Genetic erosion in small mammal populations following rain forest fragmentation in Thailand

Genetic erosion is an important population viability threatening phenomenon that is well understood in theory but virtually unstudied in nature. Although the negative effects of genetic erosion in small populations following habitat fragmentation are widely appreciated the critical early stages of the process have gone undocumented because the changes are rapid and difficult to monitor. An opportunity to study the process arose when the Khlong Saeng valley, peninsular Thailand, was flooded in 1987 with the construction of Chiew Larn reservoir. Forested islands isolated in the lake were originally inhabited by 12 species of small mammals of which three were common enough for studies of their population genetic variability. Nuclear variability was monitored in a forest rat, Maxomys surifer, tree mouse, Chiropodes gliroides, and tree shrew, Tupaia glis, using 6-7 hypervariable microsatellite loci for each species. Variability in years 5-8 post-fragmentation was studied by live-trapping on different-sized islands and in comparable areas on the adjacent undisturbed mainland. It was shown that changes in genetic variability can be monitored non-invasively in free-ranging natural populations, and that microsatellites are powerful markers for the study of genetic erosion. As predicted by theory, the loss of allelic variation occurred faster than the loss of heterozygosity in fragmented populations of all three species. Significant loss of heterozygosity and allelic variation occured in isolated populations of rats and mice in year 7 and year 8 but genetic erosion was slower in the tree shrews, which showed a decline in allelic variation only in the year 8 sample. Differences in rates of genetic erosion indicate that each species responds to fragmentation in a different manner, both demographically and genetically, reflecting differences in life history and behavior. These findings indicate that small fragmented populations of species that are the subject of conservation efforts will lose genetic variation faster than allowed for in current conservation practice; their survival requires that genetic erosion be monitored and when necessary countered by interventive management.