We found a total of 194 articles published between 1996 and 2023 that focused on pollinators or floral visitors found all, or in part, in Cameroon. Publications increase over time, with a rapid increase beginning around 2011. From 2001 onwards, researchers with Cameroonian affiliation typically accounted for more than half the publications from any given year (Fig. 1). Below, we describe the nature of this research and categorize it along various axes.
Authorship, Language, and Journal Legitimacy
Among the 194 documents retained, 98 were published in journals indexed by Scopus, 20 were in journals indexed by AJOL only, and 76 were in neither; of these one was a book, two were technical reports, while the rest were articles. 167 (86%) were in English while 27 (14%) were in French. Cameroonian researchers authored 133 of the articles and published in journals indexed in neither database 51% of the time, while international researchers authored 61 and published in journals indexed in neither database 13% of the time. A list of all documents can be found in Appendix 1, along with the databases where they were found and other classification categories.
Research methods, disciplines, and subject areas
Most studies, 101, employed both observational and experimental methods, with 74 being purely observational, two purely experimental, and 17 being reviews. 158 of the studies were primarily ecological in terms of the discipline, while 24 were better categorized as evolutionary, and 12 were chiefly social sciences. Among those with an ecological focus, most investigated plant-pollinator interactions (142), the remaining few were divided among the subjects of biodiversity (6), pollinator health and toxicology (5), pollinator life history (4), and ecosystems services (1). Ecological studies were led by both Cameroonian and international researchers. Among those with an evolutionary focus, taxonomy and systematics studies were best represented (13), followed by adaptation and speciation (7), and phylogenetics, phylogeography, and population genetics (4). Of these, all but two were led by international researchers. Social science studies focused most on the economics and political ecology of beekeeping and honey production (9), and these studies were led predominantly by national researchers.
Among the 17 reviews, two were global in scope: Aizen et al. (2019) found no trend of increasing pollinator dependence of crops in Africa, including Cameroon based on FAO data; Hill et al. (2019) cited a honey cooperative success story from Cameroon. Eleven other reviews were pan-African in scope: four on bee diversity and distribution (Chakuya et al. 2022; Coetzer & Eardley, 2019; Fabre et al. 2015; Hepburn et al. 1998), four on pollinators of different plant groups (Krasylenko et al. 2022; Nasare et al. 2022; Otieno et al. 2020; Rasplus, 1996), two on pesticide use and bee health (ASSAF, 2019; Pirk et al. 2016), one on conservation of Lepidoptera (Dongmo et al. 2018) and one on ecosystem services broadly (Egoh et al. 2012). Two reviews were solely focused on Cameroon (economically important Lepidoptera, Ouaba et al. 2022; apiculture industry, Njukang et al. 2021). The subset of original (i.e. not review) articles will be used for the remaining results.
Coverage by agroecological zone and by natural vs. agricultural systems
All of Cameroon’s five agroecological zones, in addition to montane ecosystems, were represented in the literature. As shown in Fig. 2a, the High Guinean savannah zone (pink) has been the location for the largest number of studies; in contrast, many fewer have been carried out in the Western Highlands (grey). Differences in climate and ecology across these five zones mean that certain crops do better in some zones compared to others. The dearth of studies from the Western Highlands zone is worth noting because it is the zone with the greatest percentage of land converted to farming, often farmed intensively due to land availability constraints, and has the lowest amount of natural landcover remaining (Tiamgne, 2015). The high production of vegetable crops involves unregulated pesticide use (Tambe et al. 2019). As such, applied studies that assess the impacts of farming methods and of surrounding habitat on pollinators in the Western Highlands have great potential to contribute to overcoming pollination deficits and increasing yield via recommendations for pollination services protection and management.
