Frog-biting midge communities between seasons
Sound traps broadcasting túngara frog calls were success- ful at attracting frog-biting midges both during the rainy and dry season (Table 1). During the rainy season, a total of 9052 corethrellids from eight species were captured using sound traps playing túngara frog calls. Significantly fewer midges were cap- tured using túngara frog calls in the dry season, a total of 3936 comprising six species (t-test, t = 3.37, P = 0.001; Table 2a). Both seasons were dominated by a single species, Corethrella ranapungens Borkent, which represented over 85% of the spec- imens identified in our samples. While the subdominant species differed between seasons [Corethrella edwardsi Lane (Lane, 1942) and Corethrella carariensis Borkent for the rainy and dry seasons, respectively], the difference between the subdominant species and the other less common species was slight in compar- ison. Only one species, C. carariensis, was more abundant in the dry season (48 collected) than in the rainy season (20 collected).
Both Shannon diversity and Pielou’s evenness for frog-biting midges were slightly higher in the rainy season than in the dry season (Table 2b). The slopes of the ranked abundance curves, fitted using Zipf models (Table 2c), for the rainy and dry sea- sons were similar (Fig. 2a,b), suggesting that the relative abun- dances of the midge community are similar between seasons. Although there were two species absent from the dry season that were identified in the rainy season, Corethrella appendiculata Grabham (Grabham, 1906) and Corethrella belkini Borkent, these species were rare. Use of Beal’s smoothing to take into consideration the midges attracted to túngara frog calls in both seasons revealed a high probability that both C. appendiculata (66.08%) and C. belkini (76.78%) occur in the dry season and are attracted to túngara frog calls.
Frog-biting midge communities feeding on different anuran species
During the dry season, significantly fewer midges were attracted to pug-nosed tree frog calls (1253 individuals from seven species) than to túngara frog calls (3936 individuals from six species) (t-test, t = 5.19, P < 0.00001; Table 2a). Like tún- gara frog calls, C. ranapungens was the dominant species, rep- resenting 77.43% of the specimens we identified attracted to the calls of the pug-nosed tree frog. The difference between the less common species was more pronounced, with the subdominant species in the sample, Corethrella aurita Borkent, represent- ing 16.30% of the total sample. Corethrella aurita was also the only species attracted in greater numbers to pug-nosed tree frog calls than to túngara frog calls during the dry season, represent- ing only 3.20% of the individuals sampled attracted to túngara frog call.
As a result of the slightly lower dominance of C. ranapungens, Shannon diversity and Pielou’s evenness were higher for midges attracted to pug-nosed tree frog calls than for those attracted to túngara frog calls (Table 2b). Furthermore, three midge species in the dry season attracted to pug-nosed tree frog calls were represented by only one individual. The slope of the ranked abundance curve, fitted using a Zipf–Mandelbrot model (Table 2c), was steeper than for the dry season túngara frog sample; however, this steepness is probably the result of the three species with only one individual (Fig. 2c). The steep initial slope between the first three most abundant species suggests a more uniform distribution in comparison to the midges attracted to túngara frog calls. As with túngara frog calls in the dry season, C. appendiculata was not detected in the midges attracted to pug-nosed tree frogs. Similarly, using Beal’s smoothing to take the rainy season datasets into consideration indicates that there is a high probability that C. appendiculata (67.35%) occurs in the dry season and is also attracted to pug-nosed tree frog calls.