Populations of insect pollinators are declining significantly around the world (Dirzo et al 2014; Powney 2019). A range of factors including climate change, pesticide use, and land use/cover change impact the diversity and abundance of pollinators and thus the availability of services that they provide to humans (Wagner 2020). The effects of a ‘pollination crisis’ are particularly dire for farming communities whose livelihoods depend on the cultivation of pollinator-dependent crops such as apples (Pardo and Borges 2020; Dicks et al. 2021). In this project we are investigating the nature of this urgent and complex issue in the Mukteshwar region of the Uttarakhand Himalayas in northern India, where farming communities have long depended on apples and other pollinator-dependent crops for their livelihoods. Apple production in this region has declined over the past few decades, and several studies have suggested that pollinator scarcity and climate change may be to blame (Partap and Partap 2002; Basannagari and Kala 2013). However, there is a lack of empirical data on the current status of pollinator communities in the region. To interrogate this issue, we assembled a multi-disciplinary team of researchers from India and the United States: Ginger Allington (landscape ecology), Kiran Cunningham (applied anthropology), Ann Fraser (entomology), Aman Luthra (economic geography), Shweta Rana (environmental science), Preeti Virkar (entomology), and a diverse team of researchers with a range of interests and expertise from the Center for Ecology Development and Research (CEDAR), a non-governmental research organization based in Uttarakhand (Anvita Pandey, Narendra Raikwal, Anmol Ratna, Ankita Rawat, Vishal Singh and Renu Suyal).
We had three overarching goals for this fieldwork: (a) to engage farmers as citizen scientists in documenting pollinator diversity and abundance in their orchards; (b) to assess the impacts of landscape composition and configuration on pollinator diversity and abundance; and (c) to determine the extent of pollination limitation in apples in this region. The groundwork for this project began with two field visits by a subset of team members in July and December 2019. We interviewed twenty farmers and other stakeholders about production declines and tested out a method to involve farmers in documenting the diversity of pollinators on their farms. After a COVID-induced three-year hiatus, we restarted this project in November 2022. Four CEDAR team members conducted a rapid survey of thirty farms, from which we selected fifteen to participate in this study. We selected farms across a range of altitudes, slopes and farm sizes. Altitudes of the selected farms ranged from 1680 to 2360 meters above sea level. Farm sizes ranged from less than one acre to five acres and had between 50 to 1200 individual apple trees in their orchards. Farms typically grow multiple other subsistence and market crops within and adjacent to orchards.
Citizen science, a rapidly growing research method, broadly refers to projects where members of the public collect, categorize, transcribe, or analyze data (Shirk and Bonney 2015). In addition to reducing costs of data collection, citizen science efforts can also help to develop a shared ecological understanding, co-generated by participants and researchers (Dunkley 2019). This might be particularly important for research involving pollinators because some evidence suggests that farmers around the world, including in Uttarakhand, undervalue the pollination services provided by wild pollinators, despite the potential huge benefits those organisms have for on-farm productivity (Partap and Partap 2002). We trained participating farmers to use the pan trap (or ‘bee bowl’) sample method to capture pollinators. This is a commonly used, simple method to document pollinator diversity and abundance at a site. Farmers in other parts of the world have been successfully trained as citizen scientists to conduct pan trap sampling (Garratt et al 2019).
We developed a protocol for conducting pan trap sampling, trained each farmer on the methods, and provided them with all of the supplies that they needed for the sampling and storage of the bees. We developed a ‘farmer diary’ that describes and illustrates the protocol for conducting the experiment, and also contains data sheets for each sampling date where farmers document sampling information, weather conditions, and the dominant phenophase of apple trees in their orchard. Here are images of a few pages from the diary showing visual instructions for conducting pan trap sampling.
In early February 2023, four researchers from the CEDAR team visited the field sites to conduct the first pan trap sampling and train the farmers in the process. Once trained, farmers were asked to conduct pan trap sampling on their own at two-week intervals for a total of five instances between mid-February and mid-April. Each sampling exercise required around two hours of their time for which we compensated them five thousand Indian Rupees (approximately sixty US Dollars). In this image, Renu Suyal (l) and Ankita Rawat (r) are training one of the fifteen farmers on how to set out the pan traps in his orchard. [Photo credit: Akshat Pratap]
In late March, 2023, the entire research team visited the sites, collected pan trap samples from the farmers, and engaged in informal conversations with farmers about their experience participating in the process, and about pollinators on their farm more generally. Here you can see Renu Suyal, Narendra Raikwal and Aman Luthra talking to a farmer about the state of pollinators in the region while drinking cups of delicious rhododendron squash (Photo credit: Ankita Rawat).