Agricultural systems were the focus of 100 (56%) of the studies, and tended to be led my Cameroonian researchers, while natural systems were the focus of 69 (39%) and were more likely to be led by international researchers. Ninety of the studies in agricultural systems evaluated the effects of wild and/or managed pollinator visitation on crop yield (the only managed pollinator being Apis mellifera), and in some cases this included evaluating additional effects such as fertilizers. Eighteen studies evaluated the apicultural value of wild and/or cultivated plants, while seven concerned case studies of beekeeping in different regions. Four were evolutionary in content, focusing on taxonomy and systematics of oil palm weevils (Elaeidobious spp.). The range of topics covered by studies in natural settings was broader than for agricultural studies and included pollinator behavior (14), taxonomy and systematics (12), adaptation and speciation (6), floral traits and specialization (10), spatial and temporal patterns (2), networks (1), and a methods paper. Looking at the distribution of study system by agroecological zone (Fig. 2B), we find few agriculturally based studies at high elevation, and they concern beekeeping rather than crop-based research. Only 7 (4%) studies featured both natural and agricultural systems in the same study, and these studies were concerned with identifying plants of apicultural value in areas that included both naturally occurring and cultivated species. The lack of comparative studies across types of land use represents a weakness in pollinator research in Cameroon that greatly limits its ability to inform conservation policy and practice.
Plant use categories
Grouping studies by the category of plant use (food, habitat, medicine, oil, fibre, dye, pasture, weeds, or timber) shows that plants used by people for food were the most studied, followed by plants primarily of importance to pollinators (“Habitat”, Fig. 3A). Among the food crops, vegetables had the highest number of species studied, followed closely by pulses (dry beans and peas) and oil crops used for food; per FAO categories, “oil crops” includes groundnuts, soybeans, and egusi that are high-protein and not primarily used for oil extraction in Cameroon (Fig. 3B). While the majority of studies of crop pollinators were ecological, most identifying pollinators and measuring their effects on yield, and others recording floral visitors, there were some exceptions. For example, all the studies that focused on the oil palm, Elaeis guineensis, were evolutionary in nature, investigating taxonomic and systematic relationships among palm-pollinating weevils. We found no studies looking at plant-pollinator interactions in oil palm, despite it being a cornerstone of smallholder livelihoods in Central Africa and a highly important plantation crop native to the region. Several FAO categories were not represented in any study; among these are found other economically important plantation crops such as coffee, cocoa, and cola nuts. The pollinators of tree fruits were very poorly studied as well. A multitude of other nuts and seeds, harvested from the wild, and therefore not technically crops, are also regularly sold and consumed in Cameroon (Awono et al. 2016). The plant-pollinator relationships that sustain these wild harvested products have yet to be investigated as well.
Pollinator dependence of crops
Most crops grown as annuals are self-fertile and to some degree, can self-pollinate, meaning they do not require pollinator visits to set seed. Nevertheless, numerous studies, including those covered in this review, have shown that pollinator visits to self-fertile crops can enhance the quantity and quality—weight, size, and viability—of seeds and fruit produced. Most major temporary crops in Cameroon that are propagated by seed have been assessed for pollinator dependence and its effect on yield under natural conditions, or semi-natural conditions where managed Apis mellifera hives occur either in the vicinity or are added for the study. Table 1 summarizes the results of studies on the effects of open pollination on yield, measured in terms of fruiting rate, seeds per fruit, and percent normal seeds in flowering crops left open to floral visitors versus those with flowers protected from visits by physical means. Yield increases due to pollinator visits are categorized ranging from “essential” (> 90% increase with pollinator visits) to “high” (41–90% increase) to modest” (between 10–40% increase) to “little” (under 10% increase) to “no increase” (when differences in yield were not significant), following the categories of Klein et al. (2007). When studies measured multiple parameters, we used the parameter “seeds per fruit” to assess % yield increase; where multiple studies exist, we noted the range. Of the 86 studies, all found a significant effect of pollinator visits on at least one of these parameters, including for self-pollinated crops such as peas (Pisum sativum) and wind-pollinated crops such as maize (Zea mays). In the case of peas, the manipulation control treatment produced similar yield results as pollinator visits, suggesting that some agitation of the opened flower enhances pollen transfer (Fameni-Tope et al. 2022). Improved yield in the presence of insect visits was the found in maize, a primarily wind-pollinated crop, where the foraging activity of bees released pollen and resulted in higher yields than in the wind-pollinated control plots (Pando et al. 2019; Dounia et al 2018; Tchuenguem-Fohouo et al. 2002).
