Introduction
The importance of pollination for terrestrial ecosystem function and services is well known. It is estimated that 78% of temperate flowering plant species rely to some extent on the activity of animal pollinators for reproduction (Ollerton et al. 2011). Many of these are critical for maintaining ecosystem functions providing important services for humans (e.g., increased water and air quality, prevention of soil erosion, timber, fruit and nut production; Kearns et al. 1998, Ashman et al. 2004, Cardinale et al. 2012). In addition, insect pollinators are critically important to global food crops, contributing to 65% of the produced crop volume (Klein et al. 2007). All these ecosystem services are often solely credited to bees (Woodcock 2012, Dicks et al. 2013) and although they are obligate nectar and pollen foragers, this assumption leaves out many other important flower visiting insects (Müller et al. 2006). In fact, there are more than 330,000 insect species such as flies, beetles, wasps, moths and butterflies which may provide pollination services that are often unaccounted for (Larson et al. 2001, Wardhaugh 2015, Rader et al. 2016, Ollerton 2017). Such diverse flower visitor assemblages have been shown to lead to evenness of pollination services; resulting in higher fruit set and fruit weight with fewer blemishes due to insufficient pollination (Nye and Anderson 1974, Lopez-Medina et al. 2006, Hodgkiss et al. 2018).
This evenness of pollination service is the result of each species having their own foraging preferences and tolerances, which leads to more frequent and thorough visitation of flowers from different pollinators, thus increasing the likelihood of conspecific pollen transfer (Fontaine et al. 2005, Hoehn et al. 2008, Blüthgen and Klein 2011, Garibaldi et al. 2013, 2014, Rogers et al. 2014). For example, some insect communities are abundant early in the growing season, maintaining large populations for pollination services for only a week or two, while others peak later in the season, whereas others have a lower population density that is sustained throughout the season (Bartomeus et al. 2013). Thus, there is always a reasonably large pollinator community with potentially complementary and overlapping population peaks (Garratt et al. 2018). Additionally, species have different foraging strategies. For instance, honey bees tend to forage along the tops of plants and travel in a linear fashion, while solitary bees and flies tend to forage on lower branches in a random pattern (Brittain et al. 2013, Brunet et al. 2019). The combination of these two foraging behaviours ensures that the entire plant receives pollination and therefore sets fruit evenly. Finally, different environmental conditions can affect which insects are foraging. High winds and cloudy days tend to keep honey bees inside their hives, while flies, bumblebees and some solitary bees will continue to forage in the rain and the cold (Morgan and Heinrich 1987, Brittain et al. 2013).
One of the leading justifications usually given for the exclusion of non-bee species in pollination studies is that bees are often the most efficient pollinators (Buchman and Nabhan 1996, Kennedy et al. 2013). Bees have a nectar and pollen-dependent diet; as such, their behaviour and foraging techniques often result in high pollen release and frequent flower visiting (Sheffield 2014, Russo et al. 2017, Campbell et al. 2018). When considering non-bee pollinators, their average pollination efficiency per flower visit may be comparatively low, but their ubiquity can lead to high visitation frequency, resulting in equal, or greater, pollen deposition than bees (Larson et al. 2001, Skevington and Dang 2002, Rader et al. 2009, Orford et al. 2015, Rader et al. 2016). This is especially true when considering Diptera, which are particularly speciose and abundant (Skevington and Dang 2002).
To ensure accurate estimation of relative contribution from all crop-pollinating taxa, it is crucial to include non-bee pollinators in crop-pollination surveys, pollination estimates, and pollinator-management practices to avoid taxonomic bias. However, it can be difficult to assign reliable species-level identification of non-bee pollinators and the plant pollen they carry. Given the taxonomic breadth of flower visitor communities, the number of taxonomic experts required for accurate species-level identification is typically unobtainable. Equally, the identification of plant pollen has traditionally been achieved using light microscopy, comparing samples to an extensive palynological collection. However, this method often restricts taxonomic resolution to the genus or family level and requires extensive expertise (Rahl 2008, Keller et al. 2015). DNA-based identification methods, such as DNA barcoding (Hebert et al. 2003) and metabarcoding (Cristescu 2014), can generate sufficient taxonomic resolution to overcome those barriers. The mitochondrial cytochrome c oxidase I (COI) gene is the community-wide accepted standard DNA barcoding region for animals (Hebert et al. 2003) while metabarcoding of the plastid barcode standard gene region of rbcL (large subunit of RuBisCo) (Hollingsworth et al. 2009) and the nuclear region ITS2 (internal transcribed spacer) has been explored for its accuracy in the qualification and quantification of pollen samples (e.g., Bell et al. 2019, Potter et al. 2019, de Melo Moura et al. 2022).
For this study we aimed to identify a community of non-bee flower visitors considered probable pollinators living on three fields of strawberry (Fragaria spp.) crops in Southern Ontario, Canada. We performed species-level identifications of the non-bee flower visiting insects by using DNA barcoding. In addition, pollen loads collected from the bodies of those flower visitors were quantified by particle characterization and identified utilizing rbcL metabarcoding. Subsequently, we assessed how non-bee flower visiting communities changed across the season and in response to variation in local environmental conditions, including temperature, humidity, solar radiation, and wind speed.