Jonathan Roughgarden

Evolution of Niche Width

A model is developed to predict evolution in a population which contains a variety of ecologically specialized phenotypes. Individuals of each phenotype are assumed to specialize on a specific region of a resource axis present in the environment. The model, incorporating density-dependent effects on the fitness of the various phenotypes, predicts the number of individuals of each phenotype through time for both asexually and sexually reproducing populations. The lizard species, Anolis roquet, illustrates the kind of population treated by the model. The resource axis is prey size; small lizards preferentially utilize small prey, while large lizards exploit large prey. Field data on this species are used to estimate some of the parameters of the model. The model shows that there is an optimum number of individuals of each phenotype for a given set of resources and a given regime of interphenotypic competition. If the population has this optimum distribution of individuals, then the fitnesses of all phenotypes are equal and the largest population size is achieved. Any other distribution yields too many individuals of certain kinds and too few of others, and a smaller total population size. The model is solved analytically to characterize this optimum distribution thoroughly. The model also shows that an asexually reproducing population, provided it has enough genetic variability, attains the optimum distribution very rapidly, while a sexually reproducing population attains the optimum distribution only slowly if at all. As a result, the sexual population often suffers from overcrowding in some phenotypic classes and undercrowding in others, and has a lower population size. The reason a sexual population may not attain the optimum distribution is that there is always a direct relationship between the shape of the distribution of phenotypes in the whole population and the shape of the distribution of phenotypes among the offspring of each mating within the population. In particular, the variance in the populations phenotype distribution is always approximately twice the variance in the offspring phenotype distribution. Therefore, if a sexual population is to attain the optimum population distribution, its offspring distribution must have a certain shape. Natural selection will mold the offspring distribution to this special shape, but presumably this molding takes a very long time. If this is true, a sexual population in a new or changed environment attains the optimum population phenotype distribution much more slowly than an asexual population. The niche width of a population is the length of the interval on the resource axis from which the population obtains most (say 95%) of its resources. There are two components to niche width: the within-phenotype component, which is due to the variety of resources used by each phenotype, and the between-phenotype component, which is due to the populations having a variety of phenotypes. The model summarized above concerns the evolution in asexual and sexual populations of the between-phenotype component of niche width. Also, an expression is derived which partitions niche width into the within-phenotype and between-phenotype components such that the total niche width is simply the sum of measures for these two components. Ecological release is an expansion in niche width by a population which originated in a packed fauna and which has newly colonized an open environment. A population may undergo release in either the within-phenotype or between-phenotype components of niche width or both. Because asexual populations attain the optimum population distribution of phenotypes more rapidly than do sexual populations, asexual ones have a greater potency for release in the between-phenotype component. This claim seems an important explanation for the colonizing ability of asexual species. The process of faunal buildup on an island is a race between a widening of the offspring phenotype distribution of the first species there and dispersion to the island by members of some other ecologically differentiated species. If the widening of the offspring phenotype distribution, and hence the niche width, is slow enough, vacant regions exist on the resource axis which facilitate establishment of emigrants from elsewhere. In plants, an alternative to achieving ecological release by widening the offspring phenotype distribution is to develop self-compatibility.

Demographic Theory for an Open Marine Population with Space-Limited Recruitment

We introduce a demographic model for a local population of sessile marine invertebrates that have a pelagic larval phase. The processes in the model are the settling of larvae onto empty space, and the growth and mortality of the settled organisms. The rate of settlement per unit of unoccupied space is assumed to be determined by factors outside of the local system. The model predicts the number of animals of each age in the local system through time. The model is offered in both discrete and continuous—time versions. The principal result is that the growth of the settled organisms is destabilizing. In the model, there is always a state where recruitment balances mortality. However, growth can interfere with recruitment and can destabilize this steady state, provided also that the settlement rate is sufficiently high. The model suggests that two qualitatively distinct pictures of population structure result, depending on the settlement rate. In the high settlement limit, the intertidal landscape is a mosaic of cohorts, punctuated with occasional gaps of vacant substrate. In the low settlement limit, the intertidal landscape has vacant space and organisms of all ages mixed together and spatial variation in abundance is caused by microgeographic variation in settlement and mortality rates.

