What makes a good pollinator? Abundant and specialised insects with long flight periods transport the most strawberry pollen

Animal pollination of crops, especially by insects, is critical for many crops and can be valued at US$195–$387 billion annually (Porto et al., 2020). Preserving or enhancing this service, in the face ongoing climate change, is, therefore, an important consideration farm managers. An first step towards promoting service a crop understanding which are most pollinators that crop. However, while effectiveness value bees (especially domesticated honeybees) well understood some (Eeraerts 2019; MacInnis & Forrest, 2019), other wild have not been as frequently studied. Nevertheless, general shown to substantially improve yield quantity quality fruit family Rosaceae (e.g. strawberries apples; Abrol MacLeod Identifying protecting pollinators, tandem with appropriate management bees, sensible strategy key however, straightforward plants open, cup-shaped flower morphologies such (Kalkman, 2004). These flowers visited insects several orders 2004), all contribute significantly (Cirtwill 2022). The effective will combine frequent visits large numbers pollen grains deposited per visit. single-visit deposition data resource-intensive collect and, remain rare (see Page 2021 meta-analysis available studies). amount carried on insect's body offers proxy different pollinators' potential easier than (although more difficult visitation alone, common source information about pollinate plants). For example, carry because they hairy (Stavert 2016) will, average, higher chance depositing sufficient visit (Cullen 2021; Földesi 2021). Abundant composite fruits strawberries, where least 70%–80% carpels must pollinated avoid malformations (Carew 2003). Combining individual frequency its crop, we estimate total moved each species (Gibson, 2012). load also reveals how much non-crop it carries. Since heterospecific reduce success (Arceo-Gómez Ashman, 2011), conspecific from plant's perspective (Ashman Arceo-Gómez, 2013). Resolving makeup loads may reveal individual-level specialisation within generalist (Lucas 2018; Somme 2015). Where individuals specialised species, apparently low-quality actually provide high-quality transport. Another difficulty managing insect community likely differ between farms. expected best based one site even present another site, reducing managers' ability apply recommendations their local context. Instead identifying specific taxa helpful identify parameters strongly related pollinator so managers promote match those parameters. This cheaper buying commercial bee colonies cases. larger 2021) hence, transport small insects. Larger tend fly longer distances (Greenleaf 2007), potentially allowing specialise spatially dispersed preferred plant species. Similarly, abundant active periods make briefly both groups generally pollen. Conversely, mixed (lower-quality) loads, unless level during shorter time periods. In study, use strawberry—an economically morphology—as model system answer following three questions: (1) Does estimated importance when pollen-load data? As had long flowering season our study sites, fortnight over whole test whether approaches highlight congruent sets (2) Are there predict importance? Again, consider pollen-based visit-based metrics short-term long-term estimates. (3) How do plants? indicates strawberry alone sustain populations pollinators. We sampled farms (8–19 ha size, cultivar: Elsanta) Somerset County, South-West England (hereafter A, B C). Strawberries were grown open fields open-sided polytunnels rotation new into ripe berries harvested On farm, established 5 (farms B, C) 10 (farm A) randomly positioned 25 × 2 m transects field (including tunnels) same number margins (natural semi-natural grasses, wildflowers, hedgerows). means A was twice intensively two All observed visiting any collected transect walks conducted fine days every ~7–10 March 16–September 29, 2009. (strawberry non-strawberry) habitat (field margin) recorded. Insects identified swabbed pollen, non-strawberry comparison reference library. further sampling details, see Supplementary S2.1. Commercial hives Bombus terrestris placed supplement