Table 1
Crops for which pollinator effect on yield has been measured, including number of studies, parameters measured and significance, and if the pollination efficiency of pollinators was measured.
Plant categories | Total studies per plant spp. | Pollinator effect on yield: open pollination vs. bagged Number of studies, ns = not significant bold = entirely pollinator-dependent | Yield increase1: essential > 90%, high 41–90%, modest 10–40%, little < 10%, no increase (not significant) | Pollination efficiency measured2 Genus; all are Apoidea except Diptera: Chysomya, Musca |
Fruiting rate | Seeds per fruit | Seed quality |
CEREALS | | | | | | |
Maize, Zea mays | 4 | 0 | 4 | 2 | modest-high | Apis* |
FRUITS | | | | | | |
Custard apple, wild, Annona senegalensis | 1 | 1 | 1 | 1 | no increase-modest | Apis* |
Mango, Mangifera indica | 1 | 0 | 1 | 0 | essential | |
Safou, Dacryodes edulis | 1 | 1 | 0 | 0 | high | Meliponula (Axestotrigona)* |
Syzygium guineense | 1 | 1 | 0 | 0 | high | Apis* |
Watermelon, Citrullus lanatus | 3 | 3 | 0 | 0 | essential | Apis, Xylocopa |
OIL CROPS | | | | | | |
Egusi, Cucumeropsis manii | 2 | 1, 1 | 1, 1 ns | 1, 1 ns | no increase-modest | |
Groundnut, Arachis hypogea | 2 | 1, 1 ns | 1, 1 ns | 1, 1 ns | no increase-modest | |
Sesame, Sesamum indicum | 2 | 2 | 2 | 2 | modest | Apis |
Soy, Glycine max | 5 | 5 | 5 | 5 | modest-high | Apis, Lasioglossum, Lipotriches, Leuconomia |
Sunflower, Helianthus annus | 2 | 1 | 2 | 1 | modest-essential | Apis*, Ceratina |
PULSES | | | | | | |
Bambara groundnut, Vigna subterranea | 2 | 2 | 2 | 2 | modest | |
Beans, red, Phaseolus vulgaris | 3 | 3 | 3 | 3 | modest | Chalicodoma, Xylocopa, Lipotriches |
Beans, white, Phaseolus coccineus | 2 | 2 | 2 | 2 | modest-high | Xylocopa |
Cowpea, niébé, Vigna unguiculata | 8 | 7, 1 ns | 8 | 7, 1 ns | little-high | Apis, Megachile, Chalicodoma, Xylocopa |
Peas, Pisum sativum | 1 | 1 | 1 | 1 | modest | Apis, Lasioglossum |
Pigeon pea, Cajanus cajan | 6 | 6 | 6 | 6 | little-modest | Apis, Ceratina, Chalicodoma, Crocisaspidia, Dactylurina, Lipotriches, Xylocopa |
VEGETABLES | | | | | | |
African eggplant, Solanum aethiopicum | 1 | 1 | 1 | 1 | little | Apis |
Cucumber, Cucumis sativus | 1 | 1 | 1 | 1 | essential | Apis* |
Huckleberry, Solanum nigrum | 1 | 1 | 1 | 1 | little | Apis |
Ndolé, bitterleaf, Veronia amygdalina | 1 | 0 | 1 | 1 | high | Dactylurina |
Okra, gombo, Abelmoschus esculentus | 5 | 1, 4 ns | 4, 1 ns | 3, 2 ns | no increase-modest | Eucaria, Tetralonia |
Onion, Allium cepa | 2 | 2 | 2 | 2 | high | Apis |
Pepper, green, Capsicum annum | 1 | 1 | 1 | 1 | little | |
False sesame, Ceratotheca sesamoides | 1 | 1 | 1 | 1 | modest | Apis |
Tomato, Solanum lycopersicum | 1 | 1 ns | 1 | 1 | high | Xylocopa |
Waterleaf, Talinum triangulare | 1 | 1 | 0 | 0 | modest | |
COSMETICS & DYES | | | | | | |
Bixa orellana | 1 | 1 | 1 | 1 | modest | Apis, Xylocopa |
Shea, Vitellaria paradoxa | 2 | 2 | 0 | 0 | high | Apis* |
FIBRE | | | | | | |
Cotton, Gossypium hirsutum | 6 | 6 | 5, 1ns | 6 | no increase-high | Amegilla, Apis |
Luffa, Luffa cyclindrica | 1 | 1 | 1 | 1 | essential | |
Sisal, Agave sisalana | 1 | 1 | 0 | 0 | essential | Apis |
