Abstract
Global biodiversity has suffered a decline primarily attributed to landscape simplification and intensified agricultural practices. Agricultural environments, characterized by homogeneity and frequent disturbances, are often suboptimal habitats for various insect species. While agricultural fields do favour pests, they generally fail to provide suitable habitats for natural enemies. The inclusion of diverse supporting habitats, such as semi-natural habitats, grassy and woody field margins etc. surrounding agricultural fields, play a crucial role in fostering effective biodiversity conservation. Moreover, determining the influence of different adjacent habitat types is essential in elucidating their influence on pest abundance and parasitism rates. Our two-year field study focused on assessing the abundance of Brassicogethes aeneus and its parasitism rate. The findings revealed that the adjacent habitat type did not significantly increase pest abundance and the parasitism rate of B. aeneus larvae consistently stayed over the threshold for effective biological control throughout the fields. This was attributed to the high proportion (35 and 38% in the 2 study years) of semi-natural habitats within most of the 1 km radius study areas. While our study did not identify any specific adjacent habitat type or habitat within a 1 km radius that directly impacted B. aeneus abundance, it emphasises the intricate interplay between the pests, parasitism and the surrounding environment because the interactive effect of distance from the crop edge and habitat type had a significant influence on B. aeneus infestation levels but not on parasitism. Decision tree analysis suggests that > 18% semi-natural habitat is needed to ensure sufficient levels of parasitism for effective biological control. A comprehensive understanding of habitats that influence not only B. aeneus but also other pests is critical for the successful implementation of IPM strategies and conservation initiatives within the agricultural sector.
Similar content being viewed by others
Introduction
The global biodiversity crisis has worsened in the last decades, mostly due to landscape simplification and the intensification of the agricultural sector1,2,3. To compensate for the biodiversity decline surrounding agricultural fields, targeted agri-environment schemes have been implemented, supporting farmers in adopting organic farming practices, by adding more flowering strips and grassy field margins to the agricultural landscape4,5. While implementing agri-environmental schemes on farmland offers advantages, it is crucial to retain semi-natural habitats (SNH) around the agricultural fields for effective biodiversity conservation. SNHs provide necessities, like food resources, shelter, and hibernation habitats, for various insect taxa6, including pollinators and predatory arthropods. Pollinators and predatory arthropods play vital roles in agricultural production by increasing yield through pollination and reducing pest density through biocontrol7,8,9,10,11.
Oilseed rape (Brassica napus L.) (OSR) is an economically important crop around the world, mainly used to produce vegetable oil and biofuel feedstock, grown in large monocultural fields, and attracts various herbivorous insects specialised on cruciferous plants. One such insect that causes damage to OSR plants is the pollen beetle (Brassicogethes aeneus Fabricius syn. Meligethes aeneus Fabricius), damaging both winter (WOSR) and spring oilseed rape plants at the green bud stage that may cause yield reduction. Brassicogethes aeneus overwinter in woody and sheltered habitats and in spring, they seek food sources and oviposition sites12. They oviposit into the buds, and while first instar larvae develop inside the buds, the second instar larvae exclusively feed on pollen from open flowers13. However, the main damage is done by adults feeding on the buds, leading to bud abortion and eventual cessation of seed development14, resulting up to 80% loss in yield on spring oilseed rape15. The control of B. aeneus in conventional OSR fields relies on synthetic insecticides, which not only reduces the abundance of the target pest but also affects the natural pest control providers and pollinators16,17,18,19. In line with the European Union’s Farm to Fork Strategy, there is a target to reduce overall pesticide use by 50% by 203020. This urges farmers to adopt alternative pest control methods and rely on the predatory arthropods and parasitoid wasps. Parasitism by parasitoids play a significant role in natural pest control. Brassicogethes aeneus is attacked by at least nine species of hymenopteran parasitoids, with three key species: Tersilochus heterocerus (Thomson), Phradis interstitialis (Thomson) and Phradis morionellus (Holmgren)21. These three parasitoid species are univoltine and koinobiont endoparasitoids. They overwinter in pupal cocoons in the soil, close to their hosts, and emerge in spring to migrate to the OSR fields22. It is known that these parasitoids differ in their host-finding behaviours and attack their hosts at different growth stages, suggesting niche separation among them23. Despite the host-finding behaviours, studies have shown that the parasitism rates of B. aeneus can vary widely from 0 to 63%24,25,26, and can depend on the distance from adjacent habitat27 and farmland management methods28,29,30. Landscape structure and composition significantly affect both B. aeneus and its parasitoids. One alternative pest control method is conservation biological control. This aims to reintroduce beneficial insects into crop systems to improve natural pest management. Previous studies have highlighted the links between conserving natural habitats and reducing pest pressure in farmland31,32. Whereas homogeneity in the landscape enhances the presence of B. aeneus in OSR field, more heterogenic landscape enhances species richness, abundance and fecundity of parasitoid wasps7,32,33. Additionally, complex landscapes, as opposed to simple ones, provide a higher parasitism rate of B. aeneus7,34. Brassicogethes aeneus abundance is influenced by changes in total OSR area size and the distance from the previous year’s OSR fields, which can effectively lower B. aeneus abundance while maintaining sufficient parasitism rates26. In this two-year study, we examined the landscape effects on B. aeneus abundance and parasitism of B. aeneus larvae in OSR fields. Our objectives were to determine the influence of adjacent habitat type and the in-field distance from the adjacent type on the abundance and parasitism rate of B. aeneus. Additionally, we sought to examine how the surrounding habitat types within a 1 km radius affect both B. aeneus abundance and the parasitism rate of B. aeneus larvae.
