Skip to main content
. 2024 Aug 7;291(2028):20240511. doi: 10.1098/rspb.2024.0511

Table 2.

Result of model selection for finding the best model (marked in bold) to predict zooplankton abundance at the end of the experiment, as well as the coefficients of each included term of the best model. The established models are labelled as null, first, second and third with increasing complexity. Results of pairwise likelihood ratio tests are shown with asterisks representing the significance level. ‘Temperature’ coefficient −0.02 shows that zooplankton abundance remained nearly constant in H mesocosms and the coefficient −0.34 for the interaction 'origin × temperature' shows that zooplankton abundance decreased with temperature in UN mesocosms. Note that the coefficient for ‘origin’ 5.58 suggests that when the experimental temperature is 0°C, the abundance in UN mesocosm would be 5.58 units higher than the abundance in H mesocosms, which is of little biological meaning for our experimental temperature gradient, as the effect is taken over by the interaction term.

model selection best model estimates
response model formula significance terms coefficients
ln(total zooplankton abundance + 1) null 1
first temperature 0.07 origin 5.58
second origin + temperature 0.02* temperature − 0.02
third origin × temperature 0.03* origin × temperature − 0.34