MEDICINAL | | | | | | |
Calotropis procera | 1 | 1 | 1 | 1 | essential | |
Castorbean, Ricinus communis | 2 | 2 | 1 | 1 | modest-essential | Musca, Chrysomya |
Cosmos sulphureus | 1 | 1 | 1 | 1 | modest | Apis |
Oxalis barrelieri | 1 | 1 | 0 | 0 | modest | Apis |
Physalis minima | 2 | 1, 1 ns | 1 | 1 | no increase-little | Apis |
Psorospermum febrifugum | 1 | 1 | 1 | 1 | modest | Dactylurina |
Urena lobata | 1 | 1 | 1 ns | 1 | no increase | Apis |
Mimosa pudica | 1 | 1 | 1 | 1 | little | Apis |
MULTIPURPOSE & PASTURE | | | | | | |
Brachiaria brizantha | 1 | 1 | 1 | 1 | little | Apis |
Callistemon rigidus | 2 | 2 | 2 | 1 | modest-high | Apis* |
WEEDS | | | | | | |
Striga hermonthica | 1 | 1 | 0 | 0 | essential | |
TOTAL STUDIES | 86 | | | | | |
1Yield increase due to pollinator visits. When multiple parameters were measured, "seeds per fruit" was used to assess % yield increase. |
2Pollination efficiency was calculated indirectly in terms of % visits or relative frequency; all others were calculated as Single Visit Efficiency |
It would be straightforward to take the additional step to translate the loss of pollination services into economic terms; the available yield data for diverse crops could be monetized to demonstrate the value of existing pollination services. It should be noted that these studies comparing yield from open pollination to yield under pollinator exclusion do not use pesticides during the flowering period, or at all, unless explicitly as an experimental treatment—but this is very rare. This careful or nonexistent pesticide use is in stark contrast to routine practices by farmers and thus is highly unlikely to represent actual farm yields from open pollination in Cameroon (Nchang et al. 2023); what it does more likely represent is potential yields if pesticide timing and use were altered and reduced.
One of the “Priority Research Areas” of Gemmell-Herren et al (2014) is determining whether and to what extent crops experience pollination deficits. Few studies (Agbor et al. 2022; Noubissié et al. 2014), explicitly evaluate whether pollination deficits exist, via pollen supplementation experiments. However, as can be seen in Table 1, many studies evaluate the pollination efficiency of a single visit by an abundant pollinator, so called because they have been observed to contact both reproductive organs and to transport pollen. When the yield of these singly visited flowers is higher than the yield under open pollination conditions (where a percent of the flowers may not have been visited, or have been visited by less effective pollinators), the difference can be expressed as a measure of pollination deficit. Such studies not only signal a pollination deficit but also provide information on which pollinator taxa are most relevant for addressing it, information that hand pollination studies alone do not provide. While some studies, particularly those occurring in the High Guinean savannah zone, are performed in the vicinity of managed honeybee hives, and thus are artificially enhancing pollinator abundance, none of the studies we read experimentally compared this enhancement to pollination in the absence of managed hives.
