EFFECT of HOST PLANT and LAND USE on CABBAGE SEED WEEVIL INFESTATION and ASSOCIATED PARASITOIDS GABRIELLA KOVÁCS a Thesis for A

EFFECT OF HOST PLANT AND LAND USE ON CABBAGE SEED WEEVIL INFESTATION AND ASSOCIATED PARASITOIDS

PEREMEESTAIME JA MAAKASUTUSE MÕJU KÕDRA-PEITKÄRSAKA KAHJUSTUSE JA PARASITEERITUSE TASEMELE

GABRIELLA KOVÁCS

A Thesis for applying for the degree of Doctor of Philosophy in Agricultural Sciences

Väitekiri ﬁlosooﬁadoktori kraadi taotlemiseks põllumajanduse erialal

Tartu 2018

Eesti Maaülikooli doktoritööd Doctoral Theses of the Estonian University of Life Sciences

EFFECT OF HOST PLANT AND LAND USE ON CABBAGE SEED WEEVIL INFESTATION AND ASSOCIATED PARASITOIDS

PEREMEESTAIME JA MAAKASUTUSE MÕJU KÕDRA-PEITKÄRSAKA KAHJUSTUSE JA PARASITEERITUSE TASEMELE

GABRIELLA KOVÁCS

A Thesis for applying for the degree of Doctor of Philosophy in Agricultural Sciences

Väitekiri ﬁlosooﬁadoktori kraadi taotlemiseks põllumajanduse erialal

Tartu 2018 Institute of Agricultural and Environmental Sciences Estonian University of Life Sciences

According to verdict No. 6-14/5-2, of November 5, 2018, the Doctoral Committee of Agricultural Sciences of the Estonian University of Life Sciences has accepted the thesis for the defence of the degree of Doctor of Philosophy in Agricultural Sciences.

Opponent: Prof. József Kiss Szent István University, Hungary

Supervisors: Assoc. Prof. Eve Veromann Estonian University of Life Sciences

Prof. Anne Luik, professor emeritus Estonian University of Life Sciences

Defence of the thesis: Estonian University of Life Sciences, room 2A1, Fr. R. Kreutzwaldi 5, Tartu on December 12, 2018, at 10:00.

The thesis was edited using Elsevier’s Language Editing Express Service. The Estonian summary was edited by Kalle Hein.

This research was supported by European Social Fund’s Doctoral Studies and Internationalization Programme DoRa, carried out by Foundation Archimedes. The publication of this thesis was supported by the Estonian University ofLife Sciences.

��

LIST OF ORIGINAL PUBLICATIONS ...... 7 1. INTRODUCTION ...... 8 2. LITERATURE REVIEW ...... 10 2.1 Oilseed rape (Brassica napus L.) ...... 10 2.2 Cabbage seed weevil (Ceutorhynchus obstrictus Marsham) .. 11 2.3 Parasitoids of Ceutorhynchus obstrictus ...... 12 2.4 Strategies to enhance natural pest control by parasitoids ... 13 2.4.1 Secondary or biocontrol plants ...... 14 2.4.2 Land use ...... 15 3. HYPOTHESES AND AIM OF THE STUDY ...... 17 4. MATERIALS AND METHODS ...... 18 4.1 Study area and experimental design ...... 18 4.1.1 Small plot studies with spring-sown host plants ...... 18 4.1.2 Land use studies in and around winter oilseed rape fields ...... 19 4.3 Insect sampling ...... 20 4.4 Statistical analysis ...... 21 4.4.1 Analysis of the effect of host plants ...... 21 4.4.2 Analysis of the influence of land use ...... 22 5. RESULTS ...... 24 5.1 The effect of host plants ...... 24 5.1.1 Infestation rate of spring-sown host plant pods with Ceutorhynchus obstrictus ...... 24 5.1.2 Parasitism rate of Ceutorhynchus obstrictus and the species composition of the parasitoid community ...... 25 5.2 The influence of land use ...... 27 5.2.1 Infestation rate of winter-sown oilseed rape with Ceutorhynchus obstrictus ...... 27 5.2.2 Parasitism rate of Ceutorhynchus obstrictus and the species composition of the parasitoid community ...... 28 6. DISCUSSION ...... 30 6.1 The effect of host plants on Ceutorhynchus obstrictus infestation and its parasitoids ...... 30 6.2 The influence of land use on Ceutorhynchus obstrictus infestation and its parasitoids ...... 31 6.3 Suggestions for future studies ...... 34 7. CONCLUSIONS ...... 35 REFERENCES ...... 37 SUMMARY IN ESTONIAN ...... 47 ACKNOWLEDGEMENTS ...... 52 ORIGINAL PUBLICATIONS ...... 55 CURRICULUM VITAE ...... 97 ELULOOKIRJELDUS ...... 100 LIST OF PUBLICATIONS ...... 102

6 LIST OF ORIGINAL PUBLICATIONS

This thesis is based on the following papers, which are referred to by their Roman numerals in the text. All papers are reproduced with the kind permission of the publishers.

