Variations in fecundity over catchment scales: Implications for caddisfly populations spanning a thermal gradient

Population sizes of stream insects depend on the number of eggs that enter the benthos. Densities of oviposited eggs have been linked to densities of oviposition habitat, but variability in fecundity has not been explored. Fecundity may differ between sites where females experience different thermal environments during development, because insects generally grow larger and produce more eggs in cold places (e.g. high elevation and latitude). We predicted that because the fecundity of emerging females should increase with elevation then the numbers of eggs in oviposited egg masses (clutch sizes) should also increase with elevation. Alternatively, individuals may routinely disperse along channels or between rivers, mixing individuals that have grown and developed in different thermal environments, in which case clutch sizes may not vary with elevation but could vary greatly within sites. We tested these hypotheses across an elevation gradient of 250–1,080 m above sea level, in seven streams in southeast Australia. We surveyed the fecundity of three species of caddisflies (Trichoptera: Hydrobiosidae) in 2009 (eight sites) and 2010 (17 sites). Female hydrobiosids attach their entire clutch of eggs to an emergent stream boulder, in a single egg mass. At each site we censused the number of egg masses in two random 50 m stretches and photographed a sample of egg masses. Clutch sizes were quantified by counting eggs from the photographs. From numbers of egg masses and mean clutch size per site, we estimated the number of individual eggs per site, to determine whether spatial differences in clutch size affect patterns of egg supply. Potential trade‐offs between egg number and size were also examined. Strong thermal gradients were observed in both years. Sites at lowest elevations were 4.8°C (2009) and 3.6°C (2010) warmer than sites at highest elevations, and accumulated 392 (2009) and 361 (2010) more degree‐days (>0°C) over a period of 88 days. Clutch sizes of each species varied markedly within and between species but did not differ along the thermal gradient. Within species, 83.7–93.7% of variation in clutch size was expressed within sites. Relatively little variation occurred between sites or rivers. Within species, counts of egg masses adequately described relative egg supply to each site, because clutch sizes were similar between sites and rivers. Comparing egg supply between species, however, required including clutch sizes to estimate supply. Egg size differed within but not between species. Significant but small variations in egg size were trivial compared to much larger variations in clutch size. Our results suggest that any spatial differences in fecundity caused by temperature may be erased by dispersal, such that females from multiple locations mix between sites prior to oviposition. A warming global climate may reduce the fecundity of ectotherms such as aquatic insects but consequences for populations are difficult to predict. Moreover, climate‐related effects on fecundity may not be apparent when dispersal by individuals between elevations mixes any effects of climate change on fecundity.