Pollen harvesting rates for Apis mellifera L. on Gossypium (Malvaceae) flowers.

Pollen harvesting rates were determined for honey bee foragers on flowers of Gossypium thurberi. Foragers visited 2.2-4.8 flowers/minute and collected 0.5-1.0 mg of pollen per minute. Pollen grains are 103-128 \im in diam., moist and sticky, allowing the collection of 894-1778 grains/minute. Pollen collection time was not significantly different for bees that collected only pollen (0.9 ? 0.1, SD, mg/minute) compared to nectar foragers (0.8 ?0.2 mg/minute) that acquired, then packed pollen. These are apparently the first empirically derived pollen-collecting rates for A. mellifera on any angiosperm. Pollen collection in relation to morphological complexity and hidden pollen is also discussed. Differences in foraging efficiency of bees gathering nectar and/or pollen are readily apparent. Nectar foraging rates for honey bees and other social bees on various plant species and types have been quantified frequently (see table 10.1 on page 173 in Winston, 1987, and references therein). This is probably because nectar is relatively easy to quantify within flowers and bees, and the standing crop of nectar within patches can be readily estimated. In contrast to the situation with nectar, there are few quantitative data on pollen collection (harvesting rates) for Apis mellifera (Park, 1922, 1928; Ribbands, 1953; Win ston, 1987) and other bee species. Such data are important because pollen is the primary proteinaceous and lipoidal food of most bees. Pollen usually consists of discrete particles of 5-200 Aim, rarely as large as 300 ?m (Roberts and Vallespir, 1978; Buchmann, unpubl.). This variability increases the difficulty of rapidly assessing the weight or number of pollen grains within flowers in the field using simple and inexpensive equipment. However, honey bees or other species that have acquired nearly full pollen loads are easily captured to permit the quantification of their respective pollen loads. Within the vast apicultural literature, there is almost no available data on the rate at which honey bees flowers, and particularly scant data on their handling times for pollen extraction from various flowers. Previous anecdotal qualitative estimates for pollen harvesting rates are found in the apicultural literature (Gillette, 1897; Jay, 1986; Maurizio, 1953; Park, 1922, 1928; Ribbands, 1949; Percival, 1947; Singh, 1950; Weaver et al., 1953), but this information is not especially useful for demonstrating how fast honey bees collect pollen under diverse me teorological conditions, or from a variety of specialized structures. Fur thermore, we have no quantitative measures of the time required for a naive forager to discover the most efficient ways to harvest pollen nor how learning 1 Mention of a proprietary product or trademark does not constitute endorsement or recommen dation by the USDA-ARS for its use over any other product. Accepted for publication 8 June 1989. This content downloaded from 157.55.39.104 on Mon, 20 Jun 2016 06:36:48 UTC All use subject to http://about.jstor.org/terms VOLUME 63, NUMBER 1 93 may decrease foraging rates. Our objective was to determine pollen harvesting rates for experienced honey bee foragers visiting simple open (haplomorphic) flowers that contained pollen and nectar. Furthermore, we present field and lab oratory methods that can be modified for a variety of bee species foraging to define pollen collecting rates on many diverse morphologies. Materials and Methods TECHNIQUE FOR OBSERVING AND COLLECTING HONEY BEES! Foragers from tWO to five managed colonies of honey bees of European stock (A. m. ligustica L.) located on the grounds of the Carl Hayden Bee Research Center in Tucson, Arizona were observed. Bees from these colonies located and visited three large cultivated plants of Gossypium thurberi within 100 m of the experimental apiary. A small number of marked foragers (10-20) from these colonies made daily visits to three of these large (4 m wide by 3 m tall) G. thurberi plants from mid-September to October 20, 1987. Bees were active on these plants from 0745 until 1100 hours MST at the end of which time there was little intrafloral pollen remaining. Each morning from October 5-7, 1987 at 0800 MST, the time when cotton flowers were at anthesis, newly arrived bees were visually inspected to determine if any corbicular-resident or loose pollen grains were present. Bees contaminated with more than a few dozen pollen grains (individual grains are easy to see due to their large size) were rejected and other bees observed. When a clean bee was found, an observer timed and followed the bee from an unobtrusive distance of 1-2 meters. Interfloral flight behavior was documented by recording number of flowers visited and whether the bee was collecting only pollen or was acquiring pollen as a normal byproduct of nectar feeding. We realize that no pollen is ever passively acquired, but simply collected by another behavior. A floral visit was defined as a bee landing on a flower and harvesting pollen. A was scored when a bee left one flower, hovered, groomed/packed pollen then returned to the same flower. Bees, marked individuals, usually visited the same flower only once during a foraging bout, thus visits almost always equalled the total number of flowers visited during a bout. Pollen loads increased rapidly on bees and usually after several minutes (2-10) timing was stopped and the bee carefully netted and killed in a vial of 95% ethanol. Collection rates were uniform through out the pollen harvesting process. Seven or eight bees were followed and captured in a period of 1-2 hours each morning on 6 days while flowers were fresh and pollen standing crops were high. Field data collection was terminated after a total of 20 foragers were collected in the above manner. LABORATORY COUNTING OF POLLEN GRAINS ON HONEY BEES! Snap-Cap Style polyethylene vials containing bees and 10 ml 95% ethanol were individually processed by submerging 1-2 cm of a Sonicator micro-tip (Ultrasonics Inc. model W-220F) into each vial for 15 seconds at a power setting of 5. This setting was previously determined to dislodge nearly all of the pollen from bees without also removing bee setae, wing fragments or other debris. The sonicator tip was washed into the resident vial, which was capped until pollen counting occurred. Pollen grain equatorial diameters were determined microscopically at 400 x magnifi cation using an ocular micrometer. Fifty pollen grains were measured yielding a mean and range which was then used to set the size windows on the particle This content downloaded from 157.55.39.104 on Mon, 20 Jun 2016 06:36:48 UTC All use subject to http://about.jstor.org/terms 94 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY