Print PageRecent comparisons of mechanistic models based on specific biology vs. presence-only distribution (correlative) models of species occurrences suggest that mechanistic models may add little value in considering the factors that may underlie distribution dynamics. Here we revisit that comparison and show that basing models of species distributions on mechanistic data (when available) can have several advantages, especially when paired with abundance or survivorship data for an independent test of the model results. For two butterfly species, the sachem (Atalopedes campestris) and the monarch (Danaus plexippus), we show that basing range projections on larval survival rates measured in laboratory experiments on temperature tolerances allows specific predictions of optimal breeding ranges. We used two extensive networks of citizen-scientist monitoring programs to test these patterns. The first is the North American Butterfly Association’s (NABA) continental-scale count program of adult butterflies. The second is the Monarch Larvae Monitoring Project (MLMP) that collects data on egg and larval abundances throughout the United States.
Results/Conclusions
For A. campestris, NABA’s data on adult butterfly abundances show that models based on laboratory data of cold tolerance do a good job of identifying northern range limits, but miss areas of low productivity in warmer climes. This suggests that some other factor limits growth where temperatures are high. In D. plexippus, we show how a substantial zone of lethal or sub-lethal temperatures should exist in the US below approximately 38 degrees latitude based on exposing developing larvae to different temperature regimes. These zones are tested using MLMP data throughout Eastern North America. We found that lethal temperatures identified in the lab correspond to lower juvenile survival rates in the field. NABA data also show lower adult abundances in these zones during the breeding season. This dynamic could provide part of the explanation for the spectacular migratory behavior of monarchs, who overwinter in Mexico, but seek breeding grounds thousands of kilometers to the north. For both species, the use of mechanistic laboratory data formed the basis of specific hypotheses that we were able to test with independent monitoring data. This led to a deeper understanding of the mechanisms underlying current ranges and identified specific gaps in knowledge for further study.