I Kovács, G.; Kaasik, R.; Metspalu, L.; Williams, I.H.; Luik, A.; Vero- mann E. 2013. Could Brassica rapa, Brassica juncea and Sinapis alba facilitate the control of the cabbage seed weevil in oilseed rape crops? Biological Control, 65: 124–129

II Kovács, G.; Kaasik, R.; Kaart, T.; Metspalu, L.; Luik, A.; Veromann, E. 2017. In search of secondary plants to enhance the efficiency of cabbage seed weevil management. BioControl, 62: 29–38

III Kovács, G.; Kaasik, R.; Treier, K.; Luik, A.; Veromann E. 2016. Do different field bordering elements affect cabbage seed weevil damage and its parasitism rate differently in winter oilseed rape? IOBC- WPRS Bulletin, 116: 75-80

IV Kovács, G.; Kaasik, R.; Lof, M.; van der Werf, W.; Kaart, T.; Hol- land, J.H.; Luik, A.; Veromann, E. 2018. Effects of land use on infestation and parasitism rates of cabbage seed weevil in oilseed rape. Pest Management Science DOI: https://doi.org/10.1002/ps.5161

The contribution of Gabriella Kovács to the papers was the following: Paper I II III IV Idea and study design * * * Data collection * * * Data analyses and interpretation * * * * Manuscript preparation * * * *

7 1. INTRODUCTION

The cabbage seed weevil (Ceutorhynchus obstrictus Marsham, other scientific name C. assimilis sensu auctt. non Paykull, Coleoptera: Curcu- lionidae) is one of the key pests of oilseed rape (OSR; Brassica napus L., Capparales: Brassicaceae) (Alford et al. 2003), the most dominant oilseed crop produced in the European Union (USDA 2018a). European OSR growers have been relying mainly on neonicotinoid (systemic insecticide – seed treatment) and pyrethroid (contact insecticide – foliar spray) applications against insect pests. The use of these treatments however has been proven to be problematic, causing resistance development in insect pests and being harmful towards non-target organisms, including beneficial arthropods (pollinators, parasitoids, predators). Hence, the need for more sustainable pest management strategies has been increasing for some time, especially in recent years, due to the changed legislations of the European Union. In 2013, the European Commission (EC) partially and temporarily banned the use of three neonicotinoids (clothianidin, thiamethoxam, and imidacloprid) in pollinator-attractive crops (Euro- pean Commission 2013) because of their harmful effect on several bee species (EFSA 2013a, b, c). Based on new evidence (EFSA 2018), the ban will be extended to all outdoor use (Butler 2018) in the future. As an outcome, farmers are left with the only widely available insecticide class, pyrethroids, which, if over-used, might worsen the resistance problem and increase detrimental non-target effects, highlighting the need for alternative approaches. In 2014, one of the most important measures of Directive 2009/128/EC on the “sustainable use of pesticides” entered into force, making integrated pest management (IPM) compulsory in the European Union (European Commission 2009). IPM prioritizes cropping practices that minimise the disruption to agro-ecosystems and rely on natural pest control mechanisms as much as possible. These mechanisms have an essential role in sustainable agriculture by providing the potential to reduce the need for chemical insecticide application.

Through conservation biological control (CBC; Barbosa 1998; Gurr et al. 1998; Jonsson et al. 2008), agricultural landscapes could be manipulated in a way that promotes biological pest control. For example, this could be achieved at the field scale by introducing biocontrol plants (Parolin et al. 2012a, 2014) (attractive or repellent plants) into the cropping system, or at the landscape scale by reconsidering/modifying land use, the composi-

8 tion and location of different habitat types within the landscape. To be able to effectively implement these practices, knowledge is required on i) the most efficient biocontrol plant species to use, and ii) key habitat types that support biological pest control.