Aman Luthra and Shweta Rana talk to a different farmer. Conversations with farmers were never limited to pollinators alone. Here, the farmer is telling us about his involvement in local politics in his younger years (Photo credit: Narendra Raikwal).
In general, we found farmers to be highly engaged in the process. Most followed the collection protocol carefully. In this image, Renu Suyal and Aman Luthra are checking the samples that a farmer had collected. Ankita Rawat meanwhile is checking to make sure that the information in the diary was complete and entered correctly (Photo credit: Ginger Allington).
During our March field visit, the weather often did not cooperate. It was rainy approximately fifty percent of the time, and we were not able to visit farms on those days. However, on the days we were forced to stay in our hotel, we used the time productively to get a jump start on processing and analyzing the samples of bees collected from each farm. The entomologists on our team (Ann Fraser and Preeti Virkar) trained the rest of the team to dry and pin the bees into organized boxes, which allowed them to begin identifying them. Here, Ann Fraser teaches the rest of the team where and how to pin the bees. (Photo credit: Ginger Allington)
The common spaces in the hotel were turned into a makeshift lab. In this image, Renu Suyal, Shweta Rana, and Kiran Cunningham pin the dried bees as Ginger Allington maps land use/cover (Photo credit: Ann Fraser).
Once bees had been pinned, the next step was to identify them to the genus level. Here, Preeti Virkar identifies bees (Photo credit: Russ Rhoads).
Ann Fraser identifies the bees, Kiran Cunningham labels them with the genus ID, and Aman Luthra enters the data into a spreadsheet (Photo credit: Shweta Rana).
The relative abundance of the different bee genera in the samples analyzed thus far. As a result of our collective effort, we were able to process almost all of the samples while we were still in the field. Thus far, a total of 1,396 individuals have been identified, representing thirteen bee genera. In addition to bees, the samples also contained syrphid flies, butterflies, moths and wasps.
During a field visit at the end of April 2023, we shared preliminary results with the farmers, engaging them in discussion about what we are doing with the specimens they collected, what an analysis of the data will show, and how that analysis might be helpful to them. This exercise demonstrated that we are committed to collaborating with the farmers to understand the state of pollinators on their farms. The image on the left shows all the bees collected from a single farm and on the right is a sample of the report that we will share with the farmers that compares what we found on their farm to data from all farms combined (Photo credit: Kiran Cunningham). Pollinator communities are affected by landscape composition and configuration at different spatial scales (Betts et al 2019). Habitat loss and fragmentation, landscape heterogeneity and land use intensification affect pollinator communities in various ways (Senapathi et al 2017). In our study area, landholdings are typically small and the mountainous terrain consists of a mosaic of land cover that could, at least in theory, support a rich and diverse pollinator community. But land use and land cover in the area has been changing dramatically over the past few decades due to various politico-economic pressures (Basannagari and Kala 2017). Out-migration of younger people from the area to cities in search of better livelihood opportunities has led to farm abandonment and the sale of land to property developers for a rapidly expanding vacation home real estate market. All of these changes in the landscape have the potential to impact pollinator populations (Senapathi et al 2017), although the nature and extent of these impacts have not been explicitly examined in this region.
