Hydrological controls on oviposition habitat are associated with egg-laying phenology of some caddisflies

Abstract 1. Seasonal variation in resource availability can have strong effects on life histories and population densities. Emergent rocks (ERs) are an essential oviposition resource for multiple species of stream insects. The availability of ERs depends upon water depth and clast size, which vary with discharge and river geomorphology, respectively. Recruitment success for populations may depend on whether peak egg-laying periods occur at times when ER are also abundant. For multiple species that oviposit on ER, we tested whether seasonal fluctuations in ER abundance were concurrent with oviposition phenology. We also tested whether high discharge drowned ERs for sufficiently long periods to preclude egg laying, and whether this problem varied between rivers differing in channel morphology and particle size distribution. 2. We obtained a continuous timeseries of water level (WL) measured every 30 min for two years at sites on three rivers in south-eastern Australia with similar hydrology but different geomorphology. A relationship between WL and ER numbers was determined empirically at each site and these relationships were used to predict ER availability over the two years. Egg masses of ten species of caddisflies were enumerated each month for a year in one river to establish oviposition phenology. 3. Abundance of ERs was inversely related to discharge in all three rivers. ERs were most abundant during autumn and scarce during spring. Site-specific geomorphology resulted in skewed or multimodal distributions of ER abundance each year. Between years, catchment-scale hydrometeorology mediated patterns of ER availability, despite the close proximity of sites. Temporal variance in ER availability was not consistently correlated with mean WL or WL variance. ER variance increased with WL variance, when WL was below a threshold equivalent to mean annual WL. Above this threshold, most ER were likely to be submerged. 4. Oviposition phenology varied strongly among the ten species of caddisflies, with egg-laying ranging from in 1-2 months to year-round. Temporal variations in ER and egg mass abundance were not correlated for most species. Below a threshold minimum number of ER, egg masses were highly crowded onto the few available ER, which is evidence that ER were in short supply. For five species, high egg mass abundance was positively associated with periods of the year when the time above the threshold number of ERs was high. Unusually, two species laid most egg masses during winter and when the time above this threshold was short. Three species showed no association between egg mass abundance and time above this threshold; two of these species laid eggs year-round. 5. Regional hydrometeorology controlled the availability of ERs, but between-river differences were sufficient to deliver different outcomes in the availability of oviposition sites between years and seasons. Caddisflies were rarely prevented from laying eggs but periods when ERs were in short supply created crowding, which may be associated with negative fitness effects on hatching larvae. Geomorphological controls on availability of oviposition resources may have strong implications for the coexistence of species that overlap temporally in egg-laying