Impact of mulches on floral visitors in day-neutral strawberries and contribution of syrphid flies to strawberry pollination

Animal-mediated pollination is essential for the production of many specialty crops with insects, primarily bees, serving as the predominant pollinators in many agroecosystems. However, increasing management problems with honey bees and declines in native bee populations due to various environmental stressors have highlighted the importance of promoting wild bees and other insect pollinators to ensure sustainable pollination services. This study explores the impact of various mulches on the abundance and richness of floral visitations by insects and the role of syrphid flies in a day-neutral strawberry system. Four different types of mulches were evaluated at the University of Wisconsin West Madison Agricultural Research Station in Wisconsin, USA, during the 2022 and 2023 growing seasons, and included three polyethylene mulches of different color (black, white on black, and reflective metallic) and one brown soil-biodegradable paper mulch. Observations of pollinator floral visitations were conducted and insects were grouped into broad morphotypes. In addition, syrphid flies were sampled to assess their pollen loads. Mulch type did not significantly affect the overall abundance or richness of pollinators visiting flowers. Syrphid flies were the most abundant floral visitors, followed by small dark bees, green bees, Orius spp., and honey bees, while Bombus spp. were found in low numbers. Toxomerus marginatus, the margined calligrapher, was the most abundant syrphid species collected, representing 72% of all syrphid flies collected. Overall, 31% of syrphid flies captured carried strawberry pollen.  A few syrphid species carried significant amounts (>200 grains) of strawberry pollen on their body. The findings suggest that while mulch type did not impact pollinator visitation, syrphid flies may be important for strawberry pollination as they were abundant, at least in 2022, and carry strawberry pollen. Our results underscore the need for further research to understand the complex role of non-bee pollinators in strawberry pollination, and to further investigate the role of syrphid flies as pollinators in different agroecosystems.

Methods
Floral Visitor Observations

To assess pollinator visitation, observations were conducted between 9:30 am and 3:30 pm on days with no or partial cloud cover, and sustained wind speeds below 24 kmph. Observations were done every other week from July 13 to October 5 in 2022 and weekly from July 27 to October 5 in 2023.

In each row, up to five flower clusters were observed for 10 min and any flower visitors were recorded, with the order of rows observed randomized on each sampling date. The total number of flowers observed was recorded as the number of open flowers varied across clusters. A flower visitor was defined as any insect which appeared to be interacting with flowers, or which landed on open flowers. Insects visiting were identified to species when possible, including all honey bees and bumble bees. If a species was not identifiable due to short visit duration, small size, or complex identification traits (such as syrphid flies), they were recorded into one of the following visually discernible groups (morphospecies): green bees, small dark bees, large dark bees, or the most specific possible classification such as family or order (McIntosh et al, 2023). The number of individuals visiting flowers per species or morphospecies per flower was recorded.

Syrphid Collection and Pollen Assay

Syrphid flies were collected in 2023, due to their high abundance in 2022, to assess their pollen load for the percentage of strawberry pollen. Flies were collected via sweep net, taking 20 sweeps over the tops of strawberry plants in each bed, weekly.  Targeted sweeps were made to capture flies observed foraging in rows where the 20 sweeps did not capture any individuals. Captured flies were euthanized in a potassium cyanide kill jar and stored in individual 2 ml microcentrifuge tubes at -20ºC until processing.

Pollen load was measured by thawing the frozen flies for 30 min at 2ºC and submerging them in 1.25 mL of 50% ethanol. Flies were then vortexed for 30 sec before being rinsed with 0.5 ml of 50% ethanol and placed into a clean microcentrifuge tube with 70% ethanol for long term preservation. Rinsate containing pollen was collected and allowed to evaporate for 5 d at 65ºC, before being reconstituted with 40 µL of 95% ethanol and 10 µL of Calbera’s Stain (1 g acid fuchsin, 150 ml DI water, 50 ml glycerol). This solution was then added to a 10 µL haemocytometer (Hausser Scientific model 3100 improved neubauer, Horsham, PA). The total number of all pollen and strawberry pollen grains were counted under a compound microscope (Olympus BX60, Center Valley, PA). Pollen was identified only as strawberry pollen or other species’ pollen, based on reference images of pollen sourced from the Pollen Atlas (Stebler, 2024), an open-source pollen imaging database; a stained sample of pollen from a flower collected in the field was used as a secondary comparison source. Pollen ID was aided in that most other pollen observed was likely from various weeds in the family Polygonaceae, and bared little resemblance to the smaller, smooth, and roughly triangular pollen of Fragaria x ananassa.

Flies were identified to the furthest taxonomic level using the key provided by Miranda et al., 2023. The flies we were unable to identify based on our abilities were categorized as either large or small syrphids, with small syrphids being smaller and large syrphids being as large or larger than Toxomerus marginatus.

Statistical Analysis

Statistical analysis was conducted in R Studio Team version 2023.12.1+402 (RStudio: Integrated Development for R. RStudio, PBC, Boston, MA) using R Studio Team packages.  Observational data were analyzed using a negative binomial general linear model. Models were evaluated for accurate fit of data using transformed QQ plot and residuals vs. fitted plot. Further fit analysis for zero-inflation and dispersion were conducted using the package DHARMA. A type III Wald chi-squared analysis of deviance test was conducted on all models to check for significance of specific terms. For all models, following significant fixed effects, Tukey post-hoc tests were performed to evaluate pairwise comparisons among the mulch treatments and interacting effects, and no reduced models were utilized. 

A model of this type was used to analyze pollinator abundance of floral visitation as well as for the analysis of each of the seven individual bee and syrphid morphospecies and was fitted using the GLMTTB package. Total observed visits over the 10 min observation period were used as the response variable; block (levels = A, B, C, D), year (levels = 2022 and 2023), treatment (levels = black, metallic, paper, and white), and morphospecies (see Table 1.1 for levels used) as well as year: treatment, year :morphospecies, and morphospecies:treatment were included as fixed-effect variables. An offset term of the natural log of the number of flowers included in each observation was included. No random-effects variables were used. Richness and abundance of floral visitation were calculated using the package VEGAN and were analyzed using a Poisson generalized linear model with log link. Block, treatment, year, and a treatment-year interaction were included as fixed-effects variables, with flower count as an offset term and no random-effect variables included.