The feeding and oviposition preferences of the pollen beetle (Brassico- gethes aeneus Fabricius syn. Meligethes aeneus Fabricius, Coleoptera: Nitid- ulidae) and its parasitoids on plant species other than OSR are quite well studied in Europe, the preferences of C. obstrictus are studied mainly in North America, and studies on the preferences of its parasitoids are scarce. Most landscape scale studies are from western European countries; Eastern Europe is very poorly represented in the literature (Holland et al. 2017). Furthermore, landscape scale studies on OSR pests exclude C. obstrictus. The current study extends earlier research by assessing the potential of different brassicaceous species (potential secondary plants) and land use in agricultural landscapes to affect tritrophic interactions involving the cabbage seed weevil.

9 2. LITERATURE REVIEW

2.1 Oilseed rape (Brassica napus L.)

Oilseed rape (OSR) is grown for its seed to produce oil, which is used to make biodiesel, edible and industrial oil; the by-product of oil production is a protein rich meal, which is used as animal feed (USDA 2018a). It is the first most produced oilseed crop in the European Union (EU) and the second after soybean among oilseeds produced worldwide (USDA 2018b). In 2017, the harvested production of oilseed and turnip rape was approximately 21.9 million tonnes on 6.7 million hectares, making the EU (driven by France, Germany, Poland, the UK, and Romania) the largest producer of OSR in the world (Eurostat 2018; USDA 2018a), followed by Canada and China (USDA 2018b). The majority of OSR grown in the EU is winter sown; spring OSR is grown in northern Europe: Finland, Estonia, Poland, Lithuania, and Latvia. In 2017, Estonia was the third largest spring OSR producer, 70.8 thousand tonnes on 39 thousand hectares (Eurostat 2018). The land area of OSR production has grown considerably in Estonia since the early 2000s, peaking at 98.2 thousand hectares in 2010 (Fig. 1). Despite the risky winter survival, the cultivation of winter OSR has been rising in the northern countries. In 2017, almost half of the OSR cultivated in Estonia was winter sown (Fig. 1).

��

Figure 1. Cultivated area of spring (SOSR) and winter (WOSR) oilseed rape/turnip rape in Estonia, 2000–2017 (Eurostat 2018).

10 OSR in Europe is mainly attacked by six insect pest species depending on the region, attacking at different growth stages and damaging different plant parts (Alford et al. 2003; Williams 2010). The stem-eating cabbage stem flea beetle (Psylliodes chrysocephalus Linnaeus, Coleoptera: Chrysomelidae), cabbage stem weevil (Ceutorhynchus pallidactylus Mar- sham), and rape stem weevil (C. napi Gyllenhal, both Coleoptera: Curcu- lionidae); the pollen-eating pollen beetle (Brassicogethes aeneus Fabricius syn. Meligethes aeneus, Coleoptera: Nitidulidae); the seed-eating cabbage seed weevil (C. obstrictus Marsham, Coleoptera: Curculionidae); and the inner pod wall-eating brassica pod midge (Dasineura brassicae Winnertz, Diptera: Cecidomyiidae).

Due to their different biologies and behaviours, the management of these pests requires different approaches. As a result of the change in the availability of insecticides and resistance development potential, alternative and preferably non-chemical methods are encouraged. IPM proposes the following steps to follow: less susceptible crop varieties, sustainable agronomy (ensuring crop establishment, minimising soil disturbance, minimal tillage, introducing new crops into the rotation, increasing the rotation time and space), threat assessment/monitoring (thresholds), encouraging natural enemies of pests via habitat management, introducing secondary plants into the cropping system (companion plants, trap cropping), and biopesticides.

2.2 Cabbage seed weevil (Ceutorhynchus obstrictus Marsham)

The larva of the cabbage seed weevil feeds on developing seeds inside the pod of Brassicaceae species, which is why it is an extensive and considerable univoltine pest in the OSR growing areas of Europe (Alford et al. 2003; Williams 2010) and North America (Kalischuk and Dosdall 2004; Ulmer and Dosdall 2006a). Adult weevils overwinter in leaf litter away from the fields, and emerge in the spring. When the temperatures reach 15 °C, weevils search for brassicaceous species to feed mainly on their buds, flowers, and developing pods (Free and Williams 1979; Ulmer and Dosdall 2006b; Williams 2010). Females lay their eggs into approximately 40–60 mm long pods (Dosdall and Moisey 2004) that are at least 2 mm in diameter (Doucette 1947; Dmoch 1965). First, they pierce the pod with their rostrum (Doucette 1947), lay usually one egg through the hole, and to avoid further oviposition to the same pod they mark it with

