Compatibility of organic farming treatments against Monosteira unicostata with non-target arthropod fauna of almond trees canopy

Field trials had shown that 1-2 applications of kaolin and potassium salts of fatty acids combined with thyme essential oil (PSTEO) reduced the abundance of the lace bug Monosteira unicostata (Mulsant & Rey) (Hemiptera: Tingidae), an important pest of almond trees in the Mediterranean region. These products could be useful for the control of this pest in organic production of almonds, but higher number of applications could be necessary. However, the possible detrimental effects on the almond orchard ecosystem should be evaluated. In the present work, the effects observed on the non-target arthropod fauna of the almond trees canopy in those field assays are shown. First, a comprehensive report of the non-target arthropod fauna of the almond tree is provided. Regarding natural enemies, most of the predatory arthropods captured were spiders belonging to different families like Salticidae, Thomisidae, Philodromidae, Theridiidae, Araneidae or Oxyopidae. Other predatory families that appeared in significant numbers were Chrysopidae, Anthocoridae, Aeolothripidae, Coccinellidae, Phytoseiidae, Erythraeidae or Forficulidae. Among parasitoids, the most abundant families were Eulophidae, Scelionidae and Dryinidae. Kaolin reduced the abundance of natural enemies and other non-target arthropods as well as their diversity and number of species. On the contrary, PSTEO only produced a slight reduction in the number of natural enemies, whereas no effect was found on the diversity and species richness. These effects were observed despite the reduced number of applications, so greater effect is expected if its frequency is increased in order to achieve an efficient control of M. unicostata. Additional keywords: kaolin; insecticidal soaps; thyme essential oil; organic almond production; Prunus dulcis; natural enemies Abbreviations used: PRC (principal response curve analysis); PSTEO (potassium salts of fatty acids combined with thyme essential oil). Authors’ contributions: MGN, ISR and SP designed the research and wrote the manuscript. AM, GC, MGN, ISR and SP sampled arthropods. AM, MGN, ISR, SP and CEF classified arthropods. ISR analyzed the data. Citation: Sánchez-Ramos, I.; Marcotegui, A.; Pascual, S.; Fernández, C. E.; Cobos, G.; González-Núñez, M. (2017). Compatibility of organic farming treatments against Monosteira unicostata with non-target arthropod fauna of almond trees canopy. Spanish Journal of Agricultural Research, Volume 15, Issue 2, e1004. https://doi.org/10.5424/sjar/2017152-10515 Supplementary material (Tables S1, S2, S3 and S4) accompanies the paper on SJAR’s website. Received: 23 Sep 2016. Accepted: 24 May 2017 Copyright © 2017 INIA. This is an open access article distributed under the terms of the Creative Commons Attribution (CC-by) Spain 3.0 License. Funding: Spanish Ministry of Agriculture, Food and Environment (project AEG-08-021-2) Competing interests: The authors have declared that no competing interests exist. Correspondence should be addressed to Manuel González-Núñez:: mgnunez@inia.es


Introduction
The interest in organic foods grows permanently worldwide, as does the area devoted to this type of agriculture. Organic agriculture land has quadrupled since 1990, reaching a total area of 43.7 million hectares in 2014 (Willer & Lernoud, 2016).
Nuts are among the main permanent organic crops in the world, surpassed only by coffee and olives. In 2014 the worldwide area for organic nuts was 286,109 ha (Willer & Lernoud, 2016) and an important portion of this (94,646 ha) corresponded to Spain where the main organic nut crop is almond (85,241 ha) (MAGRAMA, 2015).
The production of certified organic food requires practices authorized by organic production standards (IFOAM, 2014;EC, 2015;USDA, 2016). Crop protection in organic agriculture is based on preventive strategies (well adapted varieties, balanced fertility, rotations, companion planting, green manures, functional biodiversity, habitat management, beneficial organisms) and insecticide application is restricted to products of mineral or vegetable origin included in the standards that are used as a last option when prevention has failed (Zehnder et al., 2007). Some studies have been conducted recently to search for products allowed in organic production that would be effective for controlling M. unicostata in almond crops. Sánchez-Ramos et al. (2014) found out, in laboratory tests, that kaolin sprayed on almond leaves reduced oviposition plus adult and nymphal feeding, while potassium salts of fatty acids combined with thyme essential oil caused high mortality of nymphs. The effect of these products against this pest was also tested in field trials (Marcotegui et al., 2015). In this case, both products reduced abundance of M. unicostata and the damage on leaves, being kaolin the most effective.
To conserve or improve biodiversity is a general objective of organic cropping and enhancing abundance and efficacy of the natural enemies existing community is a priority for this production system (Zehnder et al., 2007;IFOAM, 2014). Therefore, the potential impact of any insecticide application on natural enemies arthropod fauna must be previously investigated. Natural biological control of M. unicostata has not been deeply studied. Although different groups of predators (like Araneae, Forficulidae, Chrysopidae, Miridae, Anthocoridae, Coccinelidae, Carabidae, and predatory Thysanoptera, Hymenoptera and Diptera) and parasitoids (like Mymaridae) have been reported as possible natural enemies of M. unicostata (Vessia, 1961;Moleas, 1987) only Araneae and Anthocoridae have shown an obvious predatory activity against this pest in the field (Moleas, 1987).
In the present work, the effects of these organicfarming-compatible pests control products on nontarget arthropods of almond tree canopy were evaluated, under real field conditions, paying special attention to the community of natural enemies of pests.

