2.5. Statistical Analyses

Linear regression analyses of log-transformed variables were used to determine the fungus garden biomass growth and nest architecture development in relation to the increase in number of colony workers. The nest architecture variables evaluated were external area, number of entrance holes, number of fungus chambers, total fungus chambers volume, and nest depth. Bivariate scaling analyses (log–log) of worker number with (a) fungus garden biomass, (b) number of fungus chambers, (c) number of entrance holes, (d) external area, (e) total fungus chambers volume, and (f) depth of nests were performed to test allometric relationships. The growth in fold number in workers was also used to determine the proportion of growth for every variable. In addition, the dimensions of chambers and tunnels were subjected to regression analysis, and data on architecture parameters, number of workers, and fungus garden biomass according to colony age (2, 8, 20, 32, 44, and 56 months) were subjected to descriptive statistics to obtain the mean, standard deviation, and maximum and minimum values. All analyses were performed with Statistica software version 7.0 (StatSoft, Tulsa, USA) [28].