Effect of a cover crop on the aphid incidence is not explained by increased top-down regulation

Introduction

Important regulating ecosystem services such as natural pest control in agroecosystems depend on high levels of biodiversity. Therefore, conservation practices such as those carried out with conservation biological control can increase the effectiveness of pest control by the utilization of natural enemies to reduce their mortality and provide alternative resources through manipulation of the environment (Landis, Wratten & Gurr, 2000; Gurr et al., 2017). Increased plant diversity enhances natural enemy survival and activity by providing food sources (i.e., nectar, pollen and honeydew), overwintering shelter and/or alternative prey/host species, which consequently could result in greater pest control (Landis, Wratten & Gurr, 2000; Pickett, Roltsch & Corbett, 2004; Cardinale et al., 2009; Quijas, Schmid & Balvanera, 2010; Gurr et al., 2017; Gontijo, 2019). Higher natural enemy abundances are expected in agroecosystems with a higher cultivated and spontaneous plant diversity than in more simple agroecosystems (e.g., monocultures) (Rusch, Bommarco & Ekbom, 2016; Begg et al., 2017; Perović et al., 2018). Increasing plant diversity can be achieved by different means: (1) by using intercropping, which consists of cultivating two crops at the same time, such as cereals and leguminous crops (Ben-Issa, Gomez & Gautier, 2017); (2) by adding a ground cover crop in between cultivated trees in orchards (Silva et al., 2010); (3) by adding flower strips within and around the fields (Balzan, Bocci & Moonen, 2016; Hatt et al., 2017); and (4) by allowing the growth of spontaneous vegetation around the target crop plant (Bugg & Waddington, 1994). For the three first cases, the added taxonomic plant diversity is low, with one or a few plant species added, but plant selection is aimed at attracting specific natural enemies that attack the target pest. In the latter case, the taxonomic plant diversity is higher; however, the functional diversity may not be considered in terms of the specific benefit of decreasing the target pest. Spontaneous vegetation has been shown to attract natural enemies (Denys & Tscharntke, 2002). However, positive, neutral, and even negative effects of increased plant diversity on natural enemy populations have been observed in different agricultural systems (Poveda, Gómez & Martínez, 2008; Letourneau et al., 2009) and are context-dependent (Thies & Tscharntke, 1999; Thomson & Hoffmann, 2013; Karp et al., 2018). Additionally, even when the abundance of natural enemies is enhanced, it does not always translate into a greater pest control (Andow, 1991).

Even if the spontaneous vegetation could have positive effects in some instances, it is important to select the right plant diversity (functional diversity) rather than to just increase taxonomic diversity (Gagic et al., 2015; Dainese et al., 2019). For instance, the addition of plants may create reservoirs of pests (Blitzer et al., 2012; Paredes, Cayuela & Campos, 2013) and/or enhance negative interactions among natural enemies such as intra/inter-guild competition, predation (Irvin et al., 2006; Lundgren, Wyckhuys & Desneux, 2009; Blitzer et al., 2012; Paredes, Cayuela & Campos, 2013; Gómez-Marco, Urbaneja & Tena, 2016) and hyperparasitism (Van Nouhuys & Lei, 2004; Schooler, De Barro & Ives, 2011; Jeavons et al., 2022). Primary parasitoids could be hyperparasitized by secondary parasitoids, and/or their mummies could be consumed by generalist predators (Snyder & Ives, 2001), which could reduce the strength of top-down control of pests (Sanders et al., 2011). Therefore, multiple trophic levels need to be considered to improve biological conservation methods. Moreover, studying multiple natural enemy guilds potentially participating in pest control together (Jonsson et al., 2010; Tena et al., 2015) also appears to be crucial to understanding their functional redundancy or complementary nature and may lead to management strategies to reduce negative interactions (Irvin et al., 2006; Lundgren, Wyckhuys & Desneux, 2009; Gómez-Marco, Urbaneja & Tena, 2016).