Results
Landscape characteristics
In 2014 and 2015, across the studied landscape circles the average coverage of arable land was 53.8% and 58%, respectively, with a minimum coverage of 20.2% and a maximum of 82.7%. The average SNH proportion across landscape circle coverage was – 37.7% in 2014 and 34.7% in 2015, the minimum cover of SNH was 11.3% (in 2015) and maximum of 52% (in 2015), of which woody areal elements typically occupied the largest area. The average coverage of OSR fields was 14.7% in 2014 and 15.7% in 2015. The mean proportion of SNH in landscape circles of focal fields with different adjacent elements was quite similar in both study years—herbaceous linear element [containing narrow (1.5–12 m width) grassy crop margin] 33.8%, woody linear element (containing hedge or line of trees) 40.9%, and another crop field (either bordering with OSR, barley, wheat, potato, oat, or pea) 38.3% in 2014 and 33.8%, 36.4% and 34.2% respectively in 2015. Also, the mean proportion of arable land in landscape circles was relatively similar in both years. In 2014, arable land constituted 54.0% on landscape circles where focal field was bordered by herbaceous linear element, 51,6% on landscape circles where focal field was bordered by woody linear elements and 55.9% on landscape circles where focal field was bordered by another crop field In 2015, the proportions changed slightly – 59.3%, 54.6% and 60.1%, respectively. The full list of landscape characteristics is presented in the supplementary materials (Table S1–S2).
Effects of adjacent habitats and distance from the edge of the field on B. aeneus abundance
Generally, throughout the study period, the mean number of B. aeneus per WOSR plant was low and stayed firmly under the economic threshold level (in Estonia 1–2 beetles per plant at BBCH 50–51 and 3–5 at BBCH 55–5935). The mean number of B. aeneus adults was high and surpassed the economic threshold level (mean 5.72 ± SE 1.29) in only one of the fields, during the last sampling round (BBCH 55–59) in 2015. There were three sampling rounds in 2014, during which a total of 2160 WOSR plants were tapped, and four sampling rounds in 2015, during which a total of 2880 WOSR plants were tapped. Observing the yearly means of B. aeneus per study field, shows that only one field in 2015 was close to the economic threshold, when the mean number of beetles per plant was 1.86 (SE ± 0.37) (Table S3). In 2014, the mean number of B. aeneus per plant was highest in fields adjacent to woody linear (0.32 ± SE 0.79), followed by another crop (0.26 ± SE 0.56) and herbaceous linear habitat (0.19 ± SE 0.39) (Figure 1). Whereas in 2015, the greatest mean number per plant was found in fields adjacent to herbaceous linear (0.59 ± SE 0.84), followed by another crop (0.56 ± SE1.54) and woody linear (0.34 ± SE 0.41) (Figure 1). The combined effect of distance and adjacent habitat type (2014: χ2= 65.35, df= 6, p < 0.0001; 2015: χ2= 43.76, df= 6, p < 0.0001) had a significant effect on the estimated mean abundance of B. aeneus adults in both study years. Generally, the estimated mean value of B. aeneus per ten plants was higher at 2 m from the adjacent habitat and the abundance declined with distance from the boundaries (Table 1).
Effects of habitats surrounding the focal field on B. aeneus abundance
We analysed the correlations between B. aeneus abundance in focal fields and various landscape characteristics within 1 km radius from the focal field in pooled data of 2014–2015. In our study, we did not find any significant correlations between them. Furthermore, we also carried out random forest tree analyses but none of the forests could describe representative proportion of the samples. The best model described only 16% from the samples that is not enough to extrapolate these results to the whole population and to draw fundamental conclusions. However, only for 2014 a positive correlation between herbaceous linear elements and the abundance of B. aeneus was found (Pearson correlation r = 0.68; p = 0.002; Table S4). In 2014 the abundance of B. aeneus increased significantly with increasing proportion of herbaceous linear habitats within the 1 km radius of focal field however, in the next year, 2015, the correlation between herbaceous linear habitats and B. aeneus abundance was opposite, meaning the abundance of B. aeneus decreased with increasing proportions of herbaceous linear habitats but relationship was not statistically significant. Therefore, we cannot conclude that herbaceous linear habitats may increase B. aeneus abundance as our results were inconsistent. Other habitats within the 1 km radius of focal field did not have significant correlations with B. aeneus abundance, in both 2014 and 2015 (Tables S4–S5).
Effects of adjacent habitats and distance from the edge of the field on parasitism rate
The parasitism rate of B. aeneus larvae in both study years was very high and even reached up to 100% at field level on three fields adjacent to herbaceous linear and woody linear habitats, while in two focal fields adjacent to herbaceous linear habitat no parasitism was observed (Table S3). The parasitism rate in both study years was unaffected by adjacent habitat, distance from the adjacent habitat nor the combined effect of them (all p values > 0.05) (Figure 2). The mean parasitism rate in 2014 and 2015 was high with 66.24% (SE ± 2.71) and 71.85% (SE ± 3.87), respectively. The mean parasitism rate was high across all fields regardless of the adjacent habitat, both in 2014 and 2015. The highest parasitism rate was observed in focal fields adjacent to another crop 73.04% (SE ± 4.73) in 2014 and in 2015 in focal field adjacent to woody linear habitat 81.02% (SE ± 7.89).
