Unveiling the benefits and gaps of wild pollinators on nutrition and income

Insect pollinators, including bees, wasps, flies, and butterflies, play an essential role in crop production worldwide. Studies have shown that the presence of insect pollinators in crop fields can increase fruit and seed production, improve fruit quality, and enhance the nutritional value of crops (Klein et al 2007, IPBES 2016). For instance, blueberry and canola fields in Canada with high densities of bees have higher fruit/seed set and yields (Sabbahi et al 2005, Desjardins and Oliveira 2006, Button and Elle 2014). In fact, it has been estimated that more than three-quarters of the main global food crops rely to some extent on pollinator species (IPBES 2016). Hence, safeguarding their populations is crucial to maintain and improve agricultural production in addition to farmer livelihoods.

In some cases, native pollinators can be more effective at pollinating some crops than non-native species (Garibaldi et al 2013). For example, bumble bee species can 'buzz' pollinate flowers of numerous crop species, significantly increasing fruit set and weight compared to European honeybees (Cooley and Vallejo-Marín 2021). Also, the presence of diverse communities of pollinator is more likely to generate yield stability and avoid production declines than less diverse communities, since pollinator species differ in their foraging behaviour, activity patterns and responses to changing conditions (Garibaldi et al 2011, 2014). However, both the richness and visitation rates of pollinator species are inversely related to the distance between crops and the natural or semi-natural ecosystems that provide nesting habitat (Ricketts et al 2008). Therefore, farmers should maintain and restore a variety of natural habitats near croplands to promote pollination services and increase the resilience of agricultural landscapes (Duarte et al 2018, 2020).

Despite their clear advantages for food production, wild pollinator populations have been declining worldwide, especially due to habitat loss, pesticide use, climate change, and pathogens (IPBES 2016). Global attention is now focused on the best way to halt this decline and guarantee their protection to help sustain many farmers' livelihoods (Dicks et al 2016). As the level of dependency on animal pollination varies among crop types, regional agricultural economies will differ in their susceptibility to wild pollinator loss. For instance, Canada has several crops and a vast expanse of cultivated lands that depend to some extent on animal pollinators (Richards and Kevan 2002, Agriculture and Agri-Food Canada 2022a). However, because no national-scale analysis of pollination services has been completed for Canada, any pollination-friendly initiative or policy will lack the knowledge to target actions to most effectively maintain or improve pollination services.

Identifying sites where the contribution of wild pollinators is highest or that have the greatest potential to sustainably increase natural pollination and agricultural production is critical for prioritizing conservation and restoration efforts towards food security and biodiversity protection. Previous studies have combined biophysical and socio-economic data to assess pollination services at global or local scales (Lautenbach et al 2012, Koh et al 2016, Chaplin-Kramer et al 2019), while others have evaluated multiple ecosystem services at the national level but excluded pollination (Mitchell et al 2021). However, effective decision-making requires up-to-date and context-specific evidence-based information. Building on global models (Chaplin-Kramer et al 2019, 2023), our paper aims to understand how much wild pollinators in Canada potentially contribute to people in terms of nutritional values and farmers' income and provide insight into the spatial distribution of these pollination benefits.

We estimated the potential benefits of wild pollinators in agricultural landscapes across Canada at 240 m spatial resolution (data available at Duarte et al 2024). To do this, we used publicly available data on crop type locations, yields, nutrient content, and farm gate values, as well as the location of natural habitats. We considered demand sites for wild pollinator services to be the croplands that rely on pollinator species to provide nutritional and economic benefits to people. Hence, service providing areas are the nearby natural areas that serve as habitat for these pollinator species. Our approach followed the methods of Chaplin-Kramer et al (2019), with modifications for Canadian landscapes, data, and to calculate economic benefits.

