Distribuição e disponibilidade de recursos florais e estratégias de forrageamento na interação entre beija-flores e Palicourea rígida (Rubiaceae)

Abstract

Some pollination mechanisms represent mutualistics interactions, in which the plants offer nectar and pollen and the pollinators carry the pollen grains. The hummingbirds are the more specialized pollinators among the birds, and their high energetic cost to fly, force them to develop feeding strategies to obtain the highest energetic gains with the lowest costs. These strategies are defined by hummingbird morphophysiological features and by the amount and distribution of floral resources offered by the plants. Changes in resource offer or distribution pattern can modify bird behaviour and feeding strategy. Some studies show which these strategies interfere on the pollen flow, and that territorial feeding behavior could be a factor decreasing this flow. The aim of this work was analyze the spatial and temporal distribution of floral resources offered by Palicourea rigida (Rubiaceae) populations and their relation with the feeding strategies presented by hummingbird visitors, evaluating also if these strategies would interfere in the reproductive success of P. rigida. The field work was carried out in the RPPN of the Clube Caça e Pesca Itororó de Uberlândia, MG (CCPIU) and in the Parque Estadual da Serra de Caldas Novas, GO (PESCAN), from November of 2008 to February of 2009. The distribution of the plants and floral resources in the P. rigida populations was surveyed using ten 100m long transects along which all plants as far as 5m from each side were mapped. Nectar volume and sugar content measurements by using micropippetes and hand refractometer were used to estimate mean energy offer per flower. Maps of the populations with isolines for density of plants and floral resources were organized using the SURFFER program. The dispersion index was used to determine floral resources distribution and fruit set patterns. Pollen tube growth was evaluated by fluorescency microscopy in pistils from flowers exposed to pollination by territorial and non-territorial hummingbirds. Observations of feeding strategies were carried out along transects, both in the CCPIU and PESCAN populations, in patches with different floral resource offer. Further observations were done in the CCPIU, by manipulating the amount of floral resources in patches and verifying if these variations would change hummingbird feeding strategy. The spatial distribution of floral seems to be associated with larger scale plant distribution and no finer well defined patches could be observed. There was no variation of these general patterns through time. Differences in the fruit set rate between plants that produced high e low buds, and the larger number of pollen tubes growing in the flowers that were pollinated by non-territorial hummingbirds indicated that the territorial activity reduces effective pollination. However, territorial feeding strategy was not a limitation for fruit production, since there was a strong correlation between the number of flower buds and fruit-set. Eupetomena macroura and Amazilia fimbriata were the main species of territorial hummingbirds. The mean number of flowers that supported territorial activity was 45,48 ± 19,68 flowers/patch in the CCPIU, and 46,8 ± 15,9 flowers/patch in the PESCAN. Nonetheless, these values varied according to specie, since for E. macroura the value that supported territorial behavior was 58,36 ± 6,78 flowers/patch in the CCPIU and 59,35 ± 12,02 flowers/patch in the PESCAN, while to A. fimbriata these value was 32,89 ± 5,61 flowers/patch in the CCPIU and 37,19 ± 5,77 flowers/patch in the PESCAN. The floral resource manipulation modified the territorial hummingbird species and feeding strategies, but these modifications were noticed only when resources decreased. Increase in flower number or reward per patch seems to have made territorial behavior more complex since they led to the rise of the number invasions by non-territorial hummingbirds. These two factors are related, so a patch with a lot of flowers exhibited a larger number of non-territorial invasions, which could minimize the negative effects caused by the territorialism on cross pollen flow.