Analysis of Climatic Factors Influencing Migrations of the Salamander Ambystoma talpoideum

Abstract

Migrations of breeding adult and metamorphosing juvenile mole salamanders, A. talpoideum, were studied in 5 populations in South Carolina [USA] from Sept. 1978 through July 1982. At each breeding site immigrations and emigrations were monitored through the use of terrestrial drift fences with pitfall traps. Onset of breeding immigrations occurred as early as Sept. but migrations en masse did not occur until Nov., Dec. or Jan. Onset of breeding migrations was not related to a threshold temperature or to a threshold amount of rainfall. Breeding migrations always occurred during the coldest but not necessarily the wettest months of each year. Time of peak migration varied annually depending upon meteorological conditions. Total number of breeding adults or breeding population size, was significantly correlated with cumulative rainfall during the time of immigration but not to total days of rainfall or to total days below 0.degree. C. Sex ratio of immigrating adults was significantly biased towards males at the beginning of the breeding season whereas by the end of the season it was biased towards females. Emigration of adults from the breeding sites occurred each March and was consistent among sites and years. After a minimum larval period of 4 mo., metamorphosing juveniles would emigrate from the pond as early as May and continue until the breeding site dried. In relatively permanent breeding sites emigration of metamorphosing juveniles usually continued through Oct. Most larvae did not metamorphose but remained instead in the ponds and matured sexually. During the very dry year of 1980-1981 water levels in 2 relatively permanent breeding sites were substantially reduced when compared to previous years. This drying process probably induced metamorphosis and emigration of sexually mature gilled morphs. Statistical models which predict the magnitude of migrations indicated that rainfall, water level and minimum air temperature were consistently important environmental variables. Multiple regression models using these 3 environmental variables. Multiple regression models using these 3 environmental variables explained 17-55% of the variation in number of migrating salamanders each day. Unquantified factors such as duration of rainfall, time of day of rainfall and time since last rainfall may be responsible for some of the remaining variation.