2.1. Land use/cover data

2.2. Crop pollination dependency

2.3. Pollination sufficiency

2.4. Crop production

2.5. Refuse fraction, nutrient content and dietary requirements

2.6. Pollinator-dependent crop production and equivalent people fed

We multiplied all factors to calculate wild pollinator-dependent nutrient production (WPN) for each crop in each pixel, following equation (1):

$$\begin{align}& {\text{WP}}{{\text{N}}_{in}} \nonumber\\ & \quad = \sum\limits_c^C \left( {{A_{ci}}{\ast}{Y_c}{{\ast}}\left( {1 - {\text{R}}{{\text{F}}_c}} \right)\ast{\text{N}}{{\text{C}}_{nc}}\ast{\text{PD}}{{\text{F}}_c}\ast{\text{WP}}{{\text{S}}_i}} \right)\end{align} \tag{ 1 }$$

where A is the area (ha) occupied by the crop type 'c' in pixel 'ï'; Y is the potential yield of crop 'c', which was defined per Canadian province.; RF is the refuse fraction; NC is the nutrient content for nutrient 'n' in crop 'c'; PDF is the pollination dependency fraction for crop 'c'; and WPS is the wild pollination sufficiency of pixel 'ï'. We obtained the total wild pollinator-dependent nutrient production by summing the values across all crops for each pixel and for each nutrient.

These layers of nutritional benefits are expressed in different units of nutrient production (i.e. Kj, RAE, DFE). Therefore, we first modified these values by dividing each layer by the recommended annual dietary intake. Table S6 presents the aggregated sum of the results obtained in this step per nutrient and province. Then, to summarize the results into a single metric, we averaged the values in each pixel to obtain the equivalent number of people fed through pollination. This metric represents the average number of people whose nutritional needs could be met by wild pollinator-dependent crop production across various essential nutrients (Chaplin-Kramer et al 2019, 2023). We acknowledge that for some crops, a significant portion of its production is allocated for industrial use or as feed grain. However, information on this portion per crop is lacking at a national-scale for Canada. Also, assessing all potential benefits of wild pollinators for crops directed toward livestock feed would necessitate consideration of animal-related products like meat and dairy, which fall beyond the scope of our current study. In choosing the metric of equivalent people fed, we aimed to provide an estimation of the potential benefits that wild pollinators can offer, that is aligned with international reports (such as IPBES 2016) and global studies (Chaplin-Kramer et al 2019, 2023).

2.7. Farm gate values

Farm gate value of crop production refers to the net value of a farm's production as it leaves for the market, after selling costs have been subtracted. Using data from a variety of governmental sources (Statistics Canada 2022a, 2022b, 2022c), we estimated average farm gate values per metric tonne of each crop for each province between 2015–2021. Before averaging, all values were corrected for inflation using Consumer Prices Index, and 2019 as baseline (OECD 2022). We then calculated total farmer income related to wild pollinator-dependent crop production in pixel i (WPV_i ) using equation (2):

$$\begin{align}{\text{WP}}{{\text{V}}_i} = \sum\limits_c^C \left( {{A_{ci}}\ast{Y_c}\ast{\text{FG}}{{\text{V}}_c}\ast{\text{PD}}{{\text{F}}_c}\ast{\text{WP}}{{\text{S}}_i}} \right)\end{align} \tag{ 2 }$$

where FGV is the farm gate value per metric tonne of crop 'c'.

2.8. Mapping pollination benefits back to natural habitats

We mapped both the equivalent number of people fed and farmer income back to the surrounding pollinator habitat. We used a distance decay function (equation (3)), considering the distance between each focal crop pixel 'j' and the surrounding natural habitat 'I' to weight the contribution of each habitat pixel in creating the benefit (Chaplin-Kramer et al 2023). This weighted distance (Wdist_i ) considers only habitats within the same flying distance as the pollination sufficiency (dist_max of ∼2 km) and was calculated following the habitat quality model of InVEST (Sharp et al 2020). We assumed that more distant habitats would have a lower weight because fewer wild pollinators from these habitats would reach the cultivated areas

$$\begin{align}{\text{Wdis}}{{\text{t}}_i} = \text{exp} \left( { - \left( {\frac{{2.99}}{{{\text{dis}}{{\text{t}}_{\text{max} }}}}} \right)\ast{\text{dis}}{{\text{t}}_{ij}}} \right).\end{align} \tag{ 3 }$$

Along with distance, we factored in the area of natural habitat surrounding each focal crop pixel to distribute a portion of the pollination benefit back to these habitats. That is, we assumed that each natural habitat within the surrounding landscape contributes to a fraction of the benefit provision. This fraction is directly proportional to the habitat size and inversely proportional to its distance from the focal crop pixel. Therefore, we use equation (4) to calculate the weighted benefit fraction for each focal pixel with cropland (Bweighted_j ). Finally, using equation (5), we calculated the total benefit value of each pixel with natural land (Bnat), which is now the focal pixel, by considering all surrounding crop pixels and their weighted benefit fractions. Hence, we mapped the benefits back to the natural habitat accounting for the distance and area that it occupies

$$\begin{align}{\text{Bweighte}}{{\text{d}}_j} & = \frac{{{\text{Bcro}}{{\text{p}}_j}}}{{\sum\limits_i^N\left( {{\text{NA}}{{\text{T}}_i}\ast{\text{Wdis}}{{\text{t}}_i}} \right)}}\end{align} \tag{ 4 }$$

$$\begin{align}{\text{Bna}}{{\text{t}}_j} & = {\text{NA}}{{\text{T}}_j}\ast\left( {\sum\nolimits_i^N \left( {{\text{Bweighte}}{{\text{d}}_i}\ast{\text{Wdis}}{{\text{t}}_i}} \right)} \right).\end{align} \tag{ 5 }$$

In these equations, 'j' is the focal pixel with cropland in equation (4) and natural area in equation (5); 'I' corresponds to the neighbour pixels with natural areas in equation (4) and croplands in equation (5); Bcrop corresponds to the benefit value of a pixel with cropland, derived from the wild pollinator-dependent crop production; NAT is the natural area in a neighbouring pixel; Wdist is the weighted distance between the focal and neighbour pixels; Bweighted is the weighted benefit fraction; and Bnat would be the equivalent people fed value or farmers income (in CAD$) values of a natural area.

2.9. Benefit gap

2.10. Data aggregation into ecodistricts

The estimated number of equivalent people fed each year in Canada that depends on wild pollinators is 24.4 million. Of this total, pollinator-dependent crop production in Saskatchewan feeds the most equivalent people (6.7 million), followed by Ontario (5.8 million) and Alberta (5.0 million). Additionally, almost CAD$2.8 billion is generated each year as income for farmers due to the presence of wild pollinators in Canada. Saskatchewan ($922 M), Alberta ($693 M) and Manitoba ($386 M) are the provinces with the highest income value associated to wild pollinators and their natural habitats. On the other hand, lack of nearby pollinator habitat and pollination to dependent crops nationally leads to significant benefit gaps that if addressed could help generate an additional nearly 30 million equivalent people fed and CAD$ 3 billion in farmer income every year. Saskatchewan has the highest potential to increase benefits from pollinators for both people fed (11.5 million people) and income ($1.6B) benefit gaps. Table 1 summarizes these results per province and figure S1 show their location.

Table 1. Benefit values and gaps per province.

Province	Equivalent people fed (thousand)		Farmer's income (CAD$ million)		Service providing areas (million hectares)	Pollinator-dependent crop area (million hectares)
	Benefit	Gap	Benefit	Gap
Newfoundland and Labrador	0.0	0.0	0.1	0.0	0.01	$66 \times {1^{ - 6}}$
Prince Edward Island	110.5	12.6	12.1	0.7	0.28	0.03
Nova Scotia	36.5	8.0	23.4	2.0	0.86	0.02
New Brunswick	35.7	0.6	22.2	0.0	0.83	0.02
Quebec	2400.2	1529.5	234.1	61.3	3.22	0.50
Ontario	5838.0	7090.4	249.2	298.4	2.84	1.44
Manitoba	4060.7	5473.3	385.6	476.8	3.10	2.35
Saskatchewan	6689.9	11 490.9	922.3	1,574.0	6.95	7.18
Alberta	5032.3	4306.6	693.5	596.5	8.89	3.73
British Columbia	178.9	67.5	244.2	79.3	1.08	0.11
TOTAL	24 382.8	29 979.3	2786.7	3089.0	28.07	15.37

In 2019, Canada had 45.77 million hectares of agricultural land, encompassing crops, pastures, fallow areas, and related land uses. Our analysis focused on 34.86 million hectares of cropland, accounting for 76% of the total agricultural area. We examined data for all 38 crop types present in the LULC data, of which 17 rely on animal pollinators to some extent, covering an area of approximately 15.4 million hectares (figure S2(a)). Canola/rapeseed, soybeans, and peas are the major pollinator-dependent crops, occupying 10, 2.6, and 1.9 million hectares respectively. The pollinator-dependent crops are distributed in 382 ecodistricts of the total 417 analysed, with 158 ecodistricts having more than half of their crop area relying on pollinators (figure 1(a)).

Zoom In Zoom Out Reset image size

There are 84 ecodistricts that rank in the top 25% when it comes to the production per hectare of natural habitat for both equivalent people fed and farmer income benefits. The 6.6 million hectares of natural areas providing pollinator services within these 84 ecodistricts account for 60% of the national total number of equivalent people fed and 55% of the total farmer's income. Thus, they are critical natural habitats that could be low-hanging fruits for conservation actions, as they provide, on average, high levels of pollination services in terms of nutrition and economic values. The locations of these ecodistricts are shown in light-bright-green in figure 2(a), which are concentrated in the central portion of the Prairies, southern Ontario, and the Saint Lawrence lowlands in Quebec.

Zoom In Zoom Out Reset image size

We also identified possible targeted locations for restoration actions to increase pollinator habitat and benefits from pollination. They correspond to the 68 ecodistricts among the top 25% with the highest scores for both benefit gaps. These ecodistricts contain 7.6 million hectares of pollinator-dependent crops with insufficient wild pollination, which account for 70% of the national nutritional gap and 66% of the national economic gap. Their location is shown in black in figure 2(b). Additionally, there are 50 ecodistricts that are priority for both conservation and restoration actions to enhance pollinator services, and they are located mostly in Saskatchewan and southern Manitoba, represented with a light-pink outline in figure S3.

Our findings indicate that forests contribute to the largest share of both benefits among different habitat types (table 2). They help feed the equivalent of 12.4 million people and generate CAD$1.3 billion per year for farmers. Their service providing areas extend across 14.5 million hectares, primarily surrounding croplands in southern Ontario and Quebec, known for their high yield values, but also across the Prairie region (figure 1(b)). Grasslands and wetlands also play crucial roles in providing pollination services in Canada (table 2). Together, they help feed nearly 10 million people and generate CAD$1.3 billion for farmers, with a significant impact in the Prairie region (figure 1(b)).

Our study highlights the crucial role that wild pollinators play in food security and sustaining farmers' income in Canada. To provide context, the estimated equivalent number of people with nutritional needs met by these pollinators exceeds half of Canada's population (Statistics Canada 2019). Although, in reality, a significant portion of crop production is allocated for export, industrial use, or as feed grains, wild pollinators benefit a far greater number of people than estimated here by partially contributing to their nutritional needs, as an average person's diet contains multiple sources of nutrients and energy. Also, pollinator-dependent crops significantly contribute to a healthy human diet, as they represent a substantial portion of the worldwide availability of several essential micronutrients (Eilers et al 2011). Moreover, our estimated economic contributions from pollinators account for 8.5% of Canada-wide farmer revenues derived from the sale of principal field crops (including both pollinator-dependent and non-pollinator-dependent crops; Agriculture and Agri-Food Canada 2022b), and 5% of the total crop-related income (including more minor crops) estimated in our study. These findings and proportions align with the expected global decline in crop production in the absence of animal pollination, which is between 3% and 8% (Aizen et al 2009).

Our estimates demonstrate that Canada stands to gain significantly by promoting the protection of wild pollinators. Financial mechanisms and environmental programs that encourage cooperation among farmers could support this transition (Goldman and Tallis 2009, Ayambire et al 2021), as restoring natural habitats often benefits crop production for multiple landowners throughout the surrounding landscape and contributes to the establishment of a landscape structure capable of sustaining pollinator services (Duarte et al 2018, 2020). In our study, we identified specific locations where agri-environmental programs could prioritize investments by evaluating ecodistricts that provide more than 55% of the national potential conservation benefits and more than 66% of the national potential restoration benefits (respectively, ecodistricts in light-bright-green in figure 2(a) and in black in figure 2(b)). Within each of these ecodistricsts, the fine-scale results produced in this work (in 240 m resolution) could be used to inform participatory planning derived from these programs (figures S2 and S4).

Investments in wild pollinator habitat also hold the potential to serve as climate solutions. Our findings highlight the critical role of Canadian forest ecosystems in providing pollination services, indicating a potential win–win scenario where conserving and restoring pollinator habitats can simultaneously enhance carbon storage in plants biomass (Drever et al 2021). This aspect could also further increase the eligibility of farmers for funding through the recently introduced federal Agricultural Climate Solutions program (Agriculture and Agri-Food Canada 2021). This and other similar funding sources could help compensate restoration implementation costs to create new natural habitats for pollinators. However, this result stems from our approach of considering all habitats as equally suitable for pollinators. Research in temperate regions frequently indicates that open forests generally exhibit a higher abundance and/or richness of pollinators compared to closed forests (Hanula et al 2016). Therefore, it is essential to implement proper management practices to create or sustain these favourable habitat conditions, such as thinning, prescribed fire, and/or shrub removal (Hanula et al 2016), otherwise pollinator services could significantly diminish. Our sensitivity analyses demonstrate a potential 19% reduction in the total number of equivalent people fed and a 16% decrease in farmers' income if habitat suitability is compromised in forest areas (table S4), especially in southern regions of Ontario and Quebec. This also highlights the critical need for location- and species-specific pollinator data across Canada.

Furthermore, it is important to note that the other ecosystems evaluated here, often associated with both a high abundance of pollinators and habitat suitability, can also offer significant co-benefits. For instance, our analysis emphasizes the crucial role that grasslands and wetlands also play in promoting pollination services. Unfortunately, these ecosystems have experienced considerable historical loss and degradation in Canada (Federal, Provincial and Territorial Governments of Canada 2010). Reversing this trend can provide substantial benefits for people, encompassing not the ones demonstrated in this study, but also carbon sequestration, water regulation and cultural values (Saleh and Karwacki 1996, Acreman and Holden 2013, Pattison-Williams et al 2018, Drever et al 2021).

A significant challenge lies in ensuring food security for an expanding global population while simultaneously preserving biodiversity and its contributions to human well-being (Brussaard et al 2010, Chappell and LaValle 2011). Our work underscores ecodistricts with high percentage of croplands that dependent on pollinators (figure 1), and with high potential benefit provision (figure 2). In these ecodistricts, going beyond habitat conservation and restoration is crucial to avoid the loss of current pollinator communities while increasing agriculture productivity. A promising approach consists in promoting ecological intensification in these croplands through management practices that conserve biodiversity beneficial to agricultural production, such as intercropping, crop rotations, and on-farm diversification (Kovács-Hostyánszki et al 2017). Previous studies have demonstrated that this shift in management can result in no net loss or even improved yields, monetary and nutritional values of the produced crops (Pywell et al 2015, Garibaldi et al 2016).

