Dispersal Can Sharpen Parapatric Boundaries On A Spatially Varying Environment

Ecology, March, 2000 by Gisela Garcia-Ramos, Faustino Sanchez-Garduno, Philip K. Maini

GISELA GARCIA-RAMOS [1,5]

FAUSTINO SANCHEZ-GARDUNO [2]

PHILIP K. MAINI [3,4]

Abstract. Parapatry describes a geographic pattern in which the ranges of two species have separate but contiguous distributions without any physical barriers between them. We present results from a study of ecological mechanisms to explain parapatry in closely related species. These include competition, spatially varying performances, and dispersal that depends on the densities of both species. We use a model consisting of two coupled nonlinear reaction-diffusion equations with density-dependent diffusion terms and space-dependent Lotka-Volterra-like competing interaction terms. The model is analyzed by using a mixture of phase-plane analysis and numerical simulations.

Results show that competition and dispersal can lead to completely segregated species ranges. Spatial variation favored and provided stability to parapatric distribution. Parapatry occurred under several conditions, including when both species were identical in dispersal, intrinsic rate of growth, and competition but differed in their spatial performances. Results indicate that overlapping distributions and parapatry are equally expected for close species. Moreover, similar species in parapatry tend to exhibit equivalent range sizes. This model explains how species can coexist regionally while maintaining spatial exclusion. It also describes how a species that is rare in distribution can invade the range of a similar and widespread species. We discuss the limitations of using present species distributions for recognizing modes of speciation, and we suggest studying more extensively the relationship between density-dependent dispersal and interspecific competition. We show that density-dependent dispersal can favor segregation.

Key words: coexistence of species; density-dependent dispersal; dispersal vs. competition; invasion; Lotka-Volterra model, spatial variation; modeling parapatry in closely related species; parapatry, spatial model; sharp boundaries; spatial variation; species distributions; species segregation.

INTRODUCTION

Many species are spatially separated by sharp boundary ranges. The distribution of these species ranges is contiguous and without any physical barriers between them. This spatial segregation, or parapatric distribution, is a widespread phenomenon both taxonomically and geographically (Bull 1991), and is a common form of distribution between closely related species (Anderson and Eversen 1978, King 1993). A comparison of range distributions for 130 pairs of sibling species resulted in 41% parapatry, 41% overlap, and 18% separate ranges (Anderson and Eversen 1978). Parapatry is the geographic pattern that is assumed to result from parapatric speciation, allopatric speciation with secondary contact, or peripheral isolated speciation (White 1973, Bush 1975, Endler 1977, Lynch 1989). Although in some cases narrow hybrid zones occur, parapatric boundaries without a hybrid zone have been closely mapped in several studies (e.g., Thomomys talpoides complex [Thaeler 1974], Ranidella frogs [Bull 1991], Proechimys rodent s [Garcia 1981], and Sorex shrews [Neet and Hausser 1990]). Spatial segregation occurs also in species of different genera but with similar ecological requirements (e.g., the barnacles Chthamalus and Balanus [Connell 1961], and the ticks Aponomma and Amblyomma [Bull 1991]). Parapatry without hybridization and in less-related species suggests that ecological factors could explain this spatial pattern. In this paper we study an ecological mechanism for predicting parapatry in similar species. Closely related species are nearly alike in morphology and ecological requirements, especially in the case of chromosomal species that are almost identical. Most explanations of parapatric distributions assume that competitive interactions and species similarity can cause interspecific exclusion upon geographic contact (Chappell 1978, Terborgh 1985, Haffer 1989, Neet and Hausser 1990). Environmental gradients appear also to affect the distribution of contacting species, and parapatric boundaries occur following gradients i n elevation, climate, soil structure, and vegetation (Mayr 1963, Miller 1967, MacArthur 1972, States 1976, Terborgh 1985, Kohlmann et al. 1988, Bull 1991, Bull and Possingham 1995).

Patterns of distribution for two competitive species have been modeled for continuous and discrete space (McLaughlin and Roughgarden 1993). In addition to competition and environmental gradients, theoretical studies in continuous space have tested several other assumptions on growth rate, spatial quality, and mating systems, and included dispersal. However, they commonly predict complete or partial overlap in the species distribution. Models we examined include (a) dispersal in which individuals tend to move towards favorable areas of the environment and away from each other (Shigesada et al. 1979); (b) self-density diffusion and spatially varying growth rates (Namba 1989); and (c) interspecific mating that reduces the species' reproductive success (Ribeiro and Spielman 1986). These and other studies predict only partial segregation with at least some overlap in species distributions (Gopalsamy 1977, Shigesada et al. 1979, Mimura and Kawasaki 1980, Namba and Mimura 1980, Okubo 1980, Murray 1989, Namba 1989, Bull and Possingham 1995, Shigesada and Kawasaki 1997).