Effects of landscape characteristics on B. aeneus parasitism rate
The two years parasitism rate of B. aeneus was very high and reached up to 69%. The random forest analysis showed an overall R-squared value of 0.45 suggests that approximately 45.3% of the variance in the B. aeneus parasitism percentage can be explained by the predictor variables selected by the model. A decision tree (Figure 3) revealed that the parasitism rate of B. aeneus varied based on the landscape characteristics surrounding the focal field. When the semi-natural habitat proportion (SNH) around the focal field was below 18%, the parasitism rate was only 24%. When the SNH proportion around the focal field was over 18%, the predicted parasitism rate was 70%. It is important to highlight that 97% of data connected with parasitism rate were sampled from landscape circles where SNH proportion was greater than 18% and only 3% of data was collected from landscape circles with SNH percentage below 18%. In addition, almost 30% of parasitism rate data was collected from areas where SNH proportion was greater than 43%. Thus, our data reflect an agricultural landscape, that is heterogeneous due to its very high SNH coverage, which is important for conservation biological control. The decision tree also revealed that if the proportion of SNHs exceeded 43% and the proportion of herbaceous area was over 3.9% in the surrounding landscape then the median of parasitism rate of B. aeneus was 87%. When proportion of SNHs was lower than 43% then the proportion of arable land influenced the parasitism rate—resulting in higher parasitism rate (72%), if the proportion of arable land was more than 54%, whereas if the proportion of arable land was smaller than 54%, the parasitism rate was 50%.
Discussion
The field study conducted over two consecutive years showed very low abundance of B. aeneus during the damage susceptible growth stage of WOSR and even though no relation between the bordering habitat and their abundance was found. The mean larval parasitism rate of B. aeneus was extremely high with 69% and occasionally reaching up to 100%, indicating an effective natural pest control in the landscape. This is confirmed by the decision tree that showed that landscapes with more than 18% SNH coverage have a high (70%) predicted parasitism rate.
The study showed that the abundance of B. aeneus during WOSR bud stage, that is the growth stage most susceptible to damage by B. aeneus, was very low in both years and the bordering habitat had no significant impact on their abundance. Also, it is important to highlight that in our study area the agricultural landscape surrounding study fields had very high proportion of SNHs within 1 km radius around the focal fields (Table S3). Therefore, our study shows that in diverse agricultural landscapes comprised of woody and herbaceous linear elements there is a lower risk of B. aeneus infestations in WOSR. This was despite the presence of a multitude of overwintering sites near the focal fields. Thus, the number of B. aeneus was low and did not exceed the economic threshold during bud stage of plants. Studies on the B. aeneus density, plant damage, and local or regional landscape effects were shown to be inconsistent. Several studies suggest that complex landscapes may reduce pest density34,36,37,38,39,40,41, although there are also evidence of increasing pest pressure42,43. The proportion of woodland in the landscape is usually associated with an increase of B. aeneus abundance43,44,45, whereas for grasslands there were contrasting results, indicating both increases and decreases in abundance34,42. As woody areas provide a suitable overwintering habitat for B. aeneus, this can increase pest density in the spring46, however, regardless of presence of quite large woodland areas in the study sites, the results of our study did not confirm this finding as we did not find any significant correlations between beetles abundance and woodland proportions. Pest density and its relationships with the landscape configurations have complicated and multilevel ties. There are many direct and indirect factors at large and small scale that influence pest abundance such as crop rotations, pesticide usage, tillage regimes, abundance, and species richness of naturally occurring enemies of pests, pest pathogens, amount, quality and connectivity of landscape elements, soil health, weather conditions etc. In our study, the adjacent habitat of WOSR did not influence B. aeneus abundance, but the distance from the adjacent habitat and the combined effect of distance and adjacent habitat type had a significant impact on the estimated mean abundance of B. aeneus throughout the study years. The abundance of B. aeneus during inflorescence emergence stage (BBCH 50–59) was greater at 2 m and 25 m from the edge of the WOSR and decreased at 50 m and 75 m from edge. This suggests that local adjacent habitats were the source of infestation. It was not possible to identify which habitat type led to greater infestations because results were inconsistent between years. Other studies also showed that B. aeneus tends to aggregate on the field edge during bud stage of plants and expand more into the field once the plants start to flower47,48,49. This shows that there are opportunities to use Integrated Pest Management (IPM) tools such as a precision agriculture techniques and/or trap cropping to manage B. aeneus abundance and use insecticide treatments only along WOSR headlands to reduce the agrochemical inputs and preserve natural enemies24,25,48,50,51,52. However, aggregation at the edges of fields is not common for all insect pests of oilseed rape, for instance cabbage seedpod weevil (Ceutorhynchys obstrictus Marsham), does not have similar in-field immigration pattern and is evenly distributed at different distances from the edge53, also Psylliodes chrysocephala (L.) is non-uniformly distributed within the crop and does not aggregate along field edges54. This may because of differences in overwintering sites and/or dispersal abilities.
