Skip to main content
Dryad logo

Data from: Ecological requirements drive the variable responses of wheat pests and natural enemies to the landscape context


Gonzalez, Ezequiel et al. (2021), Data from: Ecological requirements drive the variable responses of wheat pests and natural enemies to the landscape context, Dryad, Dataset,


1. Semi-natural habitats (SNH) are considered essential for pest suppressive landscapes, but their influence on crop pests and natural enemies can be highly variable. Instead of SNH per se, the availability of resources, such as pollen and nectar, may be more relevant for supporting pest control.

2. Here, we assessed the spatio-temporal variation of multiple insect pests (cereal leaf beetles and aphids) and natural enemies (predators and aphid parasitoids) in wheat fields and their responses to landscape context and flower availability. We combined detailed information on pollen use by natural enemies with the specific distribution of pollen-providing plants across a gradient of landscape composition and configuration.

3. The abundance of wheat pests was tightly linked to wheat development stage. Syrphids colonised the fields early in the season, while the abundance of other enemies increased later in the season. The responses of pests to landscape structure were variable and, while some pests had low abundances in landscapes with high edge density and SNH cover, Sitobion avenae abundance was positively associated with SNH cover. Lacewings, syrphids and cereal leaf beetles were abundant in landscapes with diverse and abundant flower resources, whereas the abundance of parasitoids and Nabis sp. was driven by aphid abundance. We detected no significant indirect effects of landscape on pests via natural enemies.

4. Synthesis and applications. Our findings highlight the need for conservation biological control to go beyond “one size fits all” and consider the specific ecology of the involved organisms. Landscapes with high edge density and flowering woody plants may support natural enemies, in particular syrphids, which colonised the fields early in the season. Incentives for pest-suppressive landscapes should focus on tailored strategies that disfavour pests and simultaneously enhance natural enemies according to their ecological requirements.


Study sites

We selected 19 winter wheat fields located in landscapes covering a gradient in landscape composition in terms of woody and herbaceous SNH cover, and floral resource availability (based on the flower resources consumed by Bombus terrestris L.; Eckerter et al., 2020). At each field, three transects parallel to a focal field edge were established along three consecutive tractor tracks, at distances ranging between 5 and 48 m from the edge and at a minimum distance of 50 m from other field edges. The mean nearest neighbour distance between focal fields was 1.9 ± 1.0 km (range 0.6-3.7 km). Landscape metrics were derived for circular buffers of radius 0.5 km around each focal edge. We assessed the nearest distances to forest, measures of landscape composition (proportion of forest, arable land, cereals, urban areas, semi-natural habitats and Shannon-Wiener index of crop diversity) and landscape configuration (edge density, woody edge density, field size and mean field size) in QGIS 3.6.2 (QGIS Development Team, 2019). Landscapes were digitised as polygon layers using Sentinel-2 satellite imagery as base maps and ground-truthed in the field in May-July 2020.

Insect sampling

Wheat pests and natural enemies were sampled using direct counting on wheat stems and sweep netting. Direct counting entailed the visual inspection of 40 wheat stems per transect per sampling date (for a total of 120 stems per field). Sweep-net sampling consisted of taking 100 sweeps per transect at 1 m intervals using a 30 cm diameter net. Samplings were repeated three times from May to early July 2020 during three wheat phenological phases (flowering, milk ripening, and dough phases).

For pests, we assessed the abundance of aphids (Sitobion avenae, Metopolophium dirhodum and Rhopalosiphum padi) and cereal leaf beetles (Oulema spp.). The following groups and stages of natural enemies were recorded: lacewings (Chrysoperla spp. eggs, larvae and adults), syrphids (eggs and larvae), ladybeetles (larvae and adults from several species of Coccinellidae), predatory bugs (i.e., adult Nabis sp.) and aphid parasitoids (mummies). In addition to raw abundances, we calculated aphid parasitism as the number of mummies per transect divided by total aphid abundance.

Assessment of floral resources

We expressed the availability of floral resources in terms of the total area covered by flowering plants that could offer resources for natural enemies, the flower diversity of the landscapes, and a specific index considering the pollen consumed by Chrysoperla carnea. To calculate these variables, flowering plants in each landscape were mapped during two occasions. First, woody plants were mapped in 2017-2019 (See Supplementary Methods in the article; Eckerter et al., 2020) in sampling plots of 10x10 m covering all hedgerows and forest edges of each landscape. Second, herbaceous flowering plants were mapped in May-July 2020 along transects in all the edges between landscape elements (Supplementary Methods). Both data sources were digitised in QGIS as vector layers and the total area per pollen type was calculated for each landscape. Only insect-pollinated plants were considered as these represent sources of both pollen and nectar for insects and are more likely to be actively visited by natural enemies (Wäckers et al., 2005). Plants with tubular flowers of limited accessibility were excluded from all calculations (van Rijn & Wäckers, 2016).

Total flower area was then calculated as the sum of the area covered by insect-pollinated mapped plants in each landscape. Flower diversity was calculated as the Shannon-Wiener Index of flower diversity, considering the relative cover of all insect-pollinated plants per landscape. Finally, we calculated a specific flower availability index (FAI) based on the pollen ingested by adult green lacewings collected using sticky traps in 2019 in the same study region (Supplementary Methods). The FAI was calculated using the formula developed by Eckerter et al. (2020), which sums the relative cover of plants providing each pollen type in each landscape times their proportional use by lacewings expressed as the share of total ingested pollen volume (Supplementary Methods). Thus, the FAI allows to compare the availability of pollen between landscapes, weighing the contribution of each pollen type according to its relevance for lacewings. The index was calculated for lacewings, but should also be relevant for ladybeetles, given the similarity of the pollen diet of these two important aphid enemies (Bertrand et al. 2019).

For references and additional details, see the original article in Journal of Applied Ecology.


Alexander von Humboldt-Stiftung, Award: Georg Forster Postdoctoral Fellowship

Alexander von Humboldt-Stiftung

Alexander von Humboldt-Stiftung

Faculty of Environmental Sciences, Czech University of Life Sciences Prague, Award: I.G.A. grant 42110/1312/3118

Deutsche Forschungsgemeinschaft, Award: EN 979/3-2