Spatial variation in larval concentrations as a cause of spatial variation in settlement for the barnacle, Balanus glandula

SummarySettlement rates of the high intertidal barnacle, Balanus glandula, were monitored at three sites in the rocky intertidal zone in Central California simultaneously with measurements of larval concentrations in the adjacent water column. In both 1983 and 1984, settlement rates onto vacant substrate differed among the sites by nearly two orders of magnitude. For all sampling dates, this spatial variation in settlement mirrored the spatial distribution of Balanus glandula cyprid concentration in the water column. A perfect rank correlation was found between cyprid concentrations near a site and subsequent settlement. A noteworthy observation was that the sites switched rank in their settlement rates from 1983 to 1984. This change in settlement rankings matched a switch in rankings for cyprid concentrations.Settlement itself appears to be an important cause of the spatial pattern of cyprid concentrations. Comparing the rates of settlement to estimates of the number of cyprids available at a site suggests that settlement causes a large drain on the cyprid population as a water mass passes over successive sites. No consistent spatial patterns were found in the distribution of other major plankton groups (calanoid copepods) that are similar in size to Balanus cyprids but do not settle.The large differences in settlement rates among these sites were previously shown to be a leading cause of large differences in the structure of benthic barnacle populations. The close correspondence shown here between these large differences in settlement and differences in larval concentrations suggests that nearshore oceanic processes affecting larval arrival contribute to the control of benthic community structure.

Density‐Dependent Natural Selection

Density—dependent selective values illustrate the evolutionary effect of population—regulating processes that diminish an individuals probability of survival with increased crowding. The selective values, assumed to decrease as a linear function of density, lead in a mild environment to the evolution of phenotypes having a high carrying capacity, K, at the expense of a low intrinsic rate of increase, r. A graphical technique shows that selection causes evolution of phenotypes having a high r at the expense of a low K in harsh seasonal environments. A mathematical technique developed for analyzing evolution in coarse—grained seasonal environments reveals genetic mechanisms, including ones with full dominance, with which a moderately harsh seasonal environment causes stable polymorphism between high—r and high—K genes. The energy balance equation demonstrates the role of high—r and high—K phenotypes in the populations energy flow. A high—r phenotype makes a large expected contribution to the populations productivity under conditions of negligible crowding, and a high—K phenotype has, for a given contribution to the populations productivity under uncrowded conditions, a low sensitivity to having that contribution diminished by crowding.

Resource partitioning among competing species--a coevolutionary approach.

Abstract A reasonably general theory for predicting the outcome of coevolution among interacting species is developed. It is applied to a model for resource partitioning among competing species. Current theory for resource partitioning is based on derivations of a “limiting similarity”—i.e., a limit to how similar competitors can be to one another consistent with coexistence. This theory presumes there is a mechanism, perhaps invasion and extinction, which causes competitors to attain the limiting similarity. The view taken in this paper is that partitioning is an evolutionary compromise between pressures for character displacement and disadvantages inherent in the shift to different resource types. A set of principles is offered for the evolution of the parameters in ecological models. (1) For single population models natural selection causes the parameters ultimately to assume those values which produce the highest equilibrium population size. (2) For models of interacting populations, but without interspecific frequency-dependence, natural selection causes the parameters to assume values which produce either the highest or lowest equilibrium population size for any species depending on the sign of the “feedback” in the community obtained by deleting that species. (3) For models of interacting populations with interspecific frequency dependence natural selection leads to parameter values which produce intermediate equilibrium population sizes. A function called the conditional equilibrium population size is introduced. Provided (a) the mean fitness is a maximum in each species at a stable coevolutionary equilibrium and (b) there is negative density-dependence in each species then natural selection causes the parameters to assume values which produce the highest conditional equilibrium population size for each species. These coevolutionary principles, applied to a model for resource partitioning, entail that the niche separation between species relative to given niche widths, increases with the variety of available resources and decreases with the number of competing populations. Also, the evolution of character displacement between two species does not proceed far enough to maximize the equilibrium population sizes of the species involved. These results imply that the relationship between the niche overlap (of nearest neighbors) and species diversity is qualitatively different depending on whether the variety of resources at any place covaries with the species diversity there. Without covariation niche overlap increases with species diversity; with covariation overlap may decrease with species diversity. This study provides the beginning of a theory for the convergent evolution of community structure.

