In lieu of an abstract, here is a brief excerpt of the content:

chapter 3 Breeding Structure of Neotropical Dry-Forest Tree Species in Fragmented Landscapes James L. Hamrick and Victoria J. Apsit Landscapes that once featured continuously distributed, seasonal dry tropical forests are now characterized in much of Central America by a matrix of pastures and agricultural lands punctuated by occasional patches of remnant forest, secondary forests, and narrow riparian forest corridors. Fragmentation of these once continuous forests could adversely affect several aspects of the biology of tropical dry-forest tree species (Harris 1984; Bierregaard et al. 1992). In particular, changes in pollinator densities and behavior may disrupt or highly modify normal breeding patterns in remnant populations (e.g., Frankie et al. 1997). Such changes in breeding patterns can, in turn, modify levels and distribution of genetic diversity throughout local landscapes (Nason and Hamrick 1997). Most tropical forest tree species are predominantly outcrossing (Bawa 1974; Opler and Bawa 1978; Loveless 1992; Nason and Hamrick 1997). Population genetic studies have demonstrated that, on average, tropical tree species have quite high levels of allozyme genetic diversity and that the majority (86.5%) of the genetic diversity is found within rather than among populations (Hamrick 1994). Indirect estimates of gene flow among populations (Nm = the number of migrants per generation) of tropical tree species indicate that historical levels of gene flow have been high enough to counteract the effects of genetic drift (i.e., Nm > 1.0). Contemporary measures of gene flow made in relatively undisturbed continuous forests indicate that pollen flow rates above 25 percent often occur over distances of several hundred meters (Hamrick and Murawski 1990; Chase et al. 1996; Stacy et al. 1996). The question, then, is, Does forest fragmentation change pollinator behavior so that pollen movement is reduced to the extent that over several generations genetic diversity is lost via genetic drift and rates of inbreeding increase within the remaining fragments? In the following sections we first examine theoretical expectations of the effects of fragmentation on the 30 genetic composition of tropical dry-forest tree populations. We then review case studies of the breeding structure of several Neotropical dryforest tree species located in fragmented landscapes to determine if gene flow is sufficient to increase the low effective population sizes that often characterize fragmented tropical tree populations . THEORETICAL EFFECTS OF FOREST FRAGMENTATION IMMEDIATE EFFECTS The genetic composition of forest tree populations immediately after fragmentation depends on several factors: (1) number, size, and distribution of fragments; (2) original density of the species; and (3) original distribution of genetic variation. If numerous fragments are left after disturbance and/or if fragments are large and close together, most of the genetic diversity in the original populations will be preserved. If, as is more likely, there are relatively few, small, widely dispersed fragments, some low-frequency alleles may be lost, but most genetic diversity present in the original population will remain (Young et al. 1996; Nason and Hamrick 1997). The density and dispersion of trees in the original forest will determine the number of individuals in fragments and fragment-to-fragment variation in population numbers. Finally, if genetic variation was distributed at random within the original population (i.e., low genetic structure ), each of the fragments should retain a relatively high proportion of the overall genetic diversity and should be genetically similar to one another. On the other hand, if genetic variation was patchily distributed originally (i.e., high genetic structure), fragment populations will each contain less of the overall genetic diversity, and genetic differentiation among fragments will be higher. Overall genetic diversity for the total landscape will be maintained, however. An exception would occur if deforestation was not random with remnant forests occupying particular habitats (e.g., steep hillsides, ridgetops, riparian areas). Alleles that preferentially occur in the deforested areas would be lost, decreasing overall genetic diversity. SUBSEQUENT EFFECTS With little gene flow via pollen or seeds between fragments (i.e., Nm 1) at which genetic drift would be the predominate factor influencing genetic structure. Superficially it would appear that the high rates of pollen flow observed among fragmented Neotropical dryforest tree populations should maintain the genetic structure seen in continuous, relatively undisturbed forests. However, a pollen flow rate of 50 percent represents a gene flow rate of 25 percent (i.e., 50% of the gametes come from local maternal trees, 25% of the gametes are locally produced male gametes, and 25% are immigrant male gametes). In a forest fragment with five adults and an effective population size (Ne ) of four (owing...

Share