Abstract. Temporal variation and phenology of tropical insect communities and the role of environmental factors controlling this variation is poorly understood. A better understanding is needed, for example, to predict the effects of climate change on tropical insect communities and to assess the longterm persistence of tropical communities. We studied seasonal and inter-annual variation in tropical fruitfeeding butterflies by exploiting a unique 137-month abundance time series of .100 species, sampled at 22 locations in the medium altitude montane rain forest of Kibale National Park, western Uganda. Precipitation peaked twice per year, about 20 d after each equinox. Vegetation greenness peaked approximately 33 d later. Species richness and abundance of butterflies peaked about 2 and 3 months, respectively, after the greenness peak. Furthermore, temporal shifts in peaks of butterfly abundances of each 6-month cycle positively correlated with temporal shifts in peaks of vegetation greenness approximately three months before. The butterfly assemblages of ENSO warm phase years differed significantly from assemblages of the other years. To our knowledge this is the first elucidation of bi-annual rhythms in butterfly assemblages. Host plant availability could explain the seasonal cycles in butterfly abundance and species richness, because the 3-month lag observed matches with the egg-to-adult development time in the studied species.

No abstract

International audiencePredation can be one of the key factors that determine abundance in insect herbivore communities, and drive evolution of body size, and anti-predator traits, including crypsis. Population dynamics and selection pressures will depend on the identity of dominant predators in the system, and these may vary substantially among habitats. Arthropods emerge as chief predators on caterpillars in the understorey of non-montane tropical forest, whereas birds dominate elsewhere. In a tropical forest in Uganda, Africa, we evaluated marks on dummy caterpillars that differed in size, material (clay vs. dough), colourant, and plant species on which dummy caterpillars were exposed. We included live caterpillars to estimate the extent to which studies using artificial caterpillars reflect actual levels of predation. Ants and wasps were the most important damagers of dummy caterpillars, whereas bug and beetle damage was very rare, and no bird or small mammal damage was observed. Daily attack rates did not differ significantly from apparent mortality of live caterpillars (daily mortality = 12.1%), but dummy caterpillars made from dough were attacked more frequently (daily attack rate = 18.4%) than those from clay (daily attack rate = 6.9%). Caterpillars of different colour and size, and caterpillars exposed on different plant species had the same chances to be predated. This is in contrast to results from temperate area studies where birds dominate and are not affected by dummy caterpillar material, but prefer larger caterpillars. Our results are consistent with dominant predators on tropical forest caterpillars being invertebrates that are more chemically than visually oriented, so that: (1) material used for dummy caterpillars is important, (2) background matching is relatively unimportant, and (3) being large may have less of a cost. These patterns in predation might facilitate polyphagy and evolution of large body size in tropical Lepidopter

Information on the life span of organisms in the field is essential for elucidating the evolution of life span and aging. We present mark-recapture data (>30,000 marked individuals, >4000 recaptured at least once) on 47 species of fruit-feeding butterflies in a tropical forest in Uganda. The data reveal adult life spans in the field for several species that are significantly longer than previously recorded in Lepidoptera (butterflies and moths). Longevity records for species of which more than 100 individuals were recaptured ranged from 67 (Bicyclus auricruda) to 293 days (Euphaedra medon). In contrast to the majority of Lepidoptera which are short-lived, these all show exceptionally long life spans, and may thus help to better identify factors that affect aging, particularly when combined with information on temporal patterns in reproduction, strategies to avoid predation, and nutritional ecology. These key traits are readily measurable in butterflies and thus studies on fruit-feeding butterflies have much potential for gaining insight into the evolution of life span and aging, especially given the tradition of field-research on butterflies.