In both years, the parasitism rate of B. aeneus larvae was extremely high with average rate of 69%, which is close to previous finding in Estonia26 and over the minimum effective biological control level (at least 32%)55. In our study, we found that the parasitism rate was not influenced by the adjacent habitat or by distance from the field edge. Interestingly, the mean parasitism rate was consistently high across almost all focal fields regardless of the adjacent habitat type, indicating a strong and efficient level of natural pest control against B. aeneus in both years. Our findings are comparable to Kovács et al.,53, where they also showed a similar effect on C. obstrictus; the adjacent habitat type did not affect the parasitism rate which was likewise outstandingly high. In our study, in 33 of 36 fields the parasitism rate was above the threshold value of effective biological control (32–36%)55, even reaching up to 100% in three fields. Thus, the population size of beetles was efficiently controlled by hymenopteran parasitoids. Unfortunately, it was not possible to identify the impact of adjacent habitats on levels of parasitism in focal fields. The high proportion of different semi-natural and other non-cropped habitats can potentially provide suitable habitats for parasitoids and created an abundance of parasitoids at a landscape scale, thereby masking any field-scale effects. Landscape complexity and the amount of SNHs in agricultural landscape have previously shown to have a positive impact on pollen beetle (Brassicogethes ssp) parasitism34,37,41,56,57. Semi-natural habitats around and nearby arable fields provide alternative host plants32,58,59,60, and as SNHs tend to have lower disturbance, parasitoids can sustain a proficient population at landscape level and therefore provide the pest control service to agricultural fields. Furthermore, the study by Le Clec'h et al.61 indicated a dilution of the additional effect of linear elements on ecosystem services supply when linear elements were implemented in a landscape where the supply of ecosystem services was already high, compared to a landscape where the initial supply was low. The decision tree created in our study, based on the landscape and parasitism data, revealed that if the SNH percentage was below 18% in the 1 km radius landscape circle the parasitism rate was low. Thies and Tscharntke60 showed that the SNH coverage below 20% led to the decrease of B. aeneus parasitism below the threshold necessary for effective biological control. The decision tree based on our data confirms their results—our model predicted that if the percentage of SNHs was below 18% then the predicted parasitism rate was only 24% and if it was over 18% then predicted parasitism rate was 37–90%. Thus, our results show that having a complex landscape, it is possible to provide a consistent parasitism rate over the two years. Furthermore, two main parasitoid species of B. aeneus, Phradis morionellus and P. interstitialis are oppositely affected by the woodland coverage around OSR fields62 and may therefore contribute differently on the overall parasitism of B. aeneus based on woody habitat abundance and location. To ensure functioning ecosystem, pollination, biological pest control service at least 10–20% SNH is recommended per km2, although our results suggest that 20% would be the minimum level in 1 km radius of a focal field63. Our results also support the previous findings showing that the landscape with a low proportion of oilseed rape had a negative effect on the parasitism rate37. In the same study region, Kovács et al.,53, stated that the landscape in the study area is still sufficiently diverse and complex, providing different habitats for both the pest and the natural enemies. Our results classified by the decision tree highlight the importance of landscape characteristics, particularly the presence of SNH as these types of areas are the necessary for conservation biocontrol.
In conclusion, our study highlights the complicated relationship between pest, parasitoids and landscape and highlights that a specific adjacent habitat type or the habitat within the 1 km radius did not increase B. aeneus abundance. Understanding the habitats and factors that influence not only B. aeneus but also other pests, is vital for implementing successful integrated pest management strategies and conservation efforts in agricultural landscapes. Future study should investigate the processes that drive parasitism rates and evaluate other factors that may contribute to the observed trends. Such knowledge will be useful in designing efficient strategies to improve natural enemy populations and facilitate biological control of B. aeneus in agricultural landscapes.
Methods
The study was carried out in 18 non-overlapping landscape circles with a radius of 1 km around the focal fields in Tartu County, Estonia in 2014 and 2015. General study site selection, landscape circle radius of 1 km, classification of SNH-s and experimental design were based on the standardized protocols of the EU FP7 project QuESSA across participating regions and crops and are previously described in detail:53,64,65,66,67. For the project as well as our study, the 1 km radius was chosen, as it aligns with the recommended spatial scale for the dispersal of oilseed rape parasitoids in landscape sectors41,62. In both study years, 18 WOSR focal fields were selected; six bordered with woody linear, six fields with herbaceous linear and six with another crop. The minimum size of the focal field was 3 ha, the width for both woody linear and herbaceous linear was between 1.5 and 25 m and at least 50 m long. In 2014, the bordering crop was spring oilseed rape (1 field), barley (2 fields), wheat (2 fields), potato (1 field), and in 2015 we had barley (1 field), wheat (2 fields), oat (1 field), pea (2 fields). The landscape surrounding the focal field within 1 km radius for both years (2014 and 2015), was mapped separately with ArcMap 10.1 (ESRI 2012) using data from the Estonian Land Board and the Agricultural Registers and Information Board with any remaining gaps filled during field visits. The minimum distance between fields was larger than 1 km, ensuring that the landscape in the 1 km radius around the focal field did not overlap. Based on the dominating vegetation, semi-natural habitats (SNH) were classified as: herbaceous areal (semi-) natural hayfields or abandoned fields that have not developed > 30% shrub/tree canopy cover), woody areal (woodland), cover crop, herbaceous linear (narrow grassy crop margin) and woody linear (hedge, line of trees), their area and proportion within 1 km radii of the focal field was calculated using ArcMap 10.1. The focal fields in this study were managed conventionally. Each sampling transect was set at 90° from the focal field bordering element progressing into the crop. Within each field they were located at least 25 m away from other field edges, that were not the focal adjacent habitat type.