The synchronization and temporal overlapping of pests and their natural enemies are fundamental for the outcome of natural pest control (Welch & Harwood, 2014). For instance, Van Nouhuys & Lei (2004) showed a better synchronization between the parasitoid Cotesia melitaearum Wilkinson (Hymenoptera, Braconidae) and its butterfly host Melitaea cinxia L. (Lepidoptera, Nymphalidae) at the beginning of the growing season and at warmer temperatures. This resulted in greater pest control, which has been demonstrated as well for other agricultural systems (Landis & Van Der Werf, 1997; Van Nouhuys & Lei, 2004). By contrast, when synchronization between natural enemies and pests is weak (Landis & Van Der Werf, 1997; Landis, Wratten & Gurr, 2000; Yang et al., 2017), a slight delay in the arrival of natural enemies can lead to a high pest growth rate and to ineffective biological control (Raymond, Ortiz-Martínez & Lavandero, 2015; Yang et al., 2017). To ensure synchronization and efficient pest control, it is thus essential that natural enemies arrive before their host or prey (Landis & Van Der Werf, 1997; Landis, Wratten & Gurr, 2000; Yang et al., 2017).

In perennial crops like fruit orchards, the early arrival of natural enemies was shown to be possible by the establishment of ground cover crops, i.e., single plant species or a mix of plants sown in the inter-rows between fruit trees or growing spontaneous vegetation (Simoes et al., 2014; Gómez-Marco, Urbaneja & Tena, 2016; Bowers et al., 2020). Cover crops can increase the abundance of natural enemies (Aguilar-Fenollosa et al., 2011; Aguilar-Fenollosa & Jacas, 2013) and reduced pest populations (Dong et al., 2005; Irvin et al., 2006; Aguilar-Fenollosa et al., 2011; Gómez-Marco et al., 2016; Gómez-Marco, Urbaneja & Tena, 2016). For instance, in peach orchards, a cover crop of Medicago sativa L. (Fabaceae) increased predator densities, especially spiders, while reducing the population of leaf miner Lyonetia clerkella L. (Lepidoptera, Lyonetiidae) (Dong et al., 2005). However, few studies have evaluated the effect of intercropping on functional redundancies and niche complementarities among natural enemy guilds, as most of these studies have focused on single natural enemy guilds. In addition, these studies were performed during the growing season, and there are no studies on the effect of increasing populations of natural enemies in winter with alternative hosts to enhance their arrival and control pest populations in spring. Among the many pest species attacking plum orchards, aphids are the most important (Symmes et al., 2012; Cichon et al., 2013). They can cause direct as well as indirect damage, as they are the main viral disease vectors of the plum pox virus (Dragoyski, Stefanova & Kamenova, 2011) with resulting important economic losses (Cambra et al., 2006).

In this study, we investigated whether intercropping an oat cover within orchards before winter may induce the early arrival of natural enemies by providing alternative hosts and thus promote the biological control of aphid pests infesting plum orchards in spring compared to spontaneous vegetation. Previous studies have shown that aphid populations on cereals during the winter in Chile, as well as their natural enemies such as parasitoids and predators, are present (Ortiz-Martínez & Lavandero, 2018; Alfaro-Tapia et al., 2021). In addition, all aphid species attacking cereals and other wild graminaceous plants have not been observed to attack or damage plum trees shoots (Blackman & Eastop, 2000; Blackman & Eastop, 2007). However, plum aphids and aphids feeding in wild and cultivated Poaceae share similar natural enemies (Starý, 1995). Moreover, from our previous studies we can conclude that the main parasitoid species, Aphidius platensis (Hymenoptera, Braconiidae) of both plum and cereal aphids such as Myzus persicae and Rhopalosiphum padi (Hemiptera, Aphididae) respectively, are capable of shifting between aphid hosts (winter vs. spring host) (Alvarez-Baca et al., 2020). We hypothesize that (1) inter-cropped oat between plum trees attracts cereal aphids and their associated shared natural enemies during the winter, which leads to (2) an early arrival of shared natural enemy populations compared to the arrival of the main prevalent plum aphids in early spring. As a consequence, we expected (3) a lower aphid incidence on plum trees in the treatment with an oat cover crop, (4) due to a higher natural enemy abundance and parasitism rates on plum trees with no effect of hyperparasitism on the primary parasitoids.

Materials & Methods

Field experiments were approved by ANID (Agencia Nacional de Investigación y Desarrollo- Chile) project number 1180601. Farms belonged to a single owner who willingly accepted to participate in the study and submitted their approval an allowance for using the field sites to FONDECYT- ANID (Fondo Nacional de Ciencia y Tecnología. Chile) project number 1180601 ANID biosecurity committee.

