Crop diversity and within field multi-species interactions mediate herbivore abundances in cotton fields
Data files
Nov 03, 2023 version files 137.24 KB
Abstract
Insect herbivore abundances in agricultural fields partly depend on surrounding landscape compositional heterogeneity (e.g., landscape complexity). Landscape complexity can directly (e.g., dilution of host crops) and indirectly (promoting herbivore biocontrol) regulate herbivores in agricultural fields. While much is known about direct (e.g., resource concentration) and indirect effects (i.e., promoting biocontrol) of landscape complexity on herbivore populations, more work is needed to study whether landscape complexity can regulate herbivore populations by mediating within field multi-species interactions among herbivores and their shared natural enemies. During 2019 and 2020, we estimated Bemisia tabaci and Aphis gossypii abundances, their dominant predators (coccinellids, spiders, Orius spp., and Geocoris spp.), and their interaction (using molecular analysis) in 38 cotton fields along a gradient of landscape diversity across Georgia, USA. Within cotton fields, we assessed the effect of predator abundances, their frequency of feeding on herbivores, and the correlation between herbivore abundances (B. tabaci and A. gossypii) on the B. tabaci and A. gossypii abundance. At the landscape scale, crop diversity and different cover types influenced the abundance of B. tabaci and A. gossypii within cotton fields. We found a complex interaction among pests at the field scale, with higher aphid abundance correlated with decreased whitefly abundance. Our results support crop diversification for improving suppression of generalist pests in cotton landscapes through promoting biocontrol and diluting host crop area. Our result further suggests that the landscape complexity effect on whiteflies can indirectly mediate aphid abundance in cotton fields, indicating the importance of within field species interactions.
README: Crop diversity and within field multi-species interactions mediate herbivore abundances in cotton fields
https://doi.org/10.5061/dryad.8pk0p2nv7
Description of the data and file structure
Herbivore_abundance-cotton_leaf_samples.csv
This dataset is used to determine associations between whitefly and aphid abundance in cotton fields and agricultural landscape complexity in multiple spatial scales.
This dataset was further used to determine within-field predator-prey interactions and their impact on whitefly and aphid abundance in cotton fields.
Notes*
The covertpe values represent the percentage of each crop/non-crop habitat in the landscape within the represented spatial scale.
The predator and pest values represents the number of collected individual in each field (40 sweeps total).
1 The field identification indicates the county where the sample was collected in the State of Georgia.
2 The latitude of the sampled cotton field in Georgia.
3 The longitude of the sampled cotton field in Georgia.
4 The spatial scale at which the percentage covertypes is estimated in the landscape.
5 Merged value of all crops with less than 3% coverage in the landscape.
6 Merged value of all woody crops including blueberries and peaches in the landscape.
7 Merged value of all annual crops including cotton and peanuts, in the landscape.
8 Merged value of all wooded areas in the landscape.
9 Merged value of all habitats with plants that are not crops.
10 Merged value of all habitat types that are not crops, including urban areas, roads, and water bodies.
11 The total number of whitefly per field, calculated by counting whitefly numbers on ten cotton leaves per field.
12 Indicates the total number of predators (Coccinellids, Geocoris, Orius and spiders combined) screened for feeding on whiteflies and aphids in cotton fields.
13 The number of predators positive for whitefly DNA from the total number of screened predators.
14 The proportion of predators positive for whitefly DNA. The proportion was calculated by dividing the number of positives by the total number of tested predators.
15 The number of predators positive for cotton aphid DNA from the total number of screened predators.
16 The proportion of predators positive for cotton aphid DNA. The proportion was calculated by dividing the number of positives by the total number of tested predators.
Predator_abundance_1920_sweepnet.csv
This dataset is used to determine associations between predator abundance in cotton fields and agricultural landscape complexity in multiple spatial scales.
Notes*
The covertpe values represent the percentage of each crop/non-crop habitat in the landscape within the represented spatial scale.
The predator and pest values represents the number of collected individual in each field (40 sweeps total).
1 The field identification indicates the county where the sample was collected in the State of Georgia.
2 The latitude of the sampled cotton field in Georgia.
3 The longitude of the sampled cotton field in Georgia.
4 The spatial scale at which the percentage covertypes is estimated in the landscape.
5 The year of data collection, and landscape quantification.
6 Merged value of all crops with less than 3% coverage in the landscape.
7 Merged value of all woody crops including blueberries and peaches in the landscape.
8 Merged value of all annual crops including cotton and peanuts, in the landscape.
9 Merged value of all wooded areas in the landscape.
10 Merged value of all habitats with plants that are not crops.
11 Merged value of all habitat types that are not crops, including urban areas, roads, and water bodies.
12 The total number of whitefly per field calculated by counting whitefly numbers in 40 sweeps in cotton fields.
WF-AP_predation_data-1920_sweeps.csv
This dataset is used to determine associations between whitefly and aphid predation in cotton fields and agricultural landscape complexity in multiple spatial scales.