Accordingly, our analysis reveals significant benefit gaps, suggesting a substantial opportunity for a positive and profitable shift in agricultural management practices with respect to pollinators in Canada. Strikingly, restoring pollinator habitat and increasing pollinator benefits to pollinator-dependent crops have the potential to provide more benefits than are currently realized in Canada, especially in Ontario, Manitoba, and Saskatchewan. The latter province warrants special attention, as it has the highest benefits provision and the highest benefit gaps, due to its large extent of pollinator-dependent crops (table 1) such as canola/rapeseed and peas. However, the best approaches for increasing pollinator habitat will likely also be region-specific across Canada, both due to the ecology and pollinator diversity, as well economic viability and physical constraints on-farm (Clearwater et al 2016). For example, in intensive agriculture areas like the Prairies, adding pollinator habitat might conflict with the use of large machinery. In high-value farmlands (e.g. southern Ontario) removing land from production to improve pollinator habitat could face financial constraints and require different incentives. Therefore, decision-makers should carefully evaluate the benefit gap estimates as their value could decrease if pollinator-dependent crops are restored to natural habitats.

Future refinements to our understanding of crop pollination across spatial scales are needed to better target conservation actions. By using a single flight distance and considering all natural areas as pollinator habitat, we simplify the diversity of pollinator species that occur in Canada (Agriculture and Agri-Food Canada 2014). This approach is necessary when conducting a nationwide modelling where species-specific data at this scale is largely lacking. We also did not consider the role of croplands as potential habitats for pollinator species, nor other local scale elements such as field margins and hedgerows. This could result in an overestimate of the value of natural habitat near croplands (Chaplin-Kramer et al 2019, 2023). Further fine scale analyses that build on our methods and results could modify flying distance and habitat preferences to correspond to the features of local and regional pollinator species (Sharp et al 2020). For instance, while our sensitivity analyses reveal small changes in the total benefit value with varying flight distances (table S3), they do indicate a potential significant increase in pollination services if local species required less habitat area in the surrounding landscape to achieve pollination sufficiency (table S4), especially in southern regions of Ontario and Quebec. As highlighted in Chaplin-Kramer (2019), modelling pollination at large scales involves a high degree of uncertainty. The relationship between habitat and pollinator abundance, as well as the translation from pollinator abundance to crop yield, introduces complexities and potential variability (Kremen et al 2004, Kremen 2005, Lonsdorf et al 2009, Kennedy et al 2013). However, past research has shown that, despite these uncertainties, a proportional measure of habitat area around farmland, as used in our approach, outperforms more complex models (Groff et al 2016). In addition, the contribution of wild pollinators is also affected by local management choices and techniques that are not easily identified in national scale analyses, including the cultivar planted, choice of pest control, and plant nutrient management (Pywell et al 2015, Garibaldi et al 2016, IPBES 2016). Lastly, there could be other unintended consequences when closing the benefit gaps that were not evaluated in this work, such as the appearance of pests due to the increased natural areas.

Our research highlights the pressing need to develop a national strategy aimed at safeguarding wild pollinators. Such a strategy has the potential to strengthen local economies and ensure the production of nutritionally essential food. Furthermore, our analysis can support the prioritization and implementation of conservation efforts aligned with Canada's commitments under the recently agreed Kunming-Montreal Global Biodiversity Framework (Convention on Biological Diversity 2022). Goal B of this framework emphasizes the sustainable use and management of biodiversity, as well as the recognition, maintenance, and enhancement of nature's contributions to people. By integrating practices that conserve pollinators with climate-smart agriculture and sustainable land management approaches, we can achieve several benefits, including enhanced biodiversity, improved crop production and resilience, and increased carbon storage within agricultural ecosystems.

We are very grateful for the support and precious discussions provided by the NCC's Conservation Policy and Planning team, and the M2L2 team. G T D was founded by Nature Conservancy of Canada and M G E M was supported by an NSERC Discovery Grant (RGPIN-2022-04567).

The data that support the findings of this study are openly available at the following URL/DOI: https://osf.io/4ey83/.