Study area and experimental design

The study area was located at the district of Codegua, region of O’Higgins, in the Central Valley of Chile (34°08′S; 70°38′W). Four organic plum Prunus domestica L. (Rosaceae) cv. ‘D’Agen’ farms with the same rootstock and cultivar, with similar management and age structure (all orchards were planted from 2009 to 2013), were selected. Each farm was at least 10 ha. All orchards were managed following organic production guidelines, and neither synthetic pesticides nor fertilizers were used. The Central Valley of Chile is characterized by a temperate Mediterranean climate, with dry summers and mild, rainy winters (Sarricolea, Herrera-Ossandon & Meseguer-Ruiz, 2017). Temperatures vary from 25 to 35 °C in spring–summer (September–March) and between 3 to 13 °C in winter, with precipitation ranging from 22 to 130 mm during spring and from 300 to 900 mm in winter (June–August) (Montes et al., 2012; Directorate General of Civil Aviation of Chile, 2018). In each farm, two treatments were established with four replicates (each replicate consisted of a plot of 1 ha), resulting in a total of eight plots (see Table S1A for geographic coordinates of each plot). The treatments were as follow: the oat cover crop (OCC) treatment, consisting of four consecutive inter-rows of oat, Avena sativa L. (Poaceae) of at least 100 m long, with inter-rows sown during the second week of May in autumn, and the treatment without oat corresponding to four inter-rows with spontaneous vegetation (SV) (Fig. 1). Spontaneous vegetation rows consisted of the naturally occurring plants, which were periodically cut with a rotary cutter. Each tree row was separated by 5 m, and the space between plum trees along the row was 4 m (Fig. 1), with minimum, maximum and average distances between plots of 104.00 m, 665.58 m and 358.57 m respectively. (see Table S1B for more details). Tree management included regular mowing and pruning prior to the beginning of the flowering in the spring season. All four farms received similar management as they were all under the same company and certification guidelines. During spring, the SV treatment plots presented a patchily distributed presence of weeds such as Malva spp., Anoda hastata Cav. (Malvaceae), Taraxacum officinale L. (Asteraceae) and graminaceous species such as Poa annua L., Lolium spp. (Poaceae) (see also Table S2). Both treatments in each plum plot were established at least 10 rows away (about 50 m) from each other to avoid interaction between treatments.

Results

During the winter, on the plum trees, we did not find any aphid eggs or any nymphs or adults. On the inter-rows, aphids, parasitoids and hyperparasitoids arrived from the first date of winter sampling (July 10th) and increased as the season progressed (Table 1). Aphids were present in the OCC treatment during all sampling dates. For the parasitoids, we observed a similar pattern, as they were present from sampling date 1 through all the dates as the season progressed, with a higher incidence in the OCC treatment. Hyperparasitoids appeared at the fifth sampling date in the OCC treatment and remained present until the end of spring, whereas in the SV treatment, they appeared earlier: once at the first sampling date and again at the fourth sampling date until the end of the season. In both treatments, their presence was scarcer when compared to that of parasitoids.

During the spring, a total of 4,865 aphids in the SV treatment were recorded, from which 76.81% corresponded to Brachycaudus helichrysi, 19.67% to Aphis spiraecola and 3.51% to Myzus persicae. Additionally, 2,752 aphids were collected in the OCC treatment; among these, 66.13% corresponded to B. helichrysi, 27.94% to A. spiraecola and 5.92% to M. persicae. The total aphid incidence was 3% in the SV treatment and 1% in the OCC treatment. The aphid incidence in plum trees differed per sampling date (GLMM: χ2 = 114.16, Df = 4, p < 0.0001) (Fig. 2A) and per treatment (GLMM: χ2 = 19.45, Df = 1, p < 0.001) (Fig. 2B), with more aphids in the SV treatment at all dates and the highest incidence of aphid infested shoots at the end of spring with no significant interactions (p = 0.87). There were no significant differences in the proportion of shoots with parasitized aphids (incidence of parasitism) between sampling dates (GLMM: χ2 = 7.50, Df = 4, p = 0.11) (Fig. 2C) or among treatments (GLMM: χ2 = 3.62, Df = 1, p = 0.06) (Fig. 2D). The proportion of shoots with parasitized aphids tended to increase in both treatments as the season went on until the last sampling date with a very low parasitism incidence (no interaction, p = 0.99). In addition, the incidence of hyperparasitism did not differ between sampling dates (GLMM: χ2 = 1.30, Df = 2, p = 0.52) (Fig. 2E) or between treatments (GLMM: χ2 = 0.47, Df = 1, p = 0.49) (Fig. 2F). Nevertheless, it tended to show a similar pattern to the incidence of parasitism, with a slight increase over the season in both treatments, with the highest incidence during the last sampling date (no interaction, p = 0.10).