Notes*
The covertpe values represent the percentage of each crop/non-crop habitat in the landscape within the represented spatial scale.
The predator and pest values represents the number of collected individual in each field (40 sweeps total).
1 The field identification indicates the county where the sample was collected in the State of Georgia.
2 The latitude of the sampled cotton field in Georgia.
3 The longitude of the sampled cotton field in Georgia.
4 The spatial scale at which the percentage covertypes is estimated in the landscape.
5 The year of data collection, and landscape quantification.
6 Merged value of all crops with less than 3% coverage in the landscape.
7 Merged value of all woody crops including blueberries and peaches in the landscape.
8 Merged value of all annual crops including cotton and peanuts, in the landscape.
9 Merged value of all wooded areas in the landscape.
10 Merged value of all habitats with plants that are not crops.
11 Merged value of all habitat types that are not crops, including urban areas, roads, and water bodies.
12 The total number of whitefly per field calculated by counting whitefly numbers in 40 sweeps in cotton fields.
13 Indicates the total number of predators (Coccinellids, Geocoris, Orius and spiders combined) screened for feeding on whiteflies and aphids in cotton fields.
14 The number of predators positive for whitefly DNA from the total number of screened predators.
15 The proportion of predators positive for whitefly DNA. The proportion was calculated by dividing the number of positives by the total number of tested predators.
16 The number of predators positive for cotton aphid DNA from the total number of screened predators.
17 The proportion of predators positive for cotton aphid DNA. The proportion was calculated by dividing the number of positives by the total number of tested predators.
Methods
Abundance of herbivores and predators
In each sampled cotton field, we assessed the abundance of pests (i.e., cotton aphids, sweetpotato whitefly) and predators using forty 180-degree sweeps with a 38 cm diameter sweep net (0.1 mm openings). Twenty sweeps were collected along a 20 m transect in the border (i.e., 5 meters inside the field from the main road) and 20 sweeps towards the center of the field parallel to the border transect (i.e., 50 m into the field from the border). The average cotton field size in our study area was 36 hectares. Our sampling design allowed us to investigate the effect of the sampling location on predator and pest counts. Given the low abundance of whitefly and aphids in sweep-net samples in the 2019 field season (e.g., the immature stage of whitefly is attached to the leaf and occasionally found in sweep-net), we collected leaf samples (i.e., five leaves were haphazardly collected within each transect/ one leaf per plant) along with sweep-net in the 2020 field season to improve estimates of aphid and whitefly abundance (i.e., adults and immature combined) in cotton fields. After collection, we transferred sweep-net content to plastic bags and separated the predators in the field using aspirators. We transferred each predator into a separate 1.5 mL Eppendorf vial containing 99% ethanol and stored them in a cooler (~3-4 h) to minimize the possibility of DNA contamination (i.e., reduce the interaction period between predators and pests within sweep-net) and minimize degradation. Similarly, we transferred the leaf samples into plastic bags and stored them in the cooler. In the laboratory, we stored the sweep-net content and leaf samples inside a -20°C freezer until processing. Sweep-net and leaf samples were carefully checked under a microscope (Leica MZ APO, Leica Microsystems Ltd, Switzerland) to identify predators and pests using identification keys to the family and genus levels (Triplehorn et al., 2005; Ubick et al., 2017). The numbers of pests and predators were recorded, and predator samples were stored in a -20°C freezer until DNA extraction.
Landscape quantification
Georeferenced maps with crop and non-crop quantifications were obtained from CropScape (USDA National Agricultural Statistics Service, https://nassgeodata.gmu.edu/CropScape/) online map resources. The percentage area of each cover type (e.g., crops and non-crops) in the surrounding 2 km of the focal cotton field were quantified using ArcMap 10 version 10.7.1 (ArcGIS, 2021). Given no previous work on the effect of landscape complexity on whiteflies, we tested four spatial scales (0.5, 1, 1.5 and 2km) to determine the scale at which the landscape complexity had the most impact on whiteflies and their control. Twenty-seven cover types were identified within a 2 km radius around the 38 cotton fields (Figure S1 & S2, Table S2). Among these, 20 were crops, and 7 were semi-natural habitats. The major crops in these agricultural landscape regions were cotton, peanut, corn, and woody crops (notably, perennial fruit trees such as peach, blueberries and pecans). The less abundant crops (namely, crops with lower than 4% average coverage in the landscape) were merged into the other-crops category. The major semi-natural habitats were forests, wetlands, grasslands (all grassy areas including field margins), shrubland and pasture. The water bodies, including rivers and ponds, were merged into the water category. We used vegan package in R (Oksanen et al., 2015) and estimated overall landscape habitat diversity (i.e. Simpson diversity of all land cover types in a given buffer area, Simpson 1949). Similarly, crop diversity was estimated by excluding non-crop habitats from the landscape using the Simpson diversity index.