RESUMO
The buzzing foraging behavior of female bees for pollen harvesting called the attention of early pollination biologists. Flower types that demand this buzzing behavior comprise about 20,000 species of different and phylogenetically unrelated plant taxa, suggesting that it had independently evolved many times among the flowering plants. Between the late 1970s and early 1980s, theoretical papers had modeled the energetics of buzz pollination, but, up to this moment, no hypothesis was experimentally tested concerning the theoretical basis of the energetics of buzz pollination. We present a cost-effective and simple apparatus, including a digital and highly accurate frequency generator, and a device for the transference of buzz-frequency energy to the receptive floral unity. The receptive floral unities may comprise the entire or partial androecium, or the tubular corolla, or, in some cases, the whole flower. This apparatus can be easily used in both laboratory and field conditions of research, as natural air currents are avoided, and the response of pollen liberation can be quantitatively measured by pollen grain counts that can be captured by adhesion in slide poured with an isosmotic lactate-glycerol media. The maximum displacement of the hardwire beam/claw system was 0.1170 ± 0.0006 mm @ 150 Hz; 0.021 ± 0.003 mm @ 250 Hz; 0.010 ± 0.001 mm @ 350 Hz; 0.0058 ± 0.0001 mm @ 450 Hz; and 0.0082 ± 0.0005 mm @ 550 Hz. Hypothesis contrasting frequency emission and pollen liberation measured as pollen grain counts may be tested in a species flower type by simple linear regression if pollen counts are normally distributed, or ordinal logistic regression, with non-normal counts. The comparison among different flower-type requirements can be tested through appropriate statistical methods for both normally and non-normally distributed pollen grain counts.