From all the natural enemies recorded, adult aphid parasitoids were the most abundant, followed by carabid beetles, coccinellid beetles, spiders and hoverflies (Table S4). For the total natural enemy abundance, there were differences among sampling dates (GLMM: χ2 = 156.99, Df = 4, p < 0.0001) (Fig. 3A) but not between treatments (GLMM: χ2 = 2.50, Df = 1, p = 0.11) (Fig. 3B), with an increase in the abundance over time in both treatments (no interaction p = 0.25). Nevertheless, coccinellid beetles increased over time (GLMM: χ2 = 172.27, Df = 4, p < 0.0001) (Fig. 3C) and were more abundant in SV than in OCC (χ2 = 29.10, Df = 1, p < 0.0001) (Fig. 3D). A higher abundance for the SV treatment was observed mainly on dates 7 and 8 compared to the OCC treatment, without any significant interactions (p = 0.14). For carabid beetles, we found differences in the sampling date (GLMM: χ2 = 61.76, Df = 4, p < 0.0001) (Fig. 3E) but not between the treatments (GLMM: χ2 = 0.05, Df = 1, p = 0.82) (Fig. 3F), with an important increase at the end of the season (last sampling date) (interaction, p = 0.46). For the adult aphid parasitoids, significant differences among sampling dates (GLMM: χ2 = 72.84, Df = 4, p < 0.0001) (Fig. 4A) but not between treatments (GLMM: χ2 = 0.98, Df = 1, p = 0.32) (Fig. 4B) were found, for which the highest abundance was during date 7, but this decreased for the last sampling date (no interaction, p = 0.87). In the case of the hoverflies, there were no differences between sampling dates (GLMM: χ2 = 5.83, Df = 4, p = 0.21) (interaction, p = 0.73) (Fig. 4C) or among treatments (GLMM: χ2 = 0.22, Df = 1, p = 0.64) (Fig. 4D). Finally, there were differences in the abundance of spiders among sampling dates (GLMM: χ2 = 28.82, Df = 4, p < 0.0001) (Fig. 4E), but this did not differ between treatments (GLMM: χ2 = 0.00, Df = 1, p = 0.96) (Fig. 4F). In both treatments, we observed the highest abundance at the end of the season (interaction, p = 0.29).

Discussion

Our results showed that the OCC treatment attracted cereal aphids and their associated natural enemies during the winter as proposed by our first hypothesis, which leads to the early arrival of shared natural enemies prior to the arrival of the plum aphids (second hypothesis). In concordance with our third hypothesis, we found a lower aphid incidence in the OCC treatment compared to the SV treatment on the plum trees. However, this was not due to a higher parasitism incidence or higher abundances of natural enemies. Moreover, no effect on the fourth trophic level was found, as we have no evidence for greater hyperparasitism of aphid colonies on plum trees in any of the two treatments (fourth hypothesis). Although the results provide evidence that the oat inter-row can be beneficial to decrease aphid densities, this is not due to an increase of activity or abundance of natural enemies.

During winter, we ensured the early arrival of aphid parasitoids on the oat inter-rows before the arrival of the target pest early in spring. Cover crops are used as overwintering resources and can harbor alternative host species for aphid parasitoids during a considerable time span (Welch & Harwood, 2014). Assuming that parasitoids on the cover crops will eventually move from the cover crop to the adjacent crop plants (Landis, Wratten & Gurr, 2000; Welch & Harwood, 2014), this would therefore increase biological control avoiding pest outbreaks (e.g., aphids damaging fruit crop shoots). During spring, there were more parasitoids in the inter-rows in the OCC treatment, which acted as a great source of alternative hosts (cereal aphids) when compared to the SV treatment. However, even when the aphid incidence on plum trees was lower in the OCC treatment (56% less) compared to the SV treatment, this was a result of neither the parasitism rate nor the predator abundance (top-down effects). One possible explanation would be that the cereal aphid parasitoids do not switch in sufficient numbers to plum trees to control plum aphids, although in the laboratory this has been shown to be possible (Alvarez-Baca et al., 2020). It is also possible that parasitoids showed a greater preference for cereal aphids compared to plum aphids, being strongly influenced by the host from which they emerged (Morris & Fellowes, 2002).

