• Ksh. Anand Singh
• M. Arun Singh

Analysis of ecological data is complex because the data structure in any ecological set up is in itself complex. In many cases, the normality assumption is often violated and as such fitting the normal linear models to ecological data is not at all the usual way. There are variety of other models which are conflicting in themselves and so choosing an appropriate one is another point of discussion. In this paper, the abundance of earthworm species is modelled through various soil and environmental characteristics in three subtropical forest ecosystems of Manipur, India. Earthworm count is observed in Mixed reserved forest, Disturbed forest and Plantation forest ecosystems in six different locations of each observation site at a soil depth of ten cm. The count of species of a particular type of earthworm observed during the twelve months of the year is regressed on 9 soil characteristics. One of the crucial challenges which ecologists often encounter in dealing with species count data is its inherent complexity arising out of sampling procedure which is further complicated by the presence of excess number of zeroes in the dataset. When the frequency of zeroes is very large and do not readily fit into any of standard distributions mainly because of skewness and over-dispersion, the dataset is referred to as zero inflated. Starting with the Standard Poisson model, we fitted four other different models viz. quasi-Poisson, Zero-inflated Poisson, Negative Binomial and Zero-inflated Negative binomial models. The appropriateness of the models are checked using AIC, BIC and Vuong test. The appropriateness of the models depends on the particular species type and type of forest.

Quasi-Poisson, zero-inflated, AIC, BIC, Vuong test.

"MODELLING SOIL EARTHWORM COUNT DATA WITH EXCESS ZEROES PROBLEM AND CHOICE OF APPROPRIATE MODELS IN DIFFERENT FOREST ECOSYSTEMS OF MANIPUR.", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 5, page no.661-666, May-2019, Available :http://www.jetir.org/papers/JETIR1905992.pdf

"MODELLING SOIL EARTHWORM COUNT DATA WITH EXCESS ZEROES PROBLEM AND CHOICE OF APPROPRIATE MODELS IN DIFFERENT FOREST ECOSYSTEMS OF MANIPUR.", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 5, page no. pp661-666, May-2019, Available at : http://www.jetir.org/papers/JETIR1905992.pdf