B. aeneus abundance assessment
The abundance of overwintered adult B. aeneus in WOSR fields was assessed using the plant tapping method68,69,70. The abundance of B. aeneus adults was the assessed along the sampling transect at four distances: 2, 25, 50, and 75 m from the edge of the crop. At each sampling point ten random WOSR plants were selected, the plants’ main raceme was tapped over a tray to dislodge B. aeneus from the buds and flowers. Dislodged B. aeneus which fell onto the tray were counted. In 2014, sampling started at the bud stage of plants (growth stage BBCH 52–53; following the decimal code by Lancashire et al.71), and ended when the plant growth stage was BBCH 57–60 (end of bud stage and beginning of flowering stage). In total, three sampling occasions were performed in 2014. In 2015, there were four sampling occasions, sampling started when the plant growth stage BBCH 50–51(bud stage of plants) and sampling ended when the plant growth stage was BBCH 60–61 (start of flowering stage).
Collection of the B. aeneus larvae and parasitism rate
The parasitism rate of B. aeneus larvae was assessed using the funnel trap method27. Funnel traps were placed along the sampling transect at three distances: 2 m, 25 m, and 75 m from the edge of the field. The trap consisted of a plastic funnel (diameter 31 cm) and a plastic container (50 ml) attached to the funnel end. The upper part of the funnel trap remained above the ground below the side shoots while the lower part with collection cup was dug into the ground. The trap was placed close to the WOSR plants at the end flowering stage of plants (BBCH 67) and were kept in the field until the end of flowering (BBCH 70). In each year a total of 162 samples were collected. The larvae of the last growth stage were collected as they were ready to leave from flowers and drop to the ground to pupae in soil. The area of which the larvae dropping from the plants were collected was equal to the diameter of the funnel (754.4 cm2). Funnel traps were emptied once a week for a period of three weeks. The content of the funnel traps was strained through a fine mesh fabric and placed into labelled plastic bags. In the laboratory, samples were sorted, all B. aeneus larvae were counted, placed in labelled Eppendorf tubes, containing distilled water, and stored in the freezer at − 20 °C for later dissection. All larvae were dissected using green food colouring as contrast medium to detect parasitoid eggs and larvae. All parasitoid eggs and larvae were counted and, when possible, identified to species or genus using identification key by Osborne72.
Statistical analyses
Statistical analyses were carried out in R statistical software (version RStudio 2022.02.3 + 492)73. To determine the effects of adjacent habitat type and the distance from the field edge on B. aeneus abundance (count data), generalized linear mixed model (package “glmmTMB”74), with Poisson distribution and log link function was used. We generated a model for both 2014 and 2015, were the adjacent habitat type (categorical variable), distance (categorical variable), and their interaction as fixed-factors and the field site (in total 18 fields per year), sampling round and the number of days from the last insecticide application as random factors in model. Analysis of deviance of Wald statistic Type II was used to test the individual and combined effect of adjacent habitat type and the distance from the edge. We applied estimated marginal means (package “emmeans”75) to compare marginal mean of B. aeneus abundance in the interaction model. Pearson's correlation to examine the relationship between the habitats within the 1 km radius of focal field and B. aeneus abundance both in 2014 and 2015 were performed. The parasitism rate of B. aeneus larvae was determined for each sampling point by multiplying the total number of parasitized larvae by 100 and dividing the result by the total number of larvae captured with the funnel trap at a specific sampling point. To analyse the adjacent habitat type and the distance from the field edge effect on parasitism rate, generalized linear mixed model with binomial distribution was performed, followed by analysis of deviance applying Wald statistic Type II. Adjacent habitat type, the distance from the edge and the combined effect of them were used as fixed effects, and the field number as random factor. Missing values of parasitism rates were removed prior to the analysis. We used DHARMa package76 to assess the model fit.
Random forest analysis was carried out to determine the variance that can be explained by the landscape types surrounding the focal fields and afterwards a decision tree was built. For this analysis we combined both 2014 and 2015 dataset of landscape characteristics and the parasitism rate to have complete dataset. For random forest, we used previously described proportions of different landscapes within 1 km radii. These variables were used as predictors in the random forest model. Model was set to generate 5000 trees at one run and re-run ten times (creating in total 50,000 trees) to see the variation in the variable importance. We used cross-validation technique to evaluate the performance of the models. The R-squared value, which measured the proportion of variance in the observed parasitism explained by our model predictions. The decision tree was built on the predictors and target variable. We used “anova” method, which is used to solve regression task where the target variable is continuous or numeric. For the analysis, we used the ‘randomForest’77, ‘rpart'78 and ‘partykit'79 pacakages in Rstudio.
Data availability
Data used in the study is available from the authors on a request.
References
Grab, H., Danforth, B., Poveda, K. & Loeb, G. Landscape simplification reduces classical biological control and crop yield. Ecol. Appl. 28, 348–355 (2018).
IPBES. Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. https://zenodo.org/record/6417333 (2019).
Rusch, A. et al. Agricultural landscape simplification reduces natural pest control: A quantitative synthesis. Agric. Ecosyst. Environ. 221, 198–204 (2016).
