Nectar Production and Spectrum of Insect Visitors in Six Varieties of Highbush Blueberry (Vaccinium corymbosum L.) in SE Poland

e attractiveness of plants to pollinators depends strongly on flower rewards, especially nectar and pollen. Nectar mass, sugar concentration, and sugar mass are known to influence the spectrum and abundance of insect visitors. Respective data on nectar secretion in highbush blueberry (Vaccinium corymbosum) under the climatic conditions of Poland are scarce. is study was conducted in 2002–2005 to assess flower abundance, nectar production, and insect visitors in six varieties of V. corymbosum in Niemce, SE Poland. Flower abundance ranged from 1.63 ± 0.64 (‘Darrow’) to 4.07 ± 0.95 in thousands of flowers per shrub (‘Northland’). Nectar mass, sugar concentration, and nectar sugar mass increased with flower age, peaking between the sixth and ninth day. Significant differences in nectar characteristics occurred between years and between varieties. ‘Bluecrop’ and ‘Darrow’ produced the largest nectar mass (19.08 ± 7.09 and 16.60 ± 8.31 mg nectar per flower, respectively) and nectar sugar mass per flower (6.39 ± 1.52 and 5.76 ± 1.51 mg sugar per flower, respectively). e estimated sugar yield in the studied V. corymbosum varieties ranged from 9.4 ± 3.3 to 20.7 ± 3.8 g sugar per shrub (‘Croatan’ and ‘Bluecrop,’ respectively). Regarding insect visitors, only honey bees and bumble bees were observed. Honey bees comprised 81%–98% of the total number of observed insect visitors. Highbush blueberry, due to abundant blooming and high per-flower sugar yield, is thus a good source of nectar sugars for honey bees.

Highbush blueberry (Vaccinium corymbosum) has become a popular commercial crop in Europe. is species has attracted attention due to its pleasantly sweet fruits which are also a source of biologically active compounds, especially antioxidants (Cardeñosa et al., 2016;Pervin et al., 2016). Poland is one of the leading producers of blueberries -in 2019, total fruit yield exceeded 25,000 tons, and approximately 6,700 tons of fruit were exported, predominantly to countries of the European Union (Statistics Poland, 2019). However, data on nectar production by highbush blueberry under the climatic conditions of Poland are scarce and were collected only regarding very young shrubs (Jabłoński et al., 1983).
e present study was conducted to examine nectar production and insect visitors in six varieties of V. corymbosum. In particular, (i) diurnal patterns of flowering and flower abundance, (ii) nectar secretion dynamics and nectar sugar yield, and (iii) the spectrum of insect visitors were investigated.

Blooming and Insect Visitors
To assess flower abundance per shrub, each variety's number of flowers per inflorescence (n = 15), number of inflorescences per shoot (n = 15), and number of shoots per shrub (n = 15) were counted and multiplied, respectively. e spectrum and abundance of insect visitors/m 2 were recorded during full flowering. Observations were carried out for 5 min in 1-hr intervals between 6 a.m. and 8 p.m. (GMT + 2 hr) for 3 consecutive days.

Nectar Production
Nectar production was evaluated during full flowering. e pipette method described by Jabłoński (2002) was used. To prevent nectar collection by insect visitors before sampling, inflorescences were covered using tulle bags. e dynamics of nectar production were monitored at different stages of flower development (interval = 1 day) from the budding stage (1-2 hr before flower opening; 'Day 0') until corolla wilting. Nectar was collected using tarred glass pipettes, with 12 replications. A single sample contained nectar of 5-10 flowers. e concentration of sugars (% w/w) in nectar was measured using an Abbe refractometer (Carl Zeiss AG, Jena, Germany). Using nectar mass and nectar sugar concentration data, nectar sugar mass (at the peak of nectar sugar secretion) was calculated per flower and per shrub.

Weather Conditions
Meteorological data were collected from a weather station in Niemce. Mean air temperature and precipitation data of 2002-2005 were compared to long-term data collected in 1951-2005 (Table 1). Regarding the month of April, the highest air temperatures were recorded in 2002 and 2005, which exceeded the long-term norm by 1.1 and 1.6°C, respectively; in these periods, 50% less precipitation than usual was recorded. e highest air temperatures of the month of May were recorded in 2002 and 2003. In May 2005, heavy rainfalls occurred, and precipitation exceeded the long-term norm by more than 60%.

Data Analyses
Data were analyzed using STATISTICA soware v.13.1 (StatSo Poland, Cracow, Poland). Before analyses, data distribution was tested for normality. Data on nectar mass per flower and sugar mass per flower were log 10 -transformed; data on numbers of flowers per shrub and on sugar mass per shrub were square-root transformed. A one-way analysis of variance was applied to test differences between years and between varieties. Means were compared post hoc using Tukey's HSD test at = 0.05.

Results
e diurnal pattern of blooming was similar between varieties; therefore, only that of 'Bluecrop' is described. Bloom development peaked in the evening hours ( Figure 1). e number of flower buds that opened between 6 p.m. and 7 p.m. constituted approximately 30% of the total number of flowers developed throughout the day.  'Bluecrop' and 'Darrow' produced the largest sugar mass per flower (6.39 ± 1.52 mg and 5.76 ± 1.51 mg sugar per flower, respectively). Sugar yield per shrub differed significantly between years (F 3, 354 = 5.546; p = 0.001) and varieties (F 5, 354 = 22.913; p < 0.001), and the estimated sugar yield ranged between 9.4 ± 3.3 g and 20.7 ± 3.8 g sugar per shrub, with the highest yield recorded in 'Bluecrop' (Figure 3).
Regarding insect visitors, only honey bees and bumble bees were observed (Figure 4). Depending on the variety, honey bee workers comprised 81%-98% of the total number of insect visitors. Visits of honey bees were distributed evenly throughout the day, while foraging activity of bumble bees peaked in the morning (9 a.m.-10 a.m.) and aernoon hours (6 p.m.-7 p.m.; Figure 1).