Example of the heavily terraced hillsides common in the region (Photo credit: Anmol Ratna)
Farms in the region are typically small and comprised of a mix of row crops (such as pulses, mustards and potatoes) planted in and around the fruit trees (Photo credit: Ginger Allington)
To better understand the impact of landscape structure on pollinator diversity and abundance, our team of geospatial researchers—Ginger Allington, Anmol Ratna and Ankita Rawat—mapped land use/cover (LULC) within a 500-meter radius around each farm. Using Google Earth imagery as a starting point, the team developed LULC classifications appropriate for this context. During our site visit, we fine-tuned the classification system and ground-verified the actual LULC surrounding each farm. These data will be used to help explain variation in pollinator abundance and diversity across the farms. Although the problem of pollination scarcity in Himalayan agricultural landscapes has been suggested, the degree of pollination limitation remains largely unexamined. Further, there is likely significant spatial variation in the degree to which pollinators are limiting crop production in heterogeneous mountainous landscapes (Garcia-Camacho and Totland 2009). Thus, to better understand the extent of pollination limitation in apples, a highly pollinator-dependent crop, we set up a pollination exclusion experiment in the fifteen orchards. Pollination exclusion experiments are used to determine the degree to which declines in fruit production are a result of inadequate (or insufficient) visitation by pollinators. We determine this by comparing fruit-set (the number of fruits that develop following pollination) on typical branches with those that have been bagged to prevent bees from visiting the flowers and those that were bagged for pollination by hand (Olhnuud et al 2022).
Preeti Virkar trains the rest of the team on how to bag and tag inflorescences and collect data on the number of buds in each inflorescence (Photo credit: Ann Fraser).
In each orchard, we selected ten apple trees. On each tree, we identified five inflorescences in the bud stage. Each inflorescence (group of buds) was assigned a treatment and tagged accordingly. Two inflorescences on each tree were bagged with a fine mesh bag to exclude pollinators entirely; one inflorescence was bagged so that it would only be hand pollinated once flowers were in bloom; one inflorescence was left open to be pollinated by bees; and one inflorescence would be both open and hand pollinated. Differences in fruit-set across the treatments will allow us to infer the extent to which apples are pollination-limited (Photo credit: Ginger Allington).
Here Anmol Ratna, Ginger Allington, and Narendra Raikwal are identifying, tagging, and bagging inflorescences on a tree (Photo credit: Rajendra Singh).
We implemented the experimental design described above, but nature does not always comply with a meticulously designed plan. An unseasonably cold spell delayed flowering and rainy days prohibited site visits. When we finally had sunny days and were ready to pollinate the two hand pollination treatments on each tree, we found some inflorescences still in bud stage, while others were too late to be hand pollinated. Even within a specific inflorescence, some flowers had not even bloomed yet, while others had already shed their petals. In the photo, one bloom was ready to be pollinated but the others were still at the bud stage. Inset photo shows a close-up of the male and female parts of an apple blossom; pollen is apparent on the anthers and the tips of the stigmas are extended and ready to receive pollen (Photo credit: Russ Rhoads and Ginger Allington).
Nature’s caprice demanded research agility. Where possible, we shifted treatments and identified new inflorescences that were at the proper stage. In some cases, especially where entire trees had not yet bloomed, we were not able to shift treatments to new inflorescences. Despite these challenges, we hope to gather sufficient data to be able to discern the extent of pollination limitation in apples in the region. Here, (l-r) Ann Fraser, Shweta Rana, and Preeti Virkar can be seen hand pollinating inflorescences that were at the correct stage. We returned to the field sites at the end of April 2023 to collect data on fruit-set across the treatments. A preliminary analysis of fruit set data shows significant pollination limitation across farms. At harvest time in July 2023, we plan to return to the field to collect data on seed set within each fruit to fully comprehend the impact of pollination limitation on fruit development. Combined with a deeper understanding of pollinator diversity and abundance, this should help farmers find ways to address the problem of pollination scarcity on their farms (Photo credits: Renu Suyal, Aman Luthra and Ginger Allington). Despite the numerous challenges we faced in the field, our diverse team pulled together and cooperated on all of the tasks—anthropologists pinned bees, and entomologists interviewed farmers. Our experience demonstrates not just how important it is to have the right team composed of diverse backgrounds, interests, and expertise in order to address a complex and urgent socio-environmental problem, but also the importance of team members who are willing to learn from each other, and band together. Perhaps even more importantly, our experience demonstrates how important it is, and how rewarding it can be, to engage research participants in the process of knowledge production. Driven by an action research framework, this project hopes to co-produce knowledge that is useful and actionable.
Farmers, such as one in this photo, were not only intellectually curious but also incredibly generous. This farmer decided to bring a pot of tea and biscuits into the orchard as Ankita Rawat and Ann Fraser determine which inflorescences to hand pollinate on this tree (Photo credit: Kiran Cunningham).