Field trials design
Field trials were conducted in 2009 and 2010 in commercial almond orchards in Murcia, Spain. The experimental design was as described in Marcotegui et al. (2015). Briefly, plantations located in Cieza (cultivars Ferragnes and Feraduel) and Cehegín (cultivar Antoñeta) were used in 2009 and 2010 respectively. In both cases the products tested were: 1) Kaolin: Surround  WP (95% w:w (wettable powder) kaolin) (BASF Aktiengesellschaft, Ludwigshafen, Germany) sprayed at a dose of 5 kg/100 L. Two applications were given each year: the first one in midspring, aimed to prevent colonization of overwintering adults of M. unicostata and the second one in early summer, to guarantee coating until the end of the crop season.
2) Potassium salts of fatty acids combined with thyme essential oil (PSTEO): OleatBio-to (40% w:w soybean and sunflower fatty acids; 5% w:w potassium salts; 6% w:w thyme essential oil) (TRABE S.A., Murcia, Spain) was sprayed at a dose of 300 mL/100 L. An application against nymphs of the second generation of M. unicostata was given at the end of spring in both years, but in 2010, the populations of the lace-bug were much higher and an additional application was required against the nymphs of the third generation in mid-summer.
Products were sprayed at the maximum field recommended concentrations in Spain (De Liñán, 2013a,b). The experimental design consisted of randomised blocks with four (2009) or seven (2010) replications. Within each block, four contiguous trees of each cultivar were randomly assigned to each treatment. See Marcotegui et al. (2015) for details.

Assessment of abundance of arthropods
The overall arthropod fauna from the canopy of almond trees was sampled using a beating method. Arthropods sampled from the four trees of each treatment were collected in a plastic bag. Beating sampling was performed monthly in spring and summer, with a total of five and six sampling dates in 2009 and 2010, respectively.
Samples were taken to the laboratory and the specimens were assigned to the following groups: 1) Natural enemies: Those belonging to families whose main feeding habit is parasitism or predation. 2) Phytophagous (target arthropods): Those causing economic damage on almond trees (pests). This group was studied in Marcotegui et al. (2015).
3) Other arthropods: Those phytophagous not described as pests on almond trees and specimens with other feeding habits or that could not be allocated to a specific feeding guild.
This work focuses on non-target arthropods, i.e. groups 1 and 3. Specimens were determined to family level when possible. Biodiversity was assessed by the number of morphospecies and the Shannon biodiversity index (Magurran, 2004).

Data analysis
The effect of the factors considered on the number of individuals captured, the number of species and the Shannon biodiversity index was tested by linear mixed-effects models (Littell et al., 1998;Wang & Goonewardene, 2004). Treatment and cultivar (only in 2009) were considered fixed factors, with block as a random factor and sampling date as a repeated measures factor. Interactions among all fixed factors were also considered in the models. The best covariance structure for the repeated-measures (date) factor was selected according to the lowest value of the Akaike and Schwarz's Bayesian information criteria fit statistics (Littell et al., 1998;Wang & Goonewardene, 2004). The models were fitted using a restricted maximum likelihood estimation method. If convergence was not achieved or the final Hessian matrix was not positive definite, the random factor was removed from the model as it was identified as redundant variable. When necessary, data were previously transformed by ln(x+1) for normality. The significance level was always p < 0.05. Statistical tests were performed using SPSS statistical program.
Analyses were performed for the periods before and after the first treatment application. In the first case, to verify the absence of significant differences among plots and in the second case, to examine the effect of the treatments on abundance and diversity of non-target arthropods. Differences in abundance and diversity of non-target arthropods among plots assigned to each treatment were evaluated separately against the untreated control in 2009 because of the different application schedule of kaolin and PSTEO. For 2010 data, differences with regard to the control were established by an LSD test when statistical significance was found.
To investigate changes in abundance and species composition of the non-target arthropod community in the canopy of almond trees, a principal response curve (PRC) analysis was performed using the program CANOCO 4.51 (Van den Brink & Ter Braak, 1999;Leps & Smilauer, 2003). The significance of the deviations from the line representing the untreated control (y=0), because of each treatment, was tested using an F-type permutation test (Monte Carlo simulation) with 499 permutations. PRCs for kaolin and PSTEO plots in relation to the untreated controls were obtained before and after the first treatment applications for each year. Additionally, to determine treatment effects on different taxa, 'species weights' were also considered in those cases in which the PRCs were significant. Data on the number of captures of each taxon were transformed to ln(x+1) before analysis.

