Identifier

etd-12192016-205053

Degree

Master of Science (MS)

Department

Renewable Natural Resources

Document Type

Thesis

Abstract

Since 1986, Louisiana’s American alligator (Alligator mississippiensis) ranching program has required the release of alligators produced from eggs collected from wild nests to maintain wild populations. This project assessed long-term harvest data (1991-2010s) to estimate survival of released alligators. First, wildlife and fishery harvest models and general inter-disciplinary survival models were evaluated to determine best fit to the data. Second, once the best fitting model was selected, release length, precipitation and temperature from release sites, and an index of hunter effort were added to investigate influences on survival estimates. Release length was included because over time the proportion and size of ranch-released alligators has been modified. The generalized linear mixed model, with a fixed intercept, and negative binomial probability distribution was selected as the best fitting model based on the minimization of differences between observed and expected values. This baseline model without covariates estimated instantaneous annual survival to be 0.87 and 0.89, for male and female alligators, respectively. The final best fitting model suggested that the larger an alligator is at release (to a certain point, benefits diminishing after 139 cm), the higher the chance of survival, and the longer it will be afield, for both males (F_1,169=5.62, p<0.05) and females (F_1,180=5.89, p<0.05). Mean precipitation was also statistically significant and positively associated with survival of both male (F_1,169=12.51, p<0.05) and female (F_1,180=12.33, p<0.05) alligators, suggesting reduced survival in drier years. Additionally, for male alligators, the coefficient of variation for mean temperature was statistically significant (F_1,169=7.19, p<0.05), suggesting lowered survival during years with more variation in temperature between months. Although not statistically significant, fit of the male model improved with the inclusion of the hunter effort covariate. Inclusion of release length along with environmental and hunter effort covariates improved precision of survival estimates resulting in values of 0.79 for males, and 0.81 for females, suggesting survival was influenced by the included covariates. Survival estimates with included environmental variables were very similar to reported wild alligator survival estimates, suggesting that ranch-released alligators respond to environmental conditions similar to wild alligators and that ranch-released alligator estimates may provide insights into wild alligator ecology.

Date

2016

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Kaller, Michael D.

DOI

10.31390/gradschool_theses.4600

Download

Share

COinS