The terminology that is commonly used across studies often conflates testing for pollinator dependence with testing for auto- vs allogamy: these studies do not manipulate the source of the pollen deposited by the pollinators (could be self, same plant or different plant), but only whether pollinators were allowed to visit the flower. As such, they do not distinguish between being self-fertile yet lacking the ability to self-pollinate and being self-infertile (Delaplane et al. 2013). Five studies (e.g. Pando et al. 2019; Noubissié et al. 2014, Djongwawe et al. 2011) fully tested a plant’s breeding system. While varietal differences in the degree of auto-vs. allogamy could be important, breeding systems for most crops are generally already well known and what is more useful is understanding the degree to which crops are pollinator-deficient (pollen limited) under natural conditions, as well as the degree to which pollinator visits improve yield (Rodger et al 2004). As stated above, the latter has been well-studied in Cameroon for a wide range of insect visited crops.
Other ecological and behavioral data typically recorded in pollination ecology studies include visitation frequency, daily rhythm of visits, floral products harvested, duration of visits and foraging speed. Most agricultural studies were performed over two cropping seasons, allowing for comparisons of visitors and focal pollinators across years or seasons. Although interannual differences in floral visitation were not typically modeled statistically in relation to possible drivers (such as climate or co-flowering resources), the potential exists for metanalyses that explore temporal and spatial variation in visitor diversity in relation to yield. One aspect that would facilitate such an analysis is that most of the studies used nearly identical methodologies in terms of study design, and the parameters measured. Additionally, studies of several crops have been replicated in different agroecological zones; results for well-studied crops like pulses, okra and cotton could be compared across agroecological regions and over time.
Pollinator and crop health
Understanding how farm management practices contribute to pollination deficits goes hand in hand with developing the socioeconomic rationale for adopting pollinator-friendly practices, both priority research areas according to Gemmill-Herren et al. (2014). Little direct work has been done on this topic; most studies do not replicate routine farmer practices with respect to pesticides, instead avoiding or limiting their use. However, this trend may be changing, as a few recent studies compare the use of insecticides (synthetic or natural) and pollinator visits on yield; they find that pesticides reduce floral visits and subsequently reduce fruit and seed quantity and quality for some but not all crops studied (e.g., Nchang et al, 2023, Adamou et al. 2022, Azo’o Ela et al. 2021). These studies indicate an increasing interest in understanding the production tradeoffs between pollination and crop protection and a recognition of the need for less toxic solutions that include integrated pest management approaches. Yet, effectively communicating these findings to farmers requires strengthening the working relationship between research and extension (Ngamo, T.L.S, pers. comm.). Additionally, two studies assessed the toxicity of commonly used pesticides to honeybees (Kenko et al. 2022, Mazi et al. 2020), six studies assess the beneficial effects of pollination services on yield when combined with natural and/or inorganic fertilizer or mycorrhizae treatments (e.g. Nenkam et a. 2023; Djakbé et al, 2017), and two studies focus on honeybee pathogens, including the dipteran parasitoid Megaselia scalaris (Cham et al. 2018) and a first report of the Varroa mite (Cham et al. 2017).
Pollinator taxa studied
Among the original studies, excepting the reviews, there were 86 studies at the community level, including all pollinators and floral visitors, 29 that were focused on a particular taxonomic group, such as sunbirds (Nectariniidae) or butterflies and moths (Lepidoptera), and 59 that were concerned with a focal species only. Over half of the community-level studies also included data on a focal species; in fact, it was common for the applied plant-pollinator studies to compare the pollination efficiency of an abundant pollinator to the pollinator community at large (open pollination). Among the focal species, Apis mellifera was the most studied (25 publications), followed by Xylocopa olivacea (8 publications). Summarizing the focal species broadly by functional group shows that bees were the most studied (Fig. 4). The focus of so many studies on the honeybee is understandable, given it is thus far the only domesticated species in Cameroon, it is straightforward to identify, and it is both native to this country and of worldwide importance to agriculture. However, studies have shown that even for bee-pollinated plants, and even when abundant, A. mellifera is not necessarily the best pollinator (Page et al. 2021; Azo’o Ela et al. 2012). Tropical Africa, including Cameroon, is home to more than a dozen species of eusocial stingless bees (Apidae: Meliponini), several that have been successfully domesticated elsewhere (Ghana, Kenya, etc.), and have the potential to provide managed pollination services (Kiatoko et al. 2023). The nests of several stingless bee species have been studied (e.g. Njoya et al. 2019), but work remains to develop meliponiculture in this country. It should be noted that, despite not practicing beekeeping, traditionally forest-based ethnic groups here have in depth knowledge of the ecology of stingless bees that could facilitate both domestication and conservation efforts.
