Optimizing line intercept sampling and estimation for feral swine damage levels in ecologically sensitive wetland plant communities

Abstract

Ecological sampling can be labor intensive, and logistically impractical in certain environments. We optimize line intercept sampling and compare estimation methods for assessing feral swine damage within fragile wetland ecosystems in Florida. Sensitive wetland sites, and the swine damage within them, were mapped using GPS technology. Evenly spaced parallel transect lines were simulated across a digital map of each site. The length of each transect and total swine damage under each transect were measured and percent swine damage within each site was estimated by two methods. The total length method (TLM) combined all transects as a single long transect, dividing the sum of all damage lengths across all transects by the combined length of all transect lines. The equal weight method (EWM) calculated the damage proportion for each transect line and averaged these proportions across all transects. Estimation was evaluated using transect spacings of 1, 3, 5, 10, 15, and 20 m. Based on relative root mean squared error and relative bias measures, the TLM produced higher quality estimates than EWM at all transect spacings. Estimation quality decreased as transect spacing increased, especially for TLM. Estimation quality also increased as the true proportion of swine damage increased. Diminishing improvements in estimation quality as transect spacings decreased suggested 5 m as an optimal tradeoff between estimation quality and labor. An inter-transect spacing of 5 m with TLM estimation appeared an optimal starting point when designing a plan for estimating swine damage, with practical, logistical, economic considerations determining final design details.