95th ESA Annual Meeting (August 1 -- 6, 2010)

PS 70-46 - Count your eggs before they invade: Models estimating egg number in clutches laid by invasive apple snails using clutch dimensions and volume

Thursday, August 5, 2010
Exhibit Hall A, David L Lawrence Convention Center
Colin H. Kyle, Ecology and Evolution, University of Chicago, Chicago, IL, Greg Dwyer, Department of Ecology and Evolution, University of Chicago, Chicago, IL, Therese Shelton, Mathematics and Computer Science, Southwestern University, Georgetown, TX and Romi Burks, Biology, Southwestern University, Georgetown, TX

Exotic invasive mollusks often cause severe environmental damage. South American apple snails Pomacea canaliculata and insularum invaded and have caused widespread environmental damage throughout Southeastern Asia. Recently, these invaders established populations in North America. Despite the necessity of monitoring their spread, manual egg counting represents the only current method to quantify apple snail reproduction, which proves both time-consuming and difficult. Quantifying reproduction for these invaders is especially important because they produce hundreds of eggs at one time in large, dense, ellipsoidal clutches and reproduction varies between climates and exotic environments. Consequently, we investigated the use of mathematical models to accurately and rapidly predict the number of eggs per clutch (EPC) of both apple snail species. To test the models, we collected, measured and manually counted P. canaliculata clutches from native populations in Uruguay (N=153) and P. insularum clutches from an exotic population in Texas (N=72). Estimating parameters from the collected data, we then constructed four generalized linear models for each species estimating EPC based on individual clutch dimension (length, width, depth) or volume calculated as an ellipsoid. AIC values compared each model’s goodness of fit to actual counts.


For both species, we found that the model using ellipsoidal volume best estimates EPC and the model using length provides the second best approximation (P. canaliculata: ΔAIC=30; P. insularum: ΔAIC=11). Although volume accounted for most of the variance in the data (P. canaliculata: R2= 0.5868; P. insularum: R2=0.5818), using length still provided reasonable estimates (P. canaliculata: R2= 0.4965; P. insularum: R2= 0.5131) and may inform researchers willing to sacrifice some precision for speed. While previously limited by having to hand count eggs, researchers can now more easily and efficiently estimate reproductive rates of newly established snail populations and conduct quantitative experiments on reproductive rates and hatching efficiency. In the battle to limit damage by destructive invaders, understanding reproductive rates of new populations represents a critical step in effectively predicting and controlling apple snail spread.  Furthermore, the observed effectiveness of using a basic geometric shape to model this biological phenomenon suggests that researchers could use similar models in other systems. Biologists studying other organisms with dense, consistently shaped egg clusters, or, theoretically, any similar structure, could use our methods to develop their own simple but accurate models.