We are studying mating patterns and their evolutionary consequences with particular reference to wild plants native to North America's tallgrass prairie. During the past century, this community has been reduced to isolated fragments within a matrix of agriculture. In fragmented habitat, the movement and foraging patterns of pollinating insects must differ from patterns in continuous habitat, and thus produce differences in plant mating structure. However, these differences have yet to be characterized, and they may be species-specific.
We focus on a metapopulation of Echinacea angustifolia, for which we have detailed spatial information. Building our empirical findings about mating patterns and selection into our modeling efforts, we will (i) develop measures of gene flow at different spatial scales and compare them to existing methods whose assumptions are valid only for uniformly structured populations (Wright 1931, Kimura and Weiss 1964). Beyond this, we will (ii) make evolutionary predictions (a) of rates and patterns of genetic divergence within metapopulations, (b) of genetic variation and capacity for ongoing adaptation to environments as they change, and (c) of probabilities and times to extinction.
In natural populations, ecological circumstances and genetic properties of individuals together determine who mates with whom and which matings yield offspring. For example, spatial proximity of individuals tends to enhance the chance of mating between them; whereas, with genetic self-incompatibility, production of progeny requires difference in mating type. These and many other situations lead to mating patterns that deviate from the analytically tractable condition of random mating. Beyond the production of young, the fitnesses of progeny depend in diverse ways on the realized pattern of mating (e.g., Price and Waser 1977, Waser et al. 2000), and these fitnesses establish the ecological distribution and the genetic composition of the subsequent mating pool. We are determining actual patterns of mating and how they affect reproduction and progeny fitness; these empirical studies will inform our models of evolutionary dynamics within biological communities.
Among non-random patterns of mating, inbreeding has attracted substantial theoretical and empirical attention that has clarified its potential evolutionary consequences. It is well established that inbreeding, whether through self-fertilization or mating between close relatives, tends to increase homozygosity of progeny, and this results in inbreeding depression. Divergent paths of mating system evolution lead from this situation, depending on genetic and ecological circumstances (Lande and Schemske 1985, Charlesworth and Charlesworth 1987, Uyenoyama et al. 1993, Waller 1993, Byers and Waller 1999). Moreover, inbreeding can induce nonlinearity in responses to selection (Shaw et al. 1998). A central goal of our work is an integrated understanding of the interplay among diverse consequences of mating structure, including inbreeding depression, genetic incompatibility, and local adaptation, in an ecological, spatial context in which, for example, pollinator movement depends on the composition and configuration of flowering plants, and spatial variation induces variation in plant fitness. Below, we outline our recent accomplishments toward understanding evolutionary implications of mating patterns in Echinacea angustifolia and our plans to extend this work through our biocomplexity research.
The narrow-leaved purple coneflower, Echinacea angustifolia (Asteraceae), is a perennial, self-incompatible plant that is pollinated by generalist insects. In his graduate research, Stuart Wagenius mapped all 48 remnant fragments with Echinacea in a 6400 ha area of western Minnesota farmland. This metapopulation comprises clusters of plants ranging from one to several thousand flowering individuals. For 28 remnants, detailed observations are available for all individual plants that flowered in 1996 - 1998 (2113 individuals) and for an additional 227 plants in a 45 ha prairie preserve. Isolation of individuals was quantified by high-precision (3 cm) mapping of all these and 403 additional plants. Locational data was organized and mapped by GIS.
In this metapopulation, Wagenius demonstrated that plants in small fragments are more pollen-limited and, hence, produce fewer seeds than those in large fragments. Seedlings grown from seeds collected in small fragments are less vigorous than those from large ones, suggesting inbreeding depression. Estimates of parameters, together with their errors, that are available for inclusion in our evolutionary models include: a) Echinacea pollination rate, b) seed production, and c) seedling vigor, all as a function of spatial distribution of flowering conspecifics.
In ongoing work, Wagenius is monitoring several thousand plants of known maternity growing in randomized arrrays in nature to assess mating types, cross-compatibilities, and components of fitness. He is also monitoring plants whose spatial circumstances he has manipulated in order to test directly the effect of spatial distribution on pollination. He is assaying 400 adult plants from 28 sub-populations for allelic diversity at eight polymorphic enzyme loci.
Characterization of the community of floral visitors is also proceeding through development of a reference collection (approx 400 specimens from 5 insect orders) and analysis of visitation (abundance and diversity) as a function of Echinacea spatial structure. This work will provide estimates, with sampling variances, of d) genetic diversity e) mortality rate, f) pollinator visitation, and g) seed yield, all as functions of local abundance and degree of isolation.
Further studies in progress will provide estimates of the following parameters, with error estimates: h) heterozygosity, which reflects mating structure, as a function of local abundance, i) seedling fitness as a function of heterozygosity, j) seed dispersal distance, k) seedling recruitment rates, and l) pollen dispersal distributions. To estimate h) and i), we are using protein electrophoresis of approximately 2000 seedlings derived from small and large sub-populations. These seedlings have already been established in a common garden, and we are monitoring them to obtain fitness measures. To estimate k), we have sown seeds into diverse conditions in the field. To estimate l), we plan to track movement of a pollen analog. These empirical findings are being used to parameterize models that explore population dynamic and evolutionary consequences of mating structure.
| © 2001- 2004 by the Regents of the University of Minnesota | |
| Last modified on |
