The behavioural and population ecology of an Australian native bee, Amphylaeus morosus Smith (Colletidae: Hylaeinae)

Abstract

This study is the first comprehensive investigation of the behavioural and population ecology of the native Australian bee, Amphylaeus morosus Smith 1879 (Colletidae: Hylaeine), and of any Australian hylaeine species. The project had three main aims. The first was to describe the lifecycle and examine the social nesting behaviour of A. morosus, the second was to assess potential floral resource overlap and resource competition between A. morosus and the introduced European honey bee (Apis mellifera L.), and the third was to investigate sex allocation in A. morosus. The research was undertaken over two years (1992-93 and 1993-94) within moist-dry eucalypt forests in temperate montane regions of the southern Great Dividing Range of Australia (Black Range and Toolangi State Forests, and the Dandenong Ranges National Park). This region may be classified as temperate 'Mediterranean', and plant communities are dominated by a Eucalyptus overstory. In this region, A. morosus nests within naturally excised fronds of the rough tree fern, Cyathea australis, which grows in sheltered run-offs within gullies. Amphylaeus morosus had a univoltine lifecycle in the present study. New nests were founded in spring, while most nests (about 70%) were reused a second time. Presumably, fronds have a finite "shelf life" due to the fibrous nature of the fronds and the generally moist conditions. A. morosus applies a waterproof cellophane to the nest prior to brood production, which takes place from late spring until early-mid summer. There is one nest per frond and the brood are laid sequentially in separate-sealed brood cells, each provisioned with nectar and pollen. New nests had only one adult female during the broodrearing phase, whereas about 24% of reused nests during this phase were multifemale (predominantly two adult females, and occasionally three adult females). This study is the first report of social nesting in any colletid bee, based on reliable data. Direct observations of within-nest behaviours was not undertaken, rather inferences on social nesting are drawn from dissections of a large sample of nests and their contents. Social nesting appears to be the result of reusing a nest, rather than the result of cooperative interactions per se among nestmates. One- and two-adult female reused nests were similar in terms of brood production per adult female (though overall differences were possibly obscured by late season mortality among adults females) and total brood mortality (principally parasitisation by the wasp, Gasteruption sp.). In contrast, new nests produced significantly fewer brood than one-adult female reused nests. New nests lagged behind reused nests in commencing brood production, presumably because comparatively more time is needed by a bee to construct and line the walls of a new nest rather than merely extend and reline a preexisting nest. The role of kin selection in promoting cooperative nesting would appear to be minor, given that average relatedness between nestmates was quite low (r = 0.26 ± 0.06 s.e.). It appears that suitable nesting fronds were not constrained in this study, and thus this factor probably did not influence multifemale nest reuse. Since A. morosus provisions cells as a linear series, and since this behaviour presents opportunities for intra-nidal oophagy, for example, then it is unlikely that nestmates in multi-female nests share reproduction equally. Relative ovary sizes, body sizes and wing wear of bees within multifemale nests were statistically similar, and hence do not support the hypothesis of differential reproductive output between nestmates. However, DNA analyses of maternity relationships within nests are needed to help quantify relative outputs between nestmates. In assessing potential floral resource competition between honey bees and A. morosus, this study measured variables that were directly associated with changes in fitness of A. morosus. Specifically, the demographic performance of A. morosus (e.g. brood production, pupal weights, survival of brood, and frequency of nests with adult females) was measured in four experimental sites (each containing six commercial-size honey bee hives) versus four control sites (no enhancement of honey bee densities) over two springsummer flowering seasons. The results of honey bee baiting experiments indicated that the honey bee treatments represented a very large perturbation to the system. Proportional use of pollens from different floral groups by A. morosus in each year was generally as follows: Eucalyptus > Fabaceae > Leptospermum; whereas that for honey bees was generally as follows: Acacia = Eucalyptus = Hypochoeris radicata > Fabaceae > Leptospermum in year one, and Eucalyptus > H. radicata > Acacia = Fabaceae > Leptospermum in year two. Overall, A. morosus and honey bees showed substantial overlap in pollen use- about 50% similarity in each year. Despite these results, no statistically significant negative impact of honey bees on the demographic performance of A. morosus was detected. Three likely and testable hypotheses may explain the lack of any detectable impact of honey bees on A. morosus. One, A. morosus and honey bees vary in their use of the same floral resources over time and space. Two, floral resources were not limiting for A. morosus. Three, increased biomass of bees in experimental sites, reduced the potential impact of honey bees on A. morosus through the effects of 'predator-saturation'. Populations of A. morosus followed Fisherian ratios of equal investment in male and female brood. Conditional sex allocation strategies best explain why male-biased numerical and investment ratios were produced in reused nests, but ratios were female-biased in new nests. Specifically, given that nests are a valuable resource that can be used more than once, then mothers in new nests should be selected to produce more daughters than sons in such nests because only daughters can exploit this resource in the next generation. However, frequency-dependent selection operating on mothers in reused nests to bias their allocation to relatively more sons, gives rise to population-wide 1:1 investment patterns. Additionally, A. morosus is protogynous, female brood were generally heavier than male brood, and comparatively more and heavier brood were produced in the second year of this study. Show more