pollination; native area locally common, B. (Martin 2019). Samplings done approval land owners “Acknowledgements” section) no permits required invertebrate collecting area. did listed extracted morphological traits (assumed constant across farms) literature abundance period (separately farm). (morphological community-based) defined below. details inter-tegument distance (ITD) period, S2.2. exploratory tool, calculated correlations tested parameter component using series Kendall's rank correlation tests (Supplementary S3.1). To approximate size movement range, used ITD: wing bases. obtained mean ITD values British Baldock al. (2019) Hackett (2019). could measured Coleoptera, removed these main analyses. Repeating analyses length width elytra Coleoptera (taken 2019) substitutes qualitatively affect conclusions S3.2). define species' annual degree year, focal farm. Species flight might course year but specialist point year. We, average weekly (number event) insect, abundance. (not flower) included. events if few available, regardless season. uniqueness pollinator's Bray–Curtis dissimilarity vector proportions made vectors above, fit structural equation models (SEMs; PI VI). model, fidelity (pQ pF vQ vF), fidelity, turn, affected (PI include direct paths explicitly terms Equations 1 2). multi-step pathway allows us through quantity, both, overall ultimately importance. linking generalised multilevel path Shipley (2009), fitting separate SEM. SEM included sub-models, effect affecting Model implemented R (R Core Team, packages PIECEWISESEM (Lefcheck, lme4 (Bates d-separation estimate. estimates Fisher's C statistic performs χ2 (2009). Note that, test, p-value greater 0.05 acceptable data. After whole-season importance, repeated procedure See S3.3 full details. unlikely meet nutritional needs single nectar indicate resources fill temporal gap provided (Filipiak, establish depend beyond infer assessed extent depending captured. More specifically, calculate (a) share visitors (b) percentage carrying (indicating habitats) (c) percentages caught habitat. cross-habitat transport, standard error proportion up measure separately margin plants. Across flies among highest (based transport; Table 1). Although ranks varied farms, hoverfly Eristalis arbustorum, honeybee Apis mellifera, bumblebees (Bombus lapidarius pratorum) importances site. Notably, most-important farm—not managed present. Visit-based lower differences (Table A. mellifera 10th terresstris 19th despite being loads. Annual strongly, significantly, positively correlated (Figure 2c; β = 0.772, p < 0.001; only non-zero singularity). Rank (τ 0.533, z 8.42, 0.001). Short-term associated 2d; 0.978, 0.001 linear relating VI; τ 0.446, 13.7, S3). Finally, correlated, association stronger S3; Figure 2a,b). models, significant effects confirms S3.4). good (p > 0.5). Both increased decreased 3a,c; 0.4, 0.001, R2 0.45; 0.3, 0.45 −0.1, 0.03, pQ; 0.5, 0.64; 0.8, 0.64 −0.2, 0.002, vQ). conversely, degrees unique (β −0.5, 0.02, 0.29 pF; −0.8, 0.005, 0.51 vF). S3.4); remaining considered 3b,d; 0.26 pQ 0.2, 0.03). Likewise, 0.19; −0.6, 0.19 0.26; Pollen Strawberry extensive needs. demonstrated strong links habitats continually available. Most (65%–86%) shared 4a), (62%–71% 4b; S8, S3.5). Pollinators captured (75%–85% insects) somewhat less (54%–57% insects). dominated group 4c). true least-important high low pollinator-dependent examining roles find agree measures precise order (Apis Bombus) (Eristalis) emerged (E. arbustorum) sampled, providing evidence hoverflies (Rader 2020; Tiusanen 2016). Bombus. Moving identities, found degree, abundance, (among parameters) suggests hold promise ‘rule-of-thumb’ indicators flowers, emphasising need areas field. There currently debate flower-visitor provision (Kortsch 2023; Rader dataset, 0.649) correlated. result UK, offer reasonable proxies Other studies come similar contribution visitor note ranked metrics. similar, discrepancy due carried: large, amounts making inferences represent currency interest: realised plant. From perspective, neglect components pollination, (Popic ideal chain resulting should measuring sequence seed set 2022; Corbet, 1998; Naeem, Rodger Stavert previous multiple