Research on other important pollinating groups, such as flies, has historically been limited to those taxa of public health concern, however work on the Syrphidae is gaining ground (Azo’o Ela et al. 2022; Jordaens et al. 2021; De Meyer et al, 2020; Ssymank, 2012). A recent study found Diptera to be diverse and dominant floral visitors, likely important to the reproduction of woody species in the lowland rainforest understory (Drager et al, 2022). Although flies are main pollinators of important tree crops grown in Cameroon, such as cocoa and cola nuts, little research has been done on these plant-pollinator relationships. Sunbirds have been the focus of various studies that demonstrate their roles as pollinators of numerous wild plant species in multiple ecosystems (e.g., Janecek et al. 2022). No studies were found of floral visitation or pollination by nectar-feeding bats, primates, or other non-avian vertebrate species, although such taxa have been studied in other contexts in Cameroon, such as One Health and biodiversity studies. More generally, nocturnal studies and those on the pollinators of canopy trees were few; difficulties of physical access and of access to technology, such as infrared video cameras, make such studies rare, but not impossible (Droissart et al. 2021).
Reliability of insect taxonomic identifications
The scientific value of pollinator research is based on the assumption that the taxa studied have been accurately identified. As Packer et al. (2018) explain, identification needs to be verifiable for research to be reproducible, but this is often not the case even in entomological studies, worldwide. Based on our evaluation of the methods described for the identification of specimens in the articles reviewed, of the 166 articles that featured a focal species, confidence in the identification of the species was high for 95 (57%), medium for 34 (21%), and low for 37 (22%) of studies. Confidence in taxonomic IDs for non-focal specimens or for community studies lacking a focal species, was high for 73 (43%), medium for 28 (17%), and low for 67 (40%) of studies. Appendix 2 includes individual rankings. None of the studies classed as medium or low confidence made mention of voucher specimens or even of high-resolution images being stored after identification, or of their storage locations. When dichotomous keys were cited, they often did not represent the taxonomic resolution of the identifications; typically, the keys were to families (e.g., Delvare and Alberlenc, 1989) and occasionally, for bees, to genera and sub-genera: Eardley et al. (2010), while the IDs given were typically to genus or species-level. The mention of comparisons with reference collections was rarely made and collaborations with taxon experts, although often the case for bee specimens in the period from 2000–2014, were rare afterward, despite a marked increase in the number of publications. Other less common concerns include species listed that are not known to be present in Africa, and names containing typographical errors. Given the concerns of reproducibility expressed above, and given that research on pollinators now dominates entomological research in Cameroon, the entomology community would benefit greatly from fostering taxonomic expertise in the Apoidea just as it has developed expertise for other non-pollinator insect groups (e.g., Grillidae, Nyobe et al. 2021; Psylloidea; Alene et al, 2021).
