Eleutherodactylus coqui, House Occupancy Rates

Burrowes, Patricia A., Hernández-Figueroa, Ábner D., Acevedo, Gustavo D., Alemán-Ríos, Junángel & Longo, Ana V., 2021, Can artificial retreat sites help frogs recover after severe habitat devastation? Insights on the use of “ coqui houses ” after Hurricane Maria in Puerto Rico, Amphibian & Reptile Conservation (e 274) 15 (1), pp. 57-70 : 64-65

Coqui House Occupancy Rates

The results of a general mixed model revealed that (1) time since the hurricane as a proxy of forest recovery, (2) microhabitat temperature, (3) house type, and (4) the interaction between time since the hurricane and microhabitat temperature were all significant predictors of oc- cupancy rate of artificial coqui houses by E. coqui ( Table 2).

Over the study period frogs that used these artificial habitats were found more often during the day with an occupancy rate up to 27.3% (mean 10.64, SD ± 4.23), and somewhat less during the night with a maximum occupancy of 13.6% (mean 4.94, SD ± 2.20) during any sampling event ( Fig. 8). The time at which frogs oc- cupied one of these artificial habitats was not indepen- dent of the type of coqui house (X 2 = 104.05, df = 1, P <0.001). Bamboo houses were used mostly during the day, as retreat and nesting sites, while PVC houses were used mostly at night for perching and/or calling (X 2 = 115.94, df = 1, P <0.001, Fig. 9).

The number of frogs occupying coqui houses during the day was found to be inversely related to the mean midday (1200–1600 h) temperature at the microhabitats (Y = -0.68x + 18.35, P = 0.0185 and R 2 = 37.13). Accordingly, frogs occupied bamboo coqui houses significantly more often in the cooler dry sea- son (shaded box) than in the warmer wet months (F = 5.90, df = 13, P = 0.0258, Fig. 8A). The frequency of daytime usage (i.e., occupancy rate) of coqui houses through time since the hurricane, which serves as a proxy of forest recovery ( Table 1), fits a bell-shaped curve that was explained by a quadratic model (y = -0.0004x 2 + 0.426x – 102.24, P = 0.0026 and R 2 = 0.40, Fig. 8A). No pattern in the nocturnal usage could be associated with time after the hurricane; nonetheless, the coqui houses (especially PVC) were used consistently throughout the study ( Fig. 8B). On four occasions, bamboo houses were used for nesting/parental care, and during this period the males stayed inside day-and-night for approximately 27 days until the eggs hatched and the juveniles dispersed ( Fig. 9).

Contrary to what we expected, the mean number of adult frogs observed per survey night was slightly greater in the control transect (18.43 ± 5.98) than in the experimental transect (17.93 ± 5.54; Paired T = -0.313, P= 0.7599; Fig. 9). The higher variance associated with the abundance of frogs in the control transect may reflect greater habitat heterogeneity in this part of the forest ( Fig. 10).