where no small step leads to improvement. We can then say that the organism has reached a local fitness peak, from which it can only move by a large step. Organisms adapt but, as in the case of the mountaineer with limited sight, there is no guarantee an organism will always reach the top of the fitness landscape.
Furthermore, adaptation follows environmental change and plants have the difficulty of having to reach a new area by dispersing their seeds. If long-distance dispersal is rare, this may be a lengthy journey. Reid (see Clark et al. 1998) calculated in 1899 that between the glaciation of the UK, some 10 000 years ago, and the Roman occupation, the oak travelled 1000 km north to reach the limit of its northern distribution in Scotland. Later diffusion models calculated that, to travel this distance, daughter oaks would have to grow, on average, 1 km from their maternal parent. This is an unrealistically large distance for an acorn to fall from a tree. This inconsistency between predictions from a simple life-history model and the study of fossil pollen in the soil was called 'Reid's paradox' (Clark et al. 1998). It was resolved when it was realized that rare, accidental long-distance dispersal sometimes occurred. This greatly increases the speed at which tree species travel over the globe. Nevertheless the point remains that plants follow climatic change and we cannot reasonably expect that as soon as a habitat becomes suitable for a certain species, its seeds are immediately present. The climate is now becoming warmer and species ranges are changing accordingly (Parmesan and Yohe 2003). If you transplanted a genotype from the south of Europe to the north and found that it survived or even outcompeted the home genotype, that would be a surprising result. On the other hand, such a result may well be due to limited gene flow so that genetic change is lagging behind environmental change. Such an observation does not alter the principle that under natural selection, the genotype best adapted to the local environment outperforms others.
No organism is probably 100% adapted to its environment, but do we know for certain that plants adapt to their environment? Plants can be transplanted to different environments and are therefore ideal study organisms for tackling such questions. Many early examples of transplant experiments can be found in the studies of Clausen (1951) and co-workers Keck and Hiesey (see also the recent review by Núñez-Farfán and Schlichting 2001). One of their experiments was to collect genotypes of Potentilla glandulosa (sticky cinquefoil) at different heights, to clone these genotypes and transplant them to Stanford (just above sea level), Mather (1400 m) and Timberline (3500 m). A single clone looked different when placed at different heights and was therefore plastic in its response to the environment. The home genotype (the squares on the diagonal) was always the largest of the three genotypes (Fig. 1.2), as you would expect if it was best adapted to its own environment. When grown at the Stanford garden differences between genotypes were small, but in the harsh mountain environment at Timberline the lowland genotype was unable to survive. It is quite likely that the differences in performance that Clausen (1951) documented were due to genetics. One could argue, however, that the clones differed in other respects also. Plants from Stanford could, for instance, have lower nitrogen concentrations that made them more vulnerable to frost on the mountain top. To control for this we need to grow at least one generation of seeds in the lab under standardized conditions. This was done by Jan van Groenendael for two genotypes of Plantago lanceolata (ribwort), one from a disturbed roadside and the other from an adjacent meadow in the Netherlands (Fig. 1.3) (Kuiper and Bos 1992). After seeds were sown in the greenhouse, plants maintained their own typical growth form. The erect plants in the meadow have to compete with tall grasses while the major problem for the roadside plants is to recover from trampling and mechanical damage. The two growth forms thus seem ideally suited for their own habitat and
Image not available in HTML version |
Fig. 1.2 Reciprocal transfer experiment with Potentilla glandulosa in three contrasting environments. The home genotype (squares on the diagonal) performs best in its own environment. Redrawn with permission from Clausen (1951), Cornell University Press.
transferring plants between the habitats showed that the home genotype usually outperformed the other genotype (Kuiper and Bos 1992). We can answer the question about genetic adaptation affirmatively for Plantago lanceolata. Several other studies have shown that plants show rapid genetic change under strong selection. The clearest examples reviewed by Linhart and Grant (1996) include genetic adaptation of lawn plants to a stringent mowing regime and of plants on mine spill to heavy metals. Adaptation was also found under less extreme conditions. Joshi et al. (2001) swapped the common grassland species Trifolium pratense (red clover), Plantago lanceolata and Dactylis glomerata (cocksfoot) between different European countries and found that the home genotype usually won.
1.5 Games plants play . . . against nature
If the critical factor for the plant is to resist an abiotic factor like frost, the optimal strategy is probably independent of the behaviour of other types in the population. This problem is known as straightforward optimization or
Image not available in HTML version |
Fig. 1.3 Plantago lanceolata genotypes from a hay meadow and a nearby disturbed roadside. The phenotypic differences were maintained when seeds of both types were germinated and grown under identical circumstances. Picture courtesy of Jan van Groenendael.
a 'game against nature'. A nice example is Dan Cohen's (1966) seed bank model. Cohen noticed that in some species less than 100% of the seeds germinate the next spring; a fraction of the seeds germinates in the subsequent years. Cohen considered this a strategy of spreading the risk in an unpredictable environment. Seedling establishment is a risky business. In some years young plants will make it through the vulnerable germination and establishment stage, but in other years a drought may start just after germination, wiping out an entire cohort. If 100% of the seeds germinate at the same time, an entire line could go extinct in a single bad year. Delayed germination of some of the seeds helps to bridge the unfavourable period. Cohen thus predicted low germination fractions in species that live in unpredictable environments with a large variance in seedling establishment. In Cohen's model plants are spreading the risk, just as stockbrokers operate in unpredictable markets by not investing all their money at once.
© Cambridge University Press