Boetzl, F. A. et al. A multitaxa assessment of the effectiveness of agri-environmental schemes for biodiversity management. Proc. Natl. Acad. Sci. U.S.A. 118, e2016038118 (2021).
Marja, R., Tscharntke, T. & Batáry, P. Increasing landscape complexity enhances species richness of farmland arthropods, agri-environment schemes also abundance – A meta-analysis. Agric. Ecosyst. Environ. 326, 107822 (2022).
Holland, J. M. et al. Structure, function and management of semi-natural habitats for conservation biological control: A review of European studies: Structure, function and management of semi-natural habitats for biological control. Pest. Manag. Sci. 72, 1638–1651 (2016).
Dainese, M. et al. A global synthesis reveals biodiversity-mediated benefits for crop production. Sci. Adv. 5, eaax0121 (2019).
Raderschall, C. A., Bommarco, R., Lindström, S. A. M. & Lundin, O. Landscape crop diversity and semi-natural habitat affect crop pollinators, pollination benefit and yield. Agric. Ecosyst. Environ. 306, 107189 (2021).
Sutter, L., Albrecht, M. & Jeanneret, P. Landscape greening and local creation of wildflower strips and hedgerows promote multiple ecosystem services. J. Appl. Ecol. 55, 612–620 (2018).
Sutter, L. & Albrecht, M. Synergistic interactions of ecosystem services: florivorous pest control boosts crop yield increase through insect pollination. Proc. R. Soc. B 283, 20152529 (2016).
Tschumi, M., Albrecht, M., Entling, M. H. & Jacot, K. High effectiveness of tailored flower strips in reducing pests and crop plant damage. Proc. R. Soc. B 282, 20151369 (2015).
Alford, D. V., Nilsson, C. & Ulber, B. Insect pests of oilseed rape crops. In Biocontrol of oilseed rape pests (ed. Alford, D. V.) 9–42 (Blackwell Science Ltd, 2003).
Williams, I. H. & Free, J. B. The feeding and mating behaviour of pollen beetles (Meligethes aeneus Fab.) and seed weevils (Ceutorhynchus assimilis Payk.) on oil-seed rape (Brassica napus L.). J. Agric. Sci. 91, 433–459 (1978).
Ekbom, B. & Borg, A. Pollen beetle (Meligethes aeneus) oviposition and feeding preference on different host plant species. Entomol. Exp. Appl. 78, 291–299 (1996).
Ahuja, I., Rohloff, J. & Bones, A. M. Defence mechanisms of Brassicaceae: implications for plant-insect interactions and potential for integrated pest management. A review. Agron Sustain Dev. 30, 311–348 (2010).
Pisa, L. et al. An update of the Worldwide Integrated Assessment (WIA) on systemic insecticides. Part 2: Impacts on organisms and ecosystems. Environ. Sci. Pollut. Res. 28, 11749–11797 (2021).
Raimets, R. et al. Synergistic interactions between a variety of insecticides and an ergosterol biosynthesis inhibitor fungicide in dietary exposures of bumble bees (Bombus terrestris L.). Pest Manag. Sci. 74, 541–546 (2018).
Sánchez-Bayo, F. et al. Are bee diseases linked to pesticides? — A brief review. Environ. Int. 89–90, 7–11 (2016).
Willow, J., Silva, A., Veromann, E. & Smagghe, G. Acute effect of low-dose thiacloprid exposure synergised by tebuconazole in a parasitoid wasp. PLoS One 14, e0212456 (2019).
European Commission. Farm to fork strategy: for a fair, healthy and environmentally-friendly food system. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions (2020).
Ulber, B., Williams, I. H., Klukowski, Z., Luik, A. & Nilsson, C. Parasitoids of oilseed rape pests in Europe: Key species for conservation biocontrol. In Biocontrol-based integrated management of oilseed rape pests (ed. Williams, I. H.) 45–76 (Springer, Netherlands, 2010).
Nilsson, C. Parasitoids of pollen beetles. In Biocontrol of oilseed rape pests (ed. Alford, D. V.) 73–85 (Blackwell Science Ltd, 2003).
Berger, J., Jönsson, M., Hedlund, K. & Anderson, P. Niche separation of pollen beetle parasitoids. Front. Ecol. Evol. 3, 45 (2015).
Kaasik, R. et al. Meligethes aeneus oviposition preferences, larval parasitism rate and species composition of parasitoids on Brassica nigra, Raphanus sativus and Eruca sativa compared with on Brassica napus. Biol. Control 69, 65–71 (2014).
Kaasik, R., Kovács, G., Toome, M., Metspalu, L. & Veromann, E. The relative attractiveness of Brassica napus, B.rapa, B. juncea and Sinapis alba to pollen beetles. BioControl 59, 19–28 (2014).
Sulg, S. et al. Spatiotemporal distancing of crops reduces pest pressure while maintaining conservation biocontrol in oilseed rape. Pest Manag. Sci. https://doi.org/10.1002/ps.7391 (2023).
Büchi, R. Mortality of pollen beetle (Meligethes spp.) larvae due to predators and parasitoids in rape fields and the effect of conservation strips. Agric. Ecosyst. Environ. 90, 255–263 (2002).
Veromann, E. et al. Effects of nitrogen fertilization on insect pests, their parasitoids, plant diseases and volatile organic compounds in Brassica napus. Crop Prot. 43, 79–88 (2013).