Parasitoids
In both years, most of the parasitoids captured were immature stages belonging to order Hymenoptera (46-54% in 2009, 24-53% in 2010) that could not be determined to family level (Table S3 [suppl.]). They were either pupae or larvae found in parasitized hosts.

Other arthropods
The remaining arthropod community was very diverse and it was represented by more than forty families of eleven orders of classes Arachnida (infraclass Acari), Entognatha and Insecta (Table S4 [ suppl.]). In both orchards, Thysanoptera was the most abundant group, but its percentage was much lower in 2009 (28-32%) compared to 2010 (80-91%), due to the very high number of thrips captured in 2010 (∼1,400-1,800). This made also that the percentages of the other groups in 2010 were very low compared to the percentages obtained in 2009, though the number of specimens captured was more or less similar. Also in both years, the second group in order of importance was Coleoptera (22-23% in 2009, 4-15% in 2010). In 2009, the next most abundant groups were Hemiptera (9-14%), Diptera (9-12%), Psocoptera (6-11%) and Hymenoptera (3-6%). This ranking of abundance was slightly modified in 2010: Psocoptera (1-2%), Hemiptera (1-2%), Hymenoptera (0-3%) and Diptera (∼1%). The remaining groups appeared in low percentages (∼2 or lower in 2009 and <0.6% in 2010). Immature stages of different orders that could not be assigned to any family were classified as "Not identified" within the corresponding order. Also, a significant number of eggs were classified as "Not identified Arthropoda" because they could not be assigned to any taxonomic category (6-11% in 2009 and less than 2% in 2010).

Effect of kaolin and potassium salts of fatty acids combined with thyme essential oil (PSTEO) on non-target arthropods
Before the treatment applications both in 2009 and 2010, no significant differences were found between kaolin or PSTEO plots and untreated control plots in abundance of non-target arthropods except for the natural enemies in the PSTEO plots in 2009, which showed lower values compared to the control plots (Table 1). After treatment application, it was found a significant reduction in the kaolin-treated plots in the abundance of natural enemies in 2009 and 2010 and in the abundance of other arthropods in 2010 compared to the control plots (Table 1). PSTEO only produced a significant reduction in the number of natural enemies in 2010.
Concerning diversity and number of species of non-target arthropods, no significant differences were found between kaolin or PSTEO plots and the untreated control plots before treatment application both in 2009 and 2010 (Table 1). After treatment application, a significant reduction in the Shannon diversity index and in the number of species was observed in the kaolin plots compared with the control plots in both years except for the Shannon index in 2010.
No differences in the community composition of nontarget arthropods were found among treated and control plots before treatment application either in 2009 and 2010 (PRC analysis, p > 0.05) ( Table 2). After treatment application, no effect was observed for both Kaolin and PSTEO-treated plots compared to the control (p > 0.05) in 2009, but a significant effect was observed in 2010 in the PSTEO-treated plots for the natural enemies community and in the kaolin-treated plots for the other non-target arthropod community (Table 2).
In those cases where the PRC analysis was significant, the contribution of different taxa to nontarget arthropod community response in the treated plots is revealed by the species scores obtained. Taxa with a positive weight over 0.5 are expected to decrease in abundance compared to the control after treatment application. In the case of the effect of PSTEO on natural enemies in 2010, immature stages of parasitoids had the highest score (4.2), and the next most affected taxa were Anthocoridae (0.8) and Phytoseiidae (0.6). Table 1. Abundance, Shannon index and number of species per sample of non-target arthropods (natural enemies and other arthropods) captured by beating in almond trees before and after being sprayed with kaolin or potassium soap with thyme essential oil (PSTEO) and in the untreated trees