whenever feasible Encouragingly, steps (flower flowers) offered coherent ranking Visitation particularly useful limited. fortnightly arbitrary short snapshot long-flowering like UK. targeting peak activity accurately (Hegland 2010), unknown, dealing suggest longer-term accurate Similar studies, tended (Ellis 2017). Long (which imply non-unique periods) moving brief periods). generalist. generalism requirement variety 2022), fidelity. Despite possibility learn abundant, long-lasting resource consistently (Amaya-Márquez, 2009), beneficial detrimental ‘ideal’ partly explains why predicting context, general, predicted throughout One exception consistency associations ITD. (as expected, Cullen visits. Instead, negative visited) without corresponding (since visits). lack relationship spatial context examined. Flight limiting factor close together. systems, insects' seek out influence de facto 2023). highly strawberry. cannot needs, consistent sucrose protein content Elsanta (Ahrenfeldt dependency supplemental showing access sources supply floral (Timberlake orders, commercially added elsewhere). Given variability wise diverse including bumblebees, hoverflies, honeybees, rather focus diversity foraging behaviours reliable (Abrol prevent malformation (Chagnon 1993), increase yields (Stewart 2017) ‘insurance’ against losses taxon. uses county diversity-focused approach improving insect-pollinated morphology. While elsewhere rankings sample sites. Indeed, reason explore importance: (for example) everywhere, requires maintaining near support life stages gaps resources. benefit preservation nesting sites tussocks steep banks (Kells Goulson, 2003) require ponds larvae Providing resources, relatively rare, ensures food (Baldock periods, presence sometimes decrease mixing (Ye 2014), mostly diets still increasing ample compensate loss individual. additional directly adding bumblebee placed. Note, inference draw hives. Our case has implications estimating service. First, farmer seeking intervention thus rule-of-thumb conclusions. Such pollen-transport accounted for). Second, indications Rather exploring taxon strawberry-farming region, long-flying cultivation systems (i.e. cropping varieties methods produce period). Managers can, target ensuring Timberlake alternative species-specific measures. Third, findings considering landscape decisions. Diverse requirements underline difficulties living alone. Overall, hope improved what of, guidelines conserving valuable Alyssa R. Cirtwill, Edith Villa-Galaviz, Rachel Gibson, Thomas Timberlake, Tomas Roslin Jane Memmott conceived ideas; Gibson data; Villa-Galaviz Cirtwill analysed led writing manuscript. authors contributed critically drafts gave final publication. would thank sites: Richard, Philip Adrian Winter Team Green Growers, Jan Butterley Nynehead Fruit. grateful Mark Pavett, John Deeming, Brian Levey Mike Wilson National Museum Wales, Cardiff identifications. members Spatial Foodweb Ecology Group comments work supported Natural Environment Research Council (NERC) (NE/T013621/1), Science Foundation (NSF) Academy Finland (AKA) (grant 334787), coordinated Belmont Forum Climate, Health Collaborative Action (proposal 1550). Further Bristol Centre Agricultural Innovation. T.R. funded European Synergy Grant 856506 (LIFEPLAN), 322266), Career Support grant Swedish University Sciences. A.C. postdoctoral fellowship (1332999). declare conflict interest. peer review history article https://www.webofscience.com/api/gateway/wos/peer-review/10.1002/2688-8319.12253. Data Dryad Digital Repository: https://doi.org/10.5061/dryad.2rbnzs7t2 (Villa-Galaviz S1. investigate (top) (bottom). Pollinator loads) then averaged S2. (above diagonal) p-values (below eight Coleoptera; before calculating correlations. moderately correction testing performed; therefore illustration only. S3. Kendall relationships (p) (v) (Q) (F) S4. S5. Coefficients incorporating Coleoptera. Width: Length: elytra. S6. S7. changes last S8. quantify crops. include: sharing (crop other. S9. List Hymenoptera, Lepidoptera recorded S10. Diptera Please note: publisher responsible functionality supporting supplied authors. Any queries (other missing content) directed author article.