NICHE WIDTH: BIOGEOGRAPHIC PATTERNS AMONG ANOLIS LIZARD POPULATIONS

One aspect of the niche width of a population refers to the variety of resources used by the entire population. This aspect may be measured by the variance of the populations resource-utilization function. Two components make up a populations niche width. The within-phenotype component measures the spread of resources used by the average individual in the population. The between-phenotype component measures the variety of individual specializations in the population. The sum of these two components yields the total niche width. Calculation of these measures is illustrated with data from Anolis lizards. If the populations niche width is mostly composed of the within-phenotype component, it is virtually a monomorphic population of generalists, whereas if mostly the between-phenotype component, it is virtually a population polymorphic with pure specialists. Anolis populations tend more to be monomorphic with generalists than polymorphic with specialists. Recent ecological theory entails four predictions supported by the data on Anolis populations. (1 and 2) A population incurs a smaller total niche width and becomes less polymorphic with specialists either as environmental productivity decreases or the number of competing species increases. (3) If environmental productivity decreases and the number of competing species also decreases, there is little change in the total niche width, but the population becomes less polymorphic with specialists. (4) Evolutionary compression or release of jaw-length variation occurs more slowly than displacement of average jaw length.

The Community Structure of Coral Reef Fishes

Observations of the community of chaetodontid fishes at the northern end of the Great Barrier Reefs are consistent with the hypothesis that it is structured in ways similar to terrestrial vertebrate communities. Our data do not support the need for alternative hypotheses centered on larval habitat preferences and stochastic recruitment. Broad geographic patterns of chaetodontids in the Pacific are similar to those seen in terrestrial vertebrates.

Fish in offshore kelp forests affect recruitment to intertidal barnacle populations

Kelp forests along the coast of central California harbor juvenile rockfish that prey on the larvae of invertebrates from the rocky intertidal zone. This predation reduces recruitment to barnacle populations to 1/50 of the level in the absence of fish. The dynamics of the intertidal community are thus strongly coupled to the dynamics of the offshore kelp community.

Construction and Analysis of a Large Caribbean Food Web

We document the construction of a relatively large food web (44 species) from the island of St. Martin in the northern Lesser Antilles, and compare it with patterns observed in other, generally smaller food webs. In constructing this web, we integrate data from a variety of studies, many of which focussed on Anolis lizards and their vertebrate predators. In addition to determining the links between predators and prey, we estimate the frequencies of predation (the link strengths), and find an approximately bell—shaped distribution with a majority of links of intermediate frequencies. Some of the properties of this web contrast strongly with those of webs in the ECOWeB compilation. In particular, our analysis shows this web to possess an unusual richness of intermediate species (relative to top predators or basal species) and of links between those intermediate species. The number and lengths of chains are also unusually high, as is the degree of omnivory. Nor does this web match the predictions of the cascade model, which predicts even higher proportions of intermediate species and links between them, and even more numerous chains. It appears that these and other differences are not due simply to the large number of species involved here, but it is not yet clear whether they should be ascribed to the completeness with which some of the diets are known, to differences between the ways this and other webs were constructed, or to unique ecological conditions on the island of St. Martin.

What Does Remote Sensing Do For Ecology

The application of remote sensing to ecological investigations is briefly discussed. Emphasis is given to the recruitment problem in marine population dynamics, the regional analysis of terrestrial ecosystems, and the monitoring of ecological changes. Impediments to the use of remote sensing data in ecology are addressed.