Veromann, E., Saarniit, M., Kevväi, R. & Luik, A. Effect of crop management on the incidence of Meligethes aeneus Fab. and their larval parasitism rate in organic and conventional winter oilseed rape. Agron. Res. 7, 548–554 (2009).
Veromann, E., Kevväi, R., Luik, A. & Williams, I. H. Do cropping system and insecticide use in spring oilseed rape affect the abundance of pollen beetles (Meligethes aeneus Fab.) on the crop?. Int. J. Pest Manag. 54, 1–4 (2008).
Begg, G. S. et al. A functional overview of conservation biological control. Crop Prot. 97, 145–158 (2017).
Bianchi, F. J. J. A., Booij, C. J. H. & Tscharntke, T. Sustainable pest regulation in agricultural landscapes: A review on landscape composition, biodiversity and natural pest control. Proc. R. Soc. B 273, 1715–1727 (2006).
Rusch, A., Valantin-Morison, M., Sarthou, J.-P. & Roger-Estrade, J. Biological control of insect pests in agroecosystems: Effects of crop management, farming systems, and seminatural habitats at the landscape scale: A review. Adv. Agron. 109, 219–259 (2010).
Thies, C. & Tscharntke, T. Landscape structure and biological control in agroecosystems. Science 285, 893–895 (1999).
Naeri-hiilamardikas. Syngenta Estonia https://www.syngenta.ee/naeri-hiilamardikas (2023).
Albrecht, M. et al. The effectiveness of flower strips and hedgerows on pest control, pollination services and crop yield: A quantitative synthesis. Ecol. Lett. 23, 1488–1498 (2020).
Beduschi, T., Tscharntke, T. & Scherber, C. Using multi-level generalized path analysis to understand herbivore and parasitoid dynamics in changing landscapes. Landscape Ecol. 30, 1975–1986 (2015).
Gladbach, D. J. et al. Crop–noncrop spillover: Arable fields affect trophic interactions on wild plants in surrounding habitats. Oecologia 166, 433–441 (2011).
Holland, J. M. et al. Semi-natural habitats support biological control, pollination and soil conservation in Europe. Rev. Agron. Sustain. Dev. 37, 31 (2017).
Skellern, M. P. & Cook, S. M. Prospects for improved off-crop habitat management for pollen beetle control in oilseed rape. Arthropod-Plant Interact. 12, 849–866 (2018).
Thies, C., Steffan-Dewenter, I. & Tscharntke, T. Effects of landscape context on herbivory and parasitism at different spatial scales. Oikos 101, 18–25 (2003).
Rusch, A., Valantin-Morison, M., Sarthou, J. P. & Roger-Estrade, J. Effect of crop management and landscape context on insect pest populations and crop damage. Agric. Ecosyst. Environ. 166, 118–125 (2013).
Zaller, J. G., Moser, D., Drapela, T., Schmöger, C. & Frank, T. Insect pests in winter oilseed rape affected by field and landscape characteristics. Basic Appl. Ecol. 9, 682–690 (2008).
Valantin-Morison, M., Meynard, J.-M. & Doré, T. Effects of crop management and surrounding field environment on insect incidence in organic winter oilseed rape (Brassica napus L.). Crop Prot. 26, 1108–1120 (2007).
Zaller, J. G., Moser, D., Drapela, T., Schmöger, C. & Frank, T. Parasitism of stem weevils and pollen beetles in winter oilseed rape is differentially affected by crop management and landscape characteristics. BioControl 54, 505–514 (2009).
Juhel, A. S. et al. Characterization of the pollen beetle, Brassicogethes aeneus, dispersal from woodlands to winter oilseed rape fields. PLoS One 12, e0183878 (2017).
Bick, E. et al. Dynamics of pollen beetle (Brassicogethes aeneus) immigration and colonization of oilseed rape (Brassica napus) in Europe. Pest. Manag. Sci. https://doi.org/10.1002/ps.7538 (2023).
Čuljak, T., Pernar, R., Juran, I., Ančić, M. & Bažok, R. Impact of oilseed rape crop management systems on the spatial distribution of Brassicogethes aeneus (Fabricius 1775): Implications for integrated pest management. Crop Prot. 89, 129–138 (2016).
Free, J. B. & Williams, I. H. The distribution of insect pests on crops of oil-seed rape (Brassica napus L.) and the damage they cause. J. Agric. Sci. 92, 139–149 (1979).
Cook, S. M., Khan, Z. R. & Pickett, J. A. The use of push-pull strategies in integrated pest management. Annu. Rev. Entomol. 52, 375–400 (2007).
Skellern, M. P. & Cook, S. M. The potential of crop management practices to reduce pollen beetle damage in oilseed rape. Arthropod-Plant Interact. 12, 867–879 (2018).
Veromann, E. et al. Fatal attraction: search for a dead-end trap crop for the pollen beetle (Meligethes aeneus). Arthropod-Plant Interact. 8, 373–381 (2014).
Kovács, G. et al. Effects of land use on infestation and parasitism rates of cabbage seed weevil in oilseed rape. Pest Manag. Sci. 75, 658–666 (2019).
Warner, D. J., Allen-Williams, L. J., Warrington, S., Ferguson, A. W. & Williams, I. H. Mapping, characterisation, and comparison of the spatio-temporal distributions of cabbage stem flea beetle (Psylliodes chrysocephala), carabids, and Collembola in a crop of winter oilseed rape (Brassica napus). Entomol. Exp. Appl. 109, 225–234 (2003).
