Using a binomial mixture model and aerial counts for an accurate estimate of Nile crocodile abundance and population size in the Kunene River, Namibia

Type Journal Article - South African Journal of Wildlife Research
Title Using a binomial mixture model and aerial counts for an accurate estimate of Nile crocodile abundance and population size in the Kunene River, Namibia
Author(s)
Volume 46
Issue 2
Publication (Day/Month/Year) 2016
Page numbers 71-86
URL http://www.bioone.org/doi/abs/10.3957/056.046.0071
Abstract
The Nile crocodile, Crocodylus niloticus, is found throughout sub-Saharan Africa, including
Namibia, Botswana and Angola. The species was transferred from CITES Appendix I to
Appendix II in 2004, although it is recognized as peripherally endangered in Namibia due to
diminishing habitat availability primarily from human encroachment. In 2013, a species
management plan was approved in Namibia to assess the management of the Namibian Nile
crocodile populations. During 2012, an aerial survey was conducted to provide an estimate
of Nile crocodile population numbers. A recently developed N-mixture model for estimation
of abundance and spatial variation was used. Detection probability correlated to animal size
and environmental covariates. Our data also suggest that small crocodiles are easier to
detect during the spring. The abundance for different size classes was influenced by river
complexity (vegetation, depth, channels) and the distribution of human settlements. An
estimated 806 individuals were counted along the 352 km Namibian portion of the Kunene
River system with a conservative estimate of 562 crocodiles regardless of size. The parameter
estimates generated by the analysis suggested that the class-structured model can produce
reliable estimates of total abundance and of local abundance for this section in the Kunene
River system.

Related studies

»