Tropical agroforestry supports insect pollinators and improves bean yield
Data files
Feb 28, 2024 version files 8.79 KB
-
Data_Kingazi_et_al..zip
3.04 KB
-
README.md
5.76 KB
Abstract
Pollination services contribute to crop productivity worldwide, but insect pollinators are declining in most agricultural landscapes, mainly due to agricultural expansion and an increase in intensive agricultural practices. To reduce the negative effects of agricultural expansion and intensification, farmers can adopt ecological interventions, such as diversifying agricultural systems through agroforestry. However, there have been variable results on the effectiveness of agroforestry in enhancing pollinators and pollination services. Furthermore, most of the information has been generated in temperate regions, while the impact of agroforestry in tropical East Africa is largely unknown. Using common beans (Phaseolus vulgaris L.) as the focal crop, this study tested whether a tropical agroforestry system called Chagga home gardens supports pollinator communities and improves pollination and crop yield. We examined 16 agroforestry plots paired with 16 non-agroforestry plots located along a gradient of woody (trees and shrubs) cover within a 1 km radius to document the abundance, species richness and visitation rates of pollinating insects on bean flowers and how they contributed to bean yield. The beans planted in the agroforestry plots had almost twice the abundance of insect pollinators, three times the richness of the species, and almost twice the visitation rates than those planted in non-agroforestry plots. We also found a significant positive effect of woody cover in the surrounding landscape on insect pollinator abundances, but not on species richness and visitation rates. Additionally, the abundance and richness of insect pollinators increased significantly with flower abundance, while the overall plant richness in a plot significantly increased insect pollinator abundance and their visitation rate. The difference in bean yield between unbagged flowers (in which insect pollinators were allowed to access bean flowers) and bagged flowers and the total yield were higher in agroforestry than in non-agroforestry plots.
Synthesis and Applications: This study showed that, compared to monoculture, agroforestry generally promotes pollination services in a tropical context. Furthermore, we found that agroforestry is likely to be particularly helpful for pollinators when it increases flower abundance and plant richness and if it is coordinated so that woody cover also increases at the landscape scale.
https://doi.org/10.5061/dryad.3tx95x6pd
Description of the data and file structure
The dataset contains three files.
-
Insect pollinator data: These data were used to analyse insect pollinator abundance, species richness, and visitation rates. The file contains 11 columns and 33 rows. All the column headings are self-explanatory. For the rows, apart from the headings row, the file has 32 rows (16 for agroforestry plots and 16 for non-agroforestry plots)
Farming system: This column contains the plots of the two farming systems studied in this paper (16 agroforestry and 16 non-agroforestry plots)
PairID: This column contains the pairs identification numbers, eg. we had 16 pairs and 32 plot, so pair 1 contains two plots, one is agroforestry and one is non-agroforestry
PlotID: This column contains plot identification number, eg. Pair 1 has two plots, P1A = Pair one agroforestry, P1N = Pair one non-agroforestry
Flower visitors abundance (Insect pollinator abundance): This contain number insect pollinators that visited bean flower in each plot during field survey. The abundance of insect pollinators per plot was calculated as the total number of individuals recorded visiting bean flowers within plots during study time.
Flower visitors richness (Insect pollinator richness): This contains number of insect pollinator taxa that visited bean flower in each plot during field survey. The species richness per plot was calculated as the total number of insect pollinator taxa that visited bean flowers within plots during study time
Visitation rate: This column contains mean number visitation per flower per observation time in each plot. Visitation rate was calculated as the total number of visits per flower per observation time
Wood cover: The mapping of woody cover at the landscape level started by dividing the study area into grids of 500 x 500 m in the QGIS software. The created grids were then converted to KML files. After that, the KML file was imported into Google Earth to guide the digitization, which was done in grids. The digitization of woody cover was done in Google Earth using Google Earth photos from the year 2022, the year that this study was conducted. After digitization, the woody cover layer was imported into QGIS to estimate the proportion (%) of woody cover within 1 km (1000 m) radii around each agroforestry and non-agroforestry plot.
Flower richness: The flower richness was estimated as the total number of wood and herbaceous plant species/taxa that were in bloom during the time of the survey.
Flower abundance (x10): For trees and shrubs, the overall abundance and species richness were obtained by recording all species and counting their individuals in each plot (Delaney et al., 2020). Furthermore, their flower abundance was subjectively estimated using a scale of 1 to 4 as per Samnegård et al. (2016). One (1) was given to a tree or shrub with <10 flowers while 2 =10-100 flowers, 3 = 101-1000 flowers and 4 = >1000 flowers. Subsequently, each scale was translated into an estimated average number of flowers, with 1 = 5 flowers, 2 = 50 flowers; 3 = 500 flowers, and 4 = 5000 flowers (Samnegård et al., 2016). For example, a tree with less than 10 flowers will be given a scale of 1 and its estimated average number of flowers will be 5. Herb floral resources and overall abundance and species richness were estimated in subplots and then extrapolated for the whole plot. A total of 9 (1m^2) subplots were established in each plot and all flowers, individuals, and species of herb plants were recorded.
Overall plant richness: This was estimated as total number of wood and herbaceous plant species in each plot
Overall plant abundance: This was estimated as total number of wood and herbaceous plants in each plot
-
NMDS data: These data were used to make the NMDS plot, The file contains insect species on the columns and farming systems plots on rows. Some of the species (rare ones) were removed during NMDS plotting but the full list of species is available in the supplementary materials of the paper
-
Bean Yield data: These data were used to analyse bean yield between pollinated and un-pollinated beans in agroforestry and non-agroforestry plots
Pair ID: This column contains the pairs identification numbers in which pollination experiment was conducted and we managed to harvest. This is because, we conducted pollination experiment in all pairs but in some plots farmers harvested their beans before we could harvest. We analysed bean yield in pairs where all the two plots were harvested so we can compare.
Plot ID: This column contains plot identification number in which pollination experiments were conducted
Habitat type: This term was used interchangeably with farming system, so it is the same thing
Pollination treatment: This column contain type of pollination experiments studied (Bagged = Bean plants were bagged with polyethylene net. The mesh prevented insects from visiting the flowers, but allowed sunlight and airflow to the bagged bean plants, Open = the bean flowers were continuously accessible to insects, wind, and autonomous self-pollination)
Plant ID in a plot: This column contain plant identifications number of the plants that were used for pollination experiment in each plot
Pods: We recorded the fruit set as the number of pods per plant.
Number of seed per pod: We recorded the seed set as the number of seeds per pod.
Yield quantity (g): The quantity of yield was measured as the weight (grams) of air-dry bean seeds per plan
- Kingazi, Nanyika; Temu, Ruwa‐Aichi; Sirima, Agnes; Jonsson, Mattias (2024). Tropical agroforestry supports insect pollinators and improves bean yield. Journal of Applied Ecology. https://doi.org/10.1111/1365-2664.14629