11 an oviposition deterring pheromone (Kozlowski et al. 1983). Although adults cause damage as well, when they feed on and oviposit into the pods, the larva causes the main damage to the crop by feeding on about five developing seeds inside the pod, which decreases the seed weight by 18–19% per infested pod (Free and Williams 1978a). Larval feeding can cause a significant yield loss if the infestation with C. obstrictus is more than 26–40% (Free and Williams 1978a; Lerin 1984; Buntin 1999). After the third instar larva has finished its development inside the pod, it exits the pod by chewing a hole through the pod wall and drops to the ground to pupate in the soil. The holes caused by adult feeding, oviposition, and larval exit might cause the pod to be more susceptible to fungal pathogen infestation (Cárcamo et al. 2001; Williams 2010). Later in the summer, new generation adults feed on various wild and/or cultivated brassicaceous plant species after emergence, and overwinter in field margins, woodlands, and perennial vegetation (Dmoch 1965; Alford et al. 2003).

The management of C. obstrictus in Europe mostly relies on pyrethroid insecticides applied as foliar sprays (Williams 2003; Heimbach and Müller 2013; Zimmer et al. 2014). The control threshold of insecticide applications for C. obstrictus varies between 0.5 and six beetles per plant depending on the country; in Estonia the threshold is 1–2 beetles on 10% of plants (Lõiveke 2003; Williams 2010). The only effective time to spray is at the early flowering period of the crop, at the peak of seed weevil abundance in the crop (Williams 2010). However, treating the flowering crop threatens the survival of beneficial insects, especially pollinators and parasitoids foraging on the flowers (Moreby et al. 2001; Ulber et al. 2010a; Gill et al. 2012; Ceuppens et al. 2015). Harming parasitoids reduces biological control and therefore potentially increases the need for further insecticide applications to control the pest (Alford et al. 1996; Murchie et al. 1997), leading to even more detrimental effects.

2.3 Parasitoids of Ceutorhynchus obstrictus

An adult parasitoid is a free-living insect that spends its immature developmental stages inside (endoparasitoid) or on (ectoparasitoid) another arthropod (host) feeding on it, and eventually killing the host. Hence, parasitoids have a potential to affect herbivorous insect populations, which is why these insects have an important role in agriculture. The most common parasitoids used in biological control belong to the orders

12 Hymenoptera and Diptera (Godfray 1994). Hymenopteran parasitoids are important in keeping the population of the cabbage seed weevil at manageable levels in OSR growing areas (Williams 2003; Dosdall and Mason 2010; Ulber et al. 2010b; Veromann et al. 2011). Although all developmental stages of C. obstrictus can be attacked by parasitoids (Mur- chie and Williams 1998; Williams 2003), the key species are larval univoltine idiobiont ectoparasitoids of the Pteromalidae family: Trichomalus perfectus (Walker), Mesopolobus morys (Walker), and Stenomalina gracilis (Walker) (all Hymenoptera: Pteromalidae) (Murchie and Williams 1998; Williams 2003; Ulber et al. 2010b; Veromann et al. 2011). Females of the most widespread and studied species T. perfectus (Murchie and Williams 1998; Williams 2003; Ulber et al. 2010b) arrive to OSR crops by the time C. obstrictus larvae have reached their second to third instar inside the pods (Dmoch 1975; Williams and Ferguson 2010). Upon finding a suitable host they lay a single egg on the host larva through the pod wall, which after hatching feeds on the host larva externally, ultimately killing it. After pupation the adult parasitoid exits the pod by chewing a hole through the pod wall (Murchie 1996; Williams 2003), mate, and leave the crop before harvest to overwinter in sheltered areas (Graham 1969; Dmoch 1975).

Larval parasitism rates reported over the years vary considerably; however, they are often above the threshold value of effective biological control (32–36%; Hawkins and Cornell 1994) (Büchi 1993; Murchie 1996; Williams 2003) and can reach even 90% (Veromann et al. 2011, 2013). Pest management approaches should focus on maintaining and possibly increasing the contribution of parasitoids to cabbage seed weevil management to increase the sustainability of cropping systems.

2.4 Strategies to enhance natural pest control by parasitoids

To utilize the pest control service provided by parasitoids, plant protection methods ought to be as safe as possible for beneficial arthropods and the cropping system must attract them. The agricultural landscape can be modified to promote the survival of natural enemies both at field and landscape scales, which would possibly increase their contribution to biological control. The implementation of such practices is called conservation biological control (Barbosa 1998; Gurr et al. 1998; Jonsson et al. 2008).