Plots
Natural enemies a Other arthropods a Shannon index a Number of species a Before After Before After Before After Before After

Discussion
One of the most important issues in integrated pest management strategies is the proper use of all the available methods to control pests, taking into account environmental concerns, like the possible negative effects of these methods on non-target fauna (Barzman et al., 2015). In this respect, it is necessary to assess the influence of control strategies on those beneficial organisms that contribute to maintain populations of damaging arthropods under economic thresholds. This is especially important in organic production systems, because the availability of allowed products for pest control is more reduced than in conventional production.
There is few information about the composition of the natural enemies' community of almond orchards. Regarding predators, our results reasonably agree with those reported by Benhadi-Marín et al. (2011), who pointed out Araneae as the most abundant group, followed by Coleoptera (mainly from Family Coccinellidae), Formicidae, Neuroptera, Hemiptera and Dermaptera. Other authors also report the occurrence of predators belonging to these groups in almond trees (Bolu, 2007;Eilers & Klein, 2009;Yanik et al., 2011;Santos et al., 2012). However, in none of these works the occurrence of predatory Acari, Thysanoptera or Diptera is reported, but Hoy et al. (1979) collected phytoseiid mites and predatory thrips in almond orchards, as well as Neuroptera and predatory Coleoptera. Of the families of spiders collected by Benhadi-Marín et al. (2011), Philodromidae, Salticidae, Theridiidae, Thomisidae, Araneidae, Oxyopidae and Gnaphosidae were also present in our orchards. Concerning parasitoids, in the work by Eilers & Klein (2009) the families Bethylidae and Encyrtidae comprised the majority of parasitoids collected, with only seven morphospecies captured, whereas we have collected up to 16 families of hymenopteran parasitoids. Regarding the possible natural enemies of M. unicostata, most of the groups reported by different authors (Vessia, 1961;Moleas, 1987) have been collected in our assays, and those that have been proved to feed on this lace bug in the field (Araneae and Anthocoridae) (Moleas, 1987) are well represented in both years of study.
In the search of more environmentally friendly strategies to control almond pests in organic production, some compounds like the aluminosilicate mineral kaolin, the insecticidal soaps based on potassium salts of fatty acids and plant essential oils have shown relatively good results in laboratory and field assays (Braham et al., 2014;Sánchez-Ramos et al., 2014;Marcotegui et al., 2015). However, although these products are considered to have safe environmental profiles (Weinzierl & Henn, 1991;Glenn et al., 1999;Isman, 2006;Markó et al., 2008;Regnault-Roger et al., 2012), they cannot be considered as selective pesticides and their impact on non-target arthropod fauna should be evaluated.
In this work, we have observed some detrimental effects of kaolin and a combination of potassium salts of fatty acids and thyme essential oil on non-target arthropod fauna. Kaolin produced the most negative effects, since this product reduced the abundance of beneficials and other arthropods, and arthropod diversity and species richness in both years of the study. PSTEO only produced a slight decrease in abundance of natural enemies in the second year. Curiously, abundance of natural enemies in 2009 was significantly higher in control plots than in PSTEO plots before treatment application. This could be considered a random effect, since plots were assigned randomly to each treatment. In addition, no significant differences were observed after treatments.
With the aim of determining the most affected taxa after the treatment applications, PRC analyses were performed. However, the differences observed regarding abundance and diversity were not fully reflected in the PRC results, because significant effects were only detected in two analyses in 2010. This contrasts with former studies that indicated that PRC analyses were more sensitive than methods based on comparison of means (Pascual et al., 2010).
As stated in our previous work, the intention was to evaluate the efficacy of the treatments with a reduced number of applications to establish cost-effective control strategies (Marcotegui et al., 2015). These applications produced a moderate effect on almond tree pests and, though M. unicostata is one of the most affected taxa, a higher number of applications might be necessary to reduce damage effectively. However, we have found negative effects, especially with kaolin, on non-target arthropod fauna even with such low number of applications. Therefore, the increase in the frequency of applications would surely exert much higher disruption on the community of beneficial and other non-target arthropods.
In conclusion, despite the fact that the products assayed in the present work are claimed to be environmentally safer than those used in conventional agriculture, they still have detrimental effects on non-target arthropod fauna because of their reduced selectivity. This is especially relevant in the case of kaolin, because a continuous coverage of the plants throughout the season is essential for the effectiveness of this product, what might lead to long-term effects on the community of beneficial arthropods. Thus, a rational use of this type of products should be implemented, for example by alternating them with other strategies with less negative effects or by taking into account the best timing for its application according to the phenology of pests.