Hawkins, B. A. & Cornell, H. V. Maximum parasitism rates and successful biological control. Science 266, 1886–1886 (1994).
Bianchi, F. J. J. A. From pattern to process: Towards mechanistic design principles for pest suppressive landscapes. Basic Appl. Ecol. 64, 157–171 (2022).
Krimmer, E., Martin, E. A., Holzschuh, A., Krauss, J. & Steffan-Dewenter, I. Flower fields and pesticide use interactively shape pollen beetle infestation and parasitism in oilseed rape fields. J. Appl. Ecol. 59, 263–273 (2022).
Frenzel, M. & Brandl, R. Diversity and composition of phytophagous insect guilds on Brassicaceae. Oecologia 113, 391–399 (1998).
Skellern Matthew, P., Clark, S. J., Ferguson, A. W., Watts, N. P. & Cook, S. M. Banker plant bonuses? The benefits and risks of including brassicas in field margins to promote conservation biocontrol of specialist pests in oilseed rape. Insects 14, 349 (2023).
Thies, C. & Tscharntke, T. Biological rape pest control in spatio-temporally changing landscapes. In Biocontrol-based integrated management of oilseed rape pests (ed. Williams, I. H.) (Springer, Netherlands, 2010).
Le Clec,h, S. et al. Effects of linear landscape elements on ecosystem services supply and trade-offs in agricultural landscapes. SSRN J https://doi.org/10.2139/ssrn.4291226 (2022).
Berger, J. S., Birkhofer, K., Hanson, H. I. & Hedlund, K. Landscape configuration affects herbivore–parasitoid communities in oilseed rape. J Pest Sci 91, 1093–1105 (2018).
There, W. E. I. A Whole earth approach to nature-positive food: biodiversity and agriculture. Science and innovations for food systems transformation 469 (Springer Cham, 2023).
Fusser, M. S., Pfister, S. C., Entling, M. H. & Schirmel, J. Effects of field margin type and landscape composition on predatory carabids and slugs in wheat fields. Agric. Ecosyst. Environ. 247, 182–188 (2017).
Pfister, S. C. et al. Positive effects of local and landscape features on predatory flies in European agricultural landscapes. Agric. Ecosyst. Environ. 239, 283–292 (2017).
Picchi, M. S., Bocci, G., Petacchi, R. & Entling, M. H. Effects of local and landscape factors on spiders and olive fruit flies. Agric. Ecosyst. Environ. 222, 138–147 (2016).
Holland, J. et al. Approaches to identify the value of seminatural habitats for conservation biological control. Insects 11, 195 (2020).
Metspalu, L. et al. Comparison of sampling methods for estimating the abundance of Meligethes aeneus on oilseed crops. Int. J. Pest Manag. 61, 312–319 (2015).
Seimandi-Corda, G., Jenkins, T. & Cook, S. M. Sampling pollen beetle (Brassicogethes aeneus) pressure in oilseed rape: which method is best?. Pest Manag. Sci. 77, 2785–2794 (2021).
Williams, I. H., Büchi, R. & Ulber, B. Sampling, trapping and rearing oilseed rape pests and their parasitoids. In Biocontrol of oilseed rape pests (ed. Alford, D. V.) (Blackwell Science Ltd, 2003).
Lancashire, P. D. et al. A uniform decimal code for growth stages of crops and weeds. Ann. Appl. Biol. 119, 561–601 (1991).
Osborne, P. Morphology of the immature stages of Meligethes aeneus (F.) and M viridescens (F.) (Coleoptera, Nitidulidae). Bull. Entomol. Res. 55, 747–759 (1965).
R Core Team. R: A language and environment for statistical computing R foundation for statistical computing. (2022).
Brooks, M. E. et al. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J. 9, 378 (2017).
Lenth, R. V. emmeans: Estimated Marginal Means, aka Least-Squares Means. (2022).
Hartig, F. & Hartig, M. F. Package ‘DHARMa’. R package (2017).
Liaw, A. & Wiener, M. Classification and regression by random Forest. R News 2, 18–22 (2002).
Therneau, T. M. & Atkinson, E. J. An introduction to recursive partitioning using the RPART routines. (1997).
Hothorn, T. & Zeileis, A. partykit: A modular toolkit for recursive partytioning in R. J. Mach. Learn. Res. 16, 3905–3909 (2015).
Acknowledgements
We would like to thank all the Quessa project partners and the farmers for letting us collect data from their fields. Special thanks go to Simone Marini, for the helpful comments and suggestions on data analysis. This study was conducted as part of the QuESSA (Quantification of Ecological Services for Sustainable Agriculture) project, which received funding from the European Union's Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 311879. The research was also funded by an Estonian Research Council grant (PRG1056).
Author information
Authors and Affiliations
Contributions
E.V., R.K., G.K., J.H. conceived the experiment(s), E.V., R.K., G.K. conducted the experiment(s), S.V. and M.L. analysed the data. S.V. wrote the first draft. R.K., M.L., G.K., J.H., E.V. reviewed and made changes to the manuscript. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Vilumets, S., Kaasik, R., Lof, M. et al. Landscape complexity effects on Brassicogethes aeneus abundance and larval parasitism rate: a two-year field study. Sci Rep 13, 22373 (2023). https://doi.org/10.1038/s41598-023-49690-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-49690-1
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.