13 2.4.1 Secondary or biocontrol plants

Secondary or biocontrol plants are plant species that are grown in close proximity to the crop, to possibly serve as promoters of biological control (Hokkanen 1991; Gurr and Wratten 1999; Landis et al. 2000; Shel- ton and Badenes-Perez 2006; Cook et al. 2007a; Parolin et al. 2012a, b, 2014), indirectly increasing crop productivity at the field level.

Biocontrol plants can have direct or indirect effects on the crop, pest or pathogen, and on natural enemies (Parolin et al. 2014). A trap plant is a well-studied secondary plant type, which as a main function, attracts pests away from the crop of interest (Hokkanen 1991; Shelton and Badenes-Perez 2006) by being more attractive to the pest than the main crop at its most damage susceptible growth stage. These plants may also be a resource (e.g. provide food source, alternative hosts, habitat etc.) for natural enemies, which can suppress pest abundance (Virk et al. 2004; Khan and Pickett 2004), as part of a biocontrol-assisted trap cropping (Shelton and Badenes-Perez 2006). Another variation of trap cropping is the use of dead-end plants, which are plants that are highly attractive to pests but do not support their development (Idris and Grafius 1996; Shelton and Nault 2004).

The effect of different host plant species (as possible biocontrol plants) on the pollen beetle and its parasitoids is quite well studied in Europe. Brassica rapa L. has shown great potential to be included in a trap crop strategy in Finland (Hokkanen et al. 1986), Switzerland (Büchi 1990), and the UK (Cook et al. 2006, 2007b), by being more attractive to the pollen beetle than B. napus during its most damage susceptible growth stage. In Estonia, other brassicaceous plant species were found to be potentially effective biocontrol plants in pollen beetle management. For instance, Brassica juncea (L.) Czern. and B. nigra (L.) supported the hymenopteran parasitoids of pollen beetles (Kaasik et al. 2014a, b). Sinapis alba L., B. nigra (L.) W. D. J. Koch, and Raphanus sativus (L.) var. oleiformis Pers. could possibly be used in trap cropping systems to potentially lure pollen beetles away from B. napus fields, because adult beetles were more attracted to them than to B. napus (Veromann et al. 2012; Kaasik et al. 2014b). Raphanus sativus (L.) var. oleiformis Pers. might act as a dead-end trap crop, as pollen beetles were attracted to it for oviposition, but their larvae could not complete their development in the buds (Veromann et al. 2014).

14 The host plant preferences of the cabbage seed weevil for feeding and/or oviposition have been studied mainly in the UK, Switzerland, the United States, and Canada. According to Free and Williams (1978b) in the UK, C. obstrictus feeds on brassicaceous plants but prefers to oviposit into OSR pods. Brassica rapa as a trap crop did not reduce C. obstrictus damage during the study of Büchi (1990) in Switzerland. During the study of Cook et al. (2006) in the UK, adult weevils did prefer the odour of B. rapa to that of a low-isothiocyanate cultivar of OSR, but due to the low abundance of weevil adults, the potential of B. rapa as a trap crop could not be evaluated. North American studies found B. rapa to be more attractive to C. obstrictus than B. napus; and they found B. nigra and S. alba to be less or not at all susceptible to C. obstrictus damage (Doucette 1947; Harmon and McCaffrey 1997; Brown et al. 1999; McCaffrey et al. 1999; Cárcamo et al. 2007). Sinapis alba is even used in developing C. obstrictus-resistant OSR (Dosdall and Kott 2006; Tansey et al. 2010), because of its resistance to C. obstrictus (Brown et al. 1999; Kalischuk and Dosdall 2004). To the best of our knowledge, the host plant preferences of C. obstrictus parasitoids have been rarely studied.

2.4.2 Land use

Plant–herbivore–natural enemy interactions can depend on landscape composition. Annual, perennial crops and non-crop areas all have their function within an agricultural landscape. Semi-natural habitats, such as field margins, permanent grasslands, and woodland (including cop- pices), can shield natural enemies and pests from agricultural practices, by providing overwintering sites, alternative hosts, and food resources (Gurr et al. 1998; Landis et al. 2000; Bianchi et al. 2006; Tscharntke et al. 2007; Rusch et al. 2012; Holland et al. 2016). Therefore, the proportion, location, and connectivity between different habitat types within the landscape have an important effect on ecosystem service provision (biological control by natural enemies, and pest incidence). Redesigning land use by optimizing the habitat composition to be able to keep pest abundance and damage under the economic threshold can be a promising plant protection method.

To be able to implement this natural pest suppression approach, knowledge on the key habitats and land use supporting biological pest control is required. The pollen beetle is well studied in a landscape context; however, the findings can be contrasting due to the variation in the