Baseline data needed
The accuracy of the taxonomic data in the agricultural pollination literature is an important issue because these publications are the main/only source of data on crop pollinator diversity. Despite the substantial and rapidly growing body of applied work on pollination, Cameroon still has no publicly available baseline diversity data, including species lists, for the best studied group, bees. When analyzing all species records for bees across African countries, Coetzer & Eardley (2019) found that the recorded number of species varied widely from one country to the next, even when countries resembled one another ecologically, leading them to conclude that the full bee diversity of most countries remains unknown and is strongly influenced by research effort. In fact, there is little infrastructure to support specimen collections in Cameroon and all specimen records assessed by Coetzer & Eardley (2019) are housed outside of the country. Cameroon has no entomological museum and only one partially digitized and published collection exists at the International Institute for Tropical Agriculture; the Apoidea in it have yet to be identified and digitized (Tindo, 2014). Research groups have reference collections, although few exist that have been curated by taxon experts, and these remain undigitized, making access very limiting, particularly for students from universities in other parts of the country. Logistical challenges include storing specimens in the humid tropics without adequate budgets for specimen cabinets or even wet storage, let alone climate control or high-resolution imaging equipment. However, simply publishing species lists for each agroecological zone, based on expert-curated reference collections, would be a way to begin generating a historical baseline that can help inform biodiversity management decisions. Incorporating a long-term pollinator monitoring protocol, however modest, in the research agenda of labs across the country would begin to generate longitudinal diversity data not tied to specific crops. Longitudinal monitoring studies are necessary for researchers to be able to evaluate significant trends in pollinator populations and advocate for science-based conservation policy (Bartomeus & Dicks, 2019).
Habitat requirements of pollinators
A diverse assemblage of wild pollinators is necessary for the persistence of natural ecosystems, for complementing the role of managed pollinators, whose importance varies by crop and ecosystem, and for providing insurance to farmers against temporal fluctuations or declines in the populations of the main pollinators, whether wild or domestic (Winfree et al. 2018). Perhaps the most understudied topic in Cameroon, compared to the pollinator literature globally, is the habitat requirements of wild pollinators. This is a topic that will increase in importance and urgency as natural areas continue to be lost and degraded in the interest of urbanization, farming and resource exploitation and industrial development. Studies that link diversity patterns to habitat type are useful for establishing habitat preferences and sensitivity to disturbance (eg. Nkoba et al. 2012, Gikungu et al. 2011, from Kenya). In Cameroon, to our knowledge, habitat use at the landscape scale has only been studied for Lepidoptera, of which not all are floral visitors (e.g., Mertens et al. 2021), and for sunbirds (e.g., Janeček et al. 2022). These studies provide highly valuable information on the spatiotemporal patterns and drivers of diversity along elevational and habitat gradients ranging from lowland rainforest to Afromontane habitat, and across seasons.
Learning what is needed for the management and conservation of wild pollinators and the ecosystem services they provide requires a comparative approach across land use types and spatial scales, an approach that is lacking in the applied pollination literature we reviewed (Fig. 2) (Dicks et al. 2021). In Cameroon, as elsewhere in the region, most agriculture is still small scale and polycultures and agroforests abound—these are factors associated with higher on farm pollinator diversity compared to larger, industrial monocultures (Gemmill-Herren et al. 2014). Crop-pollinator studies emerging from Cameroonian research groups follow a standard methodology that often includes a census of plant species currently flowering near the experimental plot that are attractive to pollinators, floral resources that can include other nearby crops, volunteer and weedy plants and species that are part of natural vegetation remnants. If these surrounding and on-farm resources could also be quantified, for example, in terms of number of floral resources, habitat type and size, and distance to field, this would provide crucial and currently non-existent data on the local effects of natural or semi-natural areas on pollination services in agriculture. In fact, what are needed are foundational studies such as Kremen et al. (2004) that elucidate natural area requirements for pollination services in agriculture, replicated in the Afrotropical context (Eardley et al. 2009). Freely available satellite data can facilitate the quantification of land use types and habitat degradation over time at larger, landscape scales. Spatial analysis, combined with pollinator diversity and efficiency sampling in and around agricultural fields, could help scientists and farmers understand the ongoing effects of land use change on pollination services. Such information would facilitate science-based recommendations for pollinator habitat management that farmers and farming communities could strive to adopt and that could inform environmental policy more broadly.