Discussion
Nectar mass, sugar concentration, and sugar mass in flowers of V. corymbosum varied between years of the present study. Variability between years was particularly prominent in 'Darrow. ' In this variety, the largest difference in nectar mass was approximately threefold (  in 2005). Irrespective of the variety, the highest nectar mass was recorded in 2005, when precipitation exceeded the long-term norm by 60%. Nectar production depends strongly on a range of abiotic factors such as precipitation, air temperature, air humidity, light availability, and CO 2 concentration, and significant interannual variation in nectar and sugar quantity were reported by numerous studies (Denisow et al., 2014;Enkegaard et al., 2016;Jachuła et al., 2019;Jachuła, Konarska, & Denisow, 2018). Nectar characteristics also differed between varieties, e.g., nectar mass per flower was more than twofold higher in 'Bluecrop' than in 'Darrow. ' Differences in nectar production between varieties is a known phenomenon which suggests the importance of genetic factors regarding nectar secretion (Bertazzini & Forlani, 2016;Bożek & Wieniarska, 2006;Masierowska & Piętka, 2014).
e studied V. corymbosum varieties produced considerable numbers of flowers, which, together with high per-flower sugar mass, resulted in high sugar yield per shrub, ranging from 9.4 to 20.7 g ('Bluecrop' and 'Darrow, ' respectively). us, approximately 47.0-103.5 kg nectar sugars can be expected from 1 ha highbush blueberry crop. In the current study, the estimated sugar productivity of 'Bluecrop' and 'Darrow' was two-to threefold higher than that estimated by Jabłoński  (1983); however, this previous study assessed sugar mass in young (3-4 years old) shrubs, and the authors emphasized that a considerably higher yield can be expected in older shrubs. High sugar productivity by highbush blueberry places this species among the best sugar-yielding cultivated shrubs in the Polish climate. For comparison, the estimated sugar yield of black currant is 4-7 kg/ha (Jabłoński et al., 1997), and that of raspberry is 39-41 kg/ha (Szklanowska et al., 1989).
In flowers of V. corymbosum, anthers open through pores at the tips (Courcelles et al., 2013), thus pollen can be released through sonication by insects, suggesting the so-called buzz-pollination syndrome (Hoffman et al., 2018). is trait can strongly affect pollinator efficiency (Courcelles et al., 2013), and more than 50 Apoidea genera are known to collect pollen released from anthers by sonication (e.g., Bombus spp., Xylocopa spp., and Andrena spp.); however, this does not occur in Apis mellifera (de Luca & Vallejo-Marín, 2013). Nonbuzzing insects can collect only small amounts of pollen from buzz-pollinated flowers, thus they are considered low-efficiency pollinators (dos Santos et al., 2009;Solís-Montero et al., 2015). According to my observations, honey bees were interested only in nectar. It was shown, however,  that even when not collecting pollen from corbiculae, A. mellifera can assume substantial amounts of pollen on its body while foraging for nectar (up to 713 ± 128 pollen tetrads of V. corymbosum) and thereby contribute to pollination (Hoffman et al., 2018). is, in turn, may explain high the fruit set in highbush blueberry when flowers were visited mainly by honey bees (Bożek, 2009). Beside the specific anther structural characteristics, flowers of some V. corymbosum varieties show long, narrow corollas (e.g., 'Bluecrop') and seem to be adapted to pollination by insects with long mouthparts. According to Courcelles et al. (2013), over 40% of visits made by honey bees to 'Bluecrop' flowers were illegitimate, i.e., the bees sucked nectar by inserting their proboscis between the flower corolla and the inferior ovary (thereby avoiding touching the anthers); the authors also observed that A. mellifera, when foraging legitimately, inserted not only the proboscis but also pushed their head into the flower to reach nectar. I did not notice nectar robbing in any of the studied variety but frequently observed honey bees inserting their whole heads in the corolla.
In the USA, which is where highbush blueberry occurs naturally, the most frequent flower foragers were Andrena spp., Bombus spp., and Xylocopa virginica (MacKenzie & Eickwort, 1996;Scott et al., 2016). A previous study on nectar production and pollination in V. corymbosum in the Polish climate showed that flowers are visited mainly by honey bees and occasionally by bumble bees, Andrena spp., and Megachile spp. (Jabłoński et al., 1983). In the present study, only honey bees (>80% of insect visitors) and bumble bees were observed to forage on V. corymbosum flowers. e low abundance of bumble bees and lack of other wild pollinators of highbush blueberry may result from competition with honey bees. Increasing evidence suggests that high abundances of honey bees can reduce population sizes of wild pollinators (Geldmann & González-Varo, 2018;Jachuła et al., 2020;Lindström et al., 2016;omson, 2016).
In conclusion, V. corymbosum is an important food crop in agricultural areas. is species flowers abundantly and produces large amounts of nectar sugars. Highbush blueberry can add to spring and late spring sugar resources, which is especially important for honey bees.