When comparing the oat cover crop to spontaneous vegetation, the resource specialization hypothesis predicts that the increased plant diversity enhances the diversity of higher trophic levels by favoring species specialized on the additional resources (Hutchinson, 1959). If the spontaneous vegetation present is abundant and persistent (which was not the case for our study), greater pest control could be achieved as: (1) a more diverse vegetation would attract more diverse natural enemies (top-down effect) (Root, 1973; Poveda, Gómez & Martínez, 2008), and (2) a patchy/complex distributed area would reduce herbivore populations (Root, 1973). However, the OCC treatment, which had only one plant species (oat) but with a high plant density and coverage, and which was specifically sown to attract cereal aphids, showed a greater effect on aphid plum populations. It was showed that functionally diversified cover crops that increase functional diversity (attraction of parasitoids and predators that affect the target pest species) could enhance ecosystem services as pest control (Heemsbergen et al., 2004; Sattler et al., 2020). In our study system, the oat inter-row had not only a more homogeneous cover but also a higher functional value than the spontaneous vegetation. In the SV treatment, wild graminaceous plants, that can be attacked by cereal aphid hosts for target parasitoids, were reduced in proportion compared to introduced weeds from other plant families (Malvaceae, Asteraceae, Convolvulaceae). Moreover, the other weeds also did not sustain important populations of alternative hosts of the target natural enemies of this study (see results, Fig. 3).

In our study, two explanations can be provided to explain the absence of a link between the higher abundance of natural enemies and a decrease of aphid populations. The first one is the possible negative interactions between natural enemies, and the second one is linked to bottom-up effects. Negative interactions among natural enemies may have disrupting effects on pest suppression, interactions such as intraguild predation (i.e., two predator species share the same prey species and could also feed on each other) (Polis, Myers & Holt, 1989; Lucas, 2005; Meisner et al., 2011; Traugott et al., 2012) and hyperparasitism (Finke & Denno, 2005; Jeavons et al., 2022). Our study would suggest that there is no relationship between the main predator abundances and the incidence of parasitism. Although coccinellid beetle abundance was higher in the SV treatment than in the OCC treatment, no measurable difference for the parasitism rates was found. However, we do not provide any direct evidence of intraguild predation or coincidental intraguild predation in this study; therefore, we cannot rule out this possibility. An increased abundance of the natural enemies present in SV treatment could be related to the increased abundance of their prey (Kandel, Tilmon & Shuster, 2016; Reznik et al., 2017). Likewise, hyperparasitoids are known to have negative impacts on parasitoid survival rates, with disrupting consequences on the population build-up of the primary parasitoid species affecting the outcome of biological control (Sullivan & Völkl, 1999; Tougeron & Tena, 2019). For instance, in Nagasaka, Takahasi & Okabayashi (2010), hyperparasitism rates on a group of Aphidiinae species, mainly Aphidius colemani Viereck (Hymenoptera, Braconidae) in sweet pepper and eggplant greenhouses using banker plants varied from 35% to 70% along the season, reducing the primary parasitism rates to less than 20% in one of the four years of sampling, with negative consequences on the control of aphids. By contrast, in our study, hyperparasitism did not seem to explain the absence of differences in parasitism rates in both treatments. These findings are in agreement with those of a study by Plećaš et al. (2014), where no differences of hyperparasitism rates (average of 10% or less) between simple and complex landscapes or between large and small-field landscapes were found, with no repercussion on the control of cereal aphids by parasitoids.

Another explanation is that the oat cover crops had a negative effect on the plum aphids through chemical and/or physical stimuli (bottom-up effects). During their host-plant selection process, herbivorous insects have to cope with different semiochemical cues from the plants as well as plant physical characteristics (i.e., color, shape and texture) (Bernays & Chapman, 1994). In addition, cover crops may release chemicals that affect movement and feeding of aphids (Beizhou et al., 2011), for instance, there is previous evidence of A. sativa as a banker plant acting as a repellent on the same aphid species if pre colonized (Glinwood & Pettersson, 2000) and therefore could possibly also be repellent to other aphid species. They could also interfere with their ability to locate their host plant due to a lack of a clear olfactory stimuli resulting from the release of odor masking substances (Kennedy, Booth & Kershaw, 1961). Therefore, in our study, we speculate that oat in the OCC treatment might not allow plum aphids to reach plum trees because of odors emitted, making the crop difficult to be perceived by the herbivore, supporting the associational resistance hypothesis (Tahvanainen & Root, 1972), whereas in the case of the SV treatment the lesser vegetational density, as plants with available aphid hosts were less present between rows, supports the resource concentration hypothesis (Root, 1973). The pest density is reduced in a more diverse and patchy habitat (Root, 1973; Poveda, Gómez & Martínez, 2008), including a more diverse chemical habitat due to mixed volatiles from different plant species (Price et al., 1980).

