Matskási István (szerk.): A Magyar Természettudományi Múzeum évkönyve 86. (Budapest 1994)

Kemenes, I. ; Demeter, A.: Uni- and multivariate analyses of the effects of environmental factors on the occurrence of otters (Lutra lutra) in Hungary

of water, density of the bank vegetation and steepness of the bank. This table also shows the cross-correlation coefficients between pairs of the three scalable environmental parameters. The occurrence of the otter is statistically significantly yet loosely correlated with the steepness of the bank, but shows stronger positive correlation with the density of vegetation and depth of water. However, the latter two parameters are also correlated with one another but not with the steepness of the bank. 2. Multivariate analysis Table 3 shows the parameter estimates for the logistic regression model. The table only shows these estimates for the variables for which the coefficients were significantly different from 0 (Wald statistic, p). Table 3. Parameter estimates for the logistic regression model, based on the data collected during the nationwide survey of the Hungarian otter populations. B: logistic regression coefficients estimated from the data, S.E.: standard error for the logistic regression coefficient, Sig: level of significance (Wald statistic). R: partial correlation coefficient of individual variable. Exp(B): the factor by which the odds of finding otters change when the independent variable increases by one unit. For B and R positive values indicate that as the variable increases in value, so does the likelihood of finding otters at a particular site. Parameter estimates for variables with a B not significantly greater than 0 (Wald statistic, p0.05) are not shown. For the list of variables, see Table 1 Variable B SE. Sig R Exp(B) Water depth 2.35 0.85 0.20 0.00 0.21 Vegetation 2.87 1.05 0.17 0.00 0.33 Disturbing factors - agriculture 2.07 0.73 0.29 0.01 0.11 - other 0.04 -3.22 0.69 0.00 -0.28 Constant -3.15 -0.49 0.00 Given these coefficients, the logistic regression equation for the probability of finding otters at a particular site can be written as Prob(otter) = —-— \+e~

z and Z = 0.85WD+1.05V+0.73A-3.22O-3.15 where WD = water depth, V = density of vegetation, A = agricultural use of area, O = other, general disturbing factors (blocked watercourse etc., see Fig. 2), 3.15 = constant. The positive B values indicate that increasing water depth and density of bank vegetation and general agricultural use increase the probability of finding otters, whereas the negative B value indicates that an increase in the so called other disturbing factors have a detrimental effect on this probability. However, the degrees to which unit changes in these independent variables change the odds of finding otters are not equal: the odds of finding otters are much more resistant to a unit change in the so-called other disturbing factors (Exp(B)=0.04, Table 3) than to unit change in any of the other factors. In other words, decreasing the density of the bank vegetation or water depth by one unit will have much gretaer detrimental effect on the chances of finding otters than an increase by one unit of the so-called other disturbing factors. It is also clear that the other variables investigated, such as steepness of the water bank, the lack of disturbing factors in general, the vicinity of towns or activities related to animal husbandry all have only insignificant effects on the