Bee diversity and their floral resource utilization along a distance gradient from the Arabuko Sokoke forest

Abstract

A study was conducted to assess the change in bee diversity along a distance gradient from the Arabuko Sokoke forest and among the three major vegetation types found in the forest. The floral utilization and resource partitioning among solitary and the honeybees was also studied. Sampling sites were selected from the three main vegetation types of the forest namely, Mixed forest, Brachystegia and Cynometra. Each vegetation type had four sampling sites; one sampling site at the forest margin and three sampling sites located in farmlands selected along a distance gradient from the forest margin; at 2 km, 4 km, and 6 km respectively. The number of bees and species collected in four 50 x 6 metres sampling plot at each sampling sites from 0800 hrs1500 hrs on sunny days with fair weather were recorded. The flower cover was estimated as the number of flowers sampled in ten 1m2 quadrat at each plot. Kruskal- Wallis test was used to test for significant difference across the distance gradient and among the vegetation types. Bee diversity was calculated as the Shannon diversity index and similarity in species composition was calculated with Horn's similarity index. Morisita index was used to calculate the overlap between the honeybees and the solitary bees in resource use across the distance gradient. Sixty three bee species belonging to three families, Apidae, Halictidae, and Megachilidae were found around the Arabuko Sokoke forest. Bee abundance increased significantly across the gradient from the forest margins (H = 15.055, df = 3, P < 0.01), while the number of species declined significantly from the forest margins (H = 14.203, df = 3, P< 0.01). Species diversity decreased significantly across the gradient from the forest margins (H = 19.39 df = 3, P< 0.01). Based on Horn's index, similarity in bee fauna across the distance gradient from the forest margins was high, ranged from CH > 0.83 - 0.89. Significantly more honeybees were found at the farmlands than at the forest margins (H = 9.93, df= 3, P = 0.02). There was a significant difference in the number of bees and species recorded among the three vegetation types (H = 37.993 df= 2, P < 0.01) and (H = 10.171, df= 2, P< 0.01) respectively. Similarity in bee fauna found between the vegetation types was high, CH > 0.80, with Cynometra and the mixed vegetation types having the most similar bee community, CH = 0.95. Significantly more member of the family Apidae were found in the mixed forest vegetation type (H = 67.07, df = 2, P< 0.01), while significantly more members of the family Halictidae were found in the Brachystegia vegetation type (H = 1l.96, df= 2, P< 0.01). There was a significant difference in the amount of flower across the distance gradient from the forest margins (H = 22.776, df = 3, P< 0.01), but not in the number of flowering species. However the number of flowering species utilized by bees differed significantly across the distance gradient (H= 22.05, df = 2, P = 0.02). Over 60% of all the bees utilized floral resources from five plant families Polygonaceae, Compositae, Amaranthaceae, Commelinaceae, and Acanthaceae. The honeybees and the solitary bees utilized similar flowering plants across the gradient from the forest. An overlap in floral resource utilization between the social honeybees and the solitary bee species was recorded. Species richness and diversity were found to decline with the increased distance from the forest margins.