Investigating status of house sparrows (passer domesticus indicus)
Student name: Mr Karthik K.
Guide: Dr Prateek Sharma
Year of completion: 2012
Host Organisation: Bombay Natural History Society
Supervisor (Host Organisation): Dr Asad Rahmani
Abstract: The current study mainly tries to get the insights on the population and distribution status of House sparrow
(Passer domesticus indicus) by collecting secondary information at a National level by running a citizen
science project and collecting field data in Bangalore. A website www.citizensparrow.in was launched to collect
data and currently there are over 9400 records from 5400 participants from 7671 locations. Nested pair
sampling approach with point count for house sparrows was followed to collect field data on the habitat
requirements of House sparrows. Wilcoxon rank tests( Non-parametric tests) results show that the sparrow
counts in the sparrow presence sites of urban and non-urban areas do not differ significantly ( p = 0.222 ). The
sparrows counts were more in the areas which had low average building height (p =0.017), areas with lesser
non-built-up spaces (p=0.001), areas with lower cluster ratio ( ratio of height of building to number of houses)
value ( p =0.017). The sparrow count data showed negative linear correlation coefficient individually with all
the three individual site covariates and also the low R-square value in the multivariate linear regression. Two
Generalized linear mixed models that is, Poisson distribution with log link function and Negative binomial
distribution with log link function were then followed to model the sparrow counts. Negative binomial
distribution was found to be more accurate with AIC value of 162.651 ( compared AIC value of 459.667 for
Poisson log link model), and also appropriate as the detection probability was assumed to be 1 which might
have resulted in inflated 0 count value of the sparrow counts. But in all three above mentioned models the
cluster ratio value is insignificant site covariate to model the sparrow count.
Key words: House sparrows, population status, citizen science program, nested pair sampling, Point counts,
Wilcoxon test, Generalized linear mixed models.
