Comparing the pollination services of honey bees and wild bees in a watermelon field

• The main representative genus of wild bees was Lasioglossum.
• A single visit from either bee type effected pollination.
• Wild bees needed a significantly lower mean number of visits to effect pollination than honeybees.
• None of the quality characteristics of the developed watermelon differed significantly for both tested pollinators.

Field studies of pollination services of honey bees and wild bees were carried out at a watermelon crop in western Crete. Specifically, pollination treatment by honey bee or mining bee species was applied to female flowers of a watermelon plant. There were also three other treatments; a no-visit treatment, an open-pollinated treatment and a ‘hand’ pollination treatment. Comparisons were based on the number of single bee visits to treated flowers, fruit abortion rates, seed sets as influenced by bee type and the quality characteristics of the developed fruit. The main representative genus of wild bees in the studied area was Lasioglossum. Whilst it was possible that a single visit from either bee type effected pollination, wild bees needed a significantly lower mean number of visits to effect pollination than honey bees. In particular they spent three times as long at each flower compared to Apis mellifera. None of the studied quality characteristics of the developed watermelon fruits (mean weight, brix, number of seeds per fruit and weight of seed) differed significantly for both tested pollinators. The current study provides theoretical and practical evidence to the growers of watermelon crop of the existence of alternative pollinator species that can be targeted for development and management because of their high pollination efficiencies. Therefore the results support the hypothesis that the native bee community can provide an equivalent service to that of managed honey bee pollinators for watermelon, a crop that has heavy pollination requirements.