On the other hand, cover crops, as well as providing refuge and alternative hosts, can be a sugar source for adult parasitoids through the provision of nectar and/or the honeydew produced by the aphids (Bugg & Waddington, 1994; Irvin et al., 2006; Vollhardt et al., 2010; Balzan, Bocci & Moonen, 2016; Luquet et al., 2021). In agroecosystems, nectar and honeydew are the most common available sources of sugar for natural enemies, especially parasitoids, which require non-prey food as part of their diet to increase their survival and fecundity (Wäckers, Van Rijn & Heimpel, 2008; Vollhardt et al., 2010). In our study, the OCC treatment clearly provided parasitoids a higher and constant source of honeydew compared to the SV. Previous studies have shown that honeydew can be as good as flower nectar in terms of quality (Vollhardt et al., 2010; Monticelli et al., 2020; Rand & Waters, 2020), as well as being the predominant source of sugar depending on the system (e.g., Luquet et al., 2021). Therefore, the OCC treatment would offer more resources, such as sugar and alternative hosts, to parasitoids. The high coverage area of the oat through all the inter-row and the temporal availability suggest that parasitoids could possibly remain foraging in the cover crop without dispersing to the plum trees (Table 1). In contrast, the SV treatment, patchy distribution with less cover would offer fewer resources, forcing parasitoids to disperse. However, whether they disperse to the plum trees instead of other crop systems is not clear.

Even when both spontaneous vegetation and cover crops habitat management strategies are beneficial for biological control, we highlight the importance of focusing on a functional plant value instead of a taxonomic diversity (Gagic et al., 2015). Spontaneous vegetation can be easily promoted by farmers, does not require soil management, and can provide some functional plant diversity. However, low plant coverage and other problems such as the proliferation of invasive plant species or plants that act as a source of an insect pest could limit its usefulness (Venzon et al., 2019). In our study, although spontaneous vegetation did provide natural enemies, the decrease of plum aphid abundance was only achieved by the use of A. sativa. When carefully selected, inter cropping can offer many benefits such as: improving soil health by fixing nitrogen (Blanco-Canqui, Claassen & Presley, 2012), protecting soil from erosion, enhancing soil organic matter by providing a high biomass production (Isik et al., 2009; Nielsen et al., 2015), increasing water quality (Dabney, Delgado & Reeves, 2001), preventing weed plant growth (Brust, Claupein & Gerhards, 2014; Finney et al., 2017) by releasing growth inhibitors at high concentrations in the roots and shoots (Kato-Noguchi et al., 1994; Kato-Noguchi, Mizutani & Hasegawa, 1994). Specifically, there is evidence that A. sativa can have other benefits in addition to those previously mentioned such as: increasing mycorrhizal fungi populations and microbial biomass (Benitez, Taheri & Lehman, 2016), increasing earthworm populations compared to plots without cover crops (Roarty, Hackett & Schmidt, 2017). Moreover, as well as other cereals, oat is part of the banker plant system that provides alternative hosts for a parasitoid or predator of a target crop pest (Micha et al., 2000; Andorno & López, 2014) (for more beneficial examples, see Table S5). All of these ecosystem services, in addition to promote an early establishment of natural enemies, and/or to have repellent effects on the plum aphid populations, are key to develop a more sustainable approach.

Conclusions

We can conclude that oat cover crops contribute positively to the control of aphid populations in plum orchards through an apparent bottom-up effect through physical and/or olfactive barriers for plum-associated aphids from the oat cover crop. In addition, it is important to mention that future studies should consider the effect of this or other cover crops on the assemblage of parasitoids and predators to fully understand the dynamics and consequences on aphid control and their natural enemies. Further studies on the species composition and food web structure in orchards with or without cover crops are needed to unravel the direct and indirect effects of these interactions on pest control.

Supplemental Information