| Abstrakt: |
We present an approach to assess the risk of extinction of a Near Threatened (NT) fish species, Notopterus chitala, in near future with an assessment of the role of abiotic factors and heavy metal level within their natural habitat. The probability of extinction has been estimated by using three different statistical models. Different modelling frameworks are selected in such a way that they cover a wide spectrum of statistical methods ranging from regression to most recent deep learning models, namely nonlinear regression, time-series models, and deep learning models. Five different variants of long short-term memory (LSTM) models have been considered under deep learning modelling framework. All the three habitats, we considered for collection of data for analyses, showed that the extinction risk is quite high with the current trend. Then, time-series analyses with dynamic regression were carried to understand whether the abiotic factors play any role on this declining population of the species. Though habitat is one of the main reasons of population decline in many aquatic species, in this case abiotic factors do not show any negative pressure on the population. We have investigated whether heavy metal pollution copper [(Cu), chromium (Cr), lead (Pb), cadmium (Cd) and zinc (Zn)] within the aquatic body has any role for population decline of the species concerned. Our results show that except salinity no abiotic factor has any influence of the population dynamics of the fish. Among the five heavy metals only Pb, Hg and Cr have some control over the population decline of the species. Abiotic factors and pollution are not the key parameters for population decline of featherback in natural habitat. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink