Revised abundance estimates of striped and short-beaked common dolphins in the Gulf of Corinth, Mediterranean Sea

Santostasi N.L., Bonizzoni S., Bearzi G., Eddy L., Gimenez O. 2016. Revised abundance estimates of striped and short-beaked common dolphins in the Gulf of Corinth, Greece. Proceedings of the 30th Annual Conference of the European Cetacean Society, Funchal, Madeira, 14-16 March 2016. Revised abundance estimates of striped and short-beaked common dolphins in the Gulf of Corinth, Greece Nina Luisa Santostasi1,4, Silvia Bonizzoni1,2,3, Giovanni Bearzi1,2,3, Lavinia Eddy1,2, Olivier Gimenez4 1) Dolphin Biology and Conservation, Piegaro PG, Italy; nsantostasi@gmail.com 2) OceanCare, Wädenswil, Switzerland 3) Texas A&M University at Galveston, Galveston TX, USA 4) Centre d’Ecologie Fonctionnelle et Evolutive, CNRS, Montpellier, France Reliable population abundance estimates are essential for conservation, but remain challenging to obtain for cetaceans. The Gulf of Corinth, in Greece, hosts local populations of striped and short-beaked common dolphins. We estimated the combined abundance of the two species (the latter only found in mixed-species groups) in years 2011-2014 using photographic identification and capture-recapture models. We considered a subset of 22,039 high-resolution, high-quality photos featuring appropriately marked dorsal fins. The proportion of unmarked individuals was calculated based on the number of photographs of marked and unmarked dorsal fins obtained daily. Our 4-year average estimate was 1,504 individuals (95% CI 1,376-1,644). This estimate is almost twice as large as a published estimate obtained in 2009 with comparable methods (835 animals; 95% CI 631-1,106). An actual 80% population increase was ruled out. The 2009 sample (23 sampled groups and 116 capture histories) was 2.7 times smaller than the one in 2011-2014 (average of xxxx sampled groups and 311 capture histories per year). Additionally, it only covered the central portion of the Gulf. To better understand the discrepancy, we tested if either insufficient spatial coverage or insufficient sample size in 2009 may have resulted in underestimation of population abundance. First, spatial resampling of the 2012 dataset within the smaller area sampled in 2009 yielded no significant difference in abundance. Second, through randomly re-sampling, we reduced the 2012 sample size to that of 2009. We performed 100 permutations, extracted a capture matrix from each simulated sample and re-estimated abundance. The resulting point estimates were extremely variable, ranging between 716 and 3,174 animals. In conclusion, while spatial coverage did not seem to affect abundance estimates, reducing the number of sampled groups negatively affected accuracy. Simulations such as those outlined here can improve abundance estimates by ensuring that they rely on an appropriate sample size.