|Legaspi, Benjamin - FPSC, STATE OF FLORIDA|
Submitted to: Environmental Entomology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 24, 2009
Publication Date: May 10, 2010
Citation: Legaspi, B.C., Legaspi, J.C. 2010. Field-level validation of a CLIMEX model for the Cactoblastis cactorum (Lepidoptera: Pyralidae) using estimated larval growth rates. Environmental Entomology. 39(2):368-377. Interpretive Summary: Bioclimatic models are computer programs used to predict the geographic distribution of organisms, such as insects, based on climate. They are especially useful in the case of invasive pests that are still spreading into new areas. In this study, scientists at the Center for Medical, Agricultural and Veterinary Entomology, USDA, Agricultural Research Service in Tallahassee, FL and scientists with the State of Florida used the CLIMEX program to develop a bioclimatic model for the cactus moth, which is a recent invasive pest in the United States and Mexico. Our model predicted that the moth could survive from the Caribbean Islands to Florida, Texas and Mexico, South Africa, southern India, parts of Southeast Asia and in the northeastern coast of Australia. Actual records show that the moth has been found in the Caribbean, Cuba, Bahamas, Puerto Rico southern Africa, Kenya, Mexico and Australia. In North America, the model suggests that the moth may be close to its predicted range along the Atlantic Coast. However, parts of Texas and most of Mexico might still be vulnerable to its spread. In St. Marks, Florida, we used field cage experiments to measure larval growth for all three generations in a calendar year. We compared growth rates measured in the field against those predicted by CLIMEX for each generation using historical weather data. The model prediction of a single growth season did not match field measurements. We ran the model again using weather data collected at a nearby weather station at the time of the experiment. The model correctly predicted three periods of growth at approximately the times we found in the field. We conclude that only time will tell if the cactus moth spreads as far in North America as CLIMEX predicts and that we can improve the accuracy of model predictions for a given time period if we use actual weather data instead of historical averages.
Technical Abstract: A CLIMEX was developed for the cactus moth, Cactoblastis cactorum (Berg) (Lepidoptera: Pyralidae). Model validation was attempted at the global scale by comparing worldwide distribution against known occurrence records, and at the field scale by comparing CLIMEX “growth indices” against field measurements of larval growth. Globally, CLIMEX predicted limited potential distribution in North America (from the Caribbean Islands to Florida, Texas and Mexico), Africa (South Africa, and parts of the eastern coast), southern India, parts of Southeast Asia and in the northeastern coast of Australia. Actual records indicate the moth has been found in the Caribbean (Antigua, Barbuda, Montserrat Saint Kitts and Nevis, Cayman Islands and US Virgin Islands), Cuba, Bahamas, Puerto Rico southern Africa, Kenya, and Australia. However, the model did not predict that distribution would extend from India to the west into Pakistan. In the United States, comparison of the predicted and actual distribution patterns suggests that the moth may be close to its predicted northern range along the Atlantic coast. Parts of Texas and most of Mexico may be vulnerable. Larval growth rates in the field were estimated by measuring differences in head capsules and body lengths of larval cohorts at weekly intervals. Growth indices plotted against measures of larval growth rates compared poorly when CLIMEX was run using the default historical weather data. CLIMEX predicted a single period conducive to insect development, in contrast to the three generations observed in the field. Invasive pests, such as Cactoblastis, have not reached equilibrium distributions and present unique opportunities to validate models by comparing predicted distributions with eventual realized geographic ranges. Only time and more complete records will tell whether Cactoblastis will extend its geographical distribution to regions predicted by the CLIMEX model. In terms of small scale temporal predictions, this study suggests that CLIMEX indices may agree with field-specific population dynamics, provided an adequate metric for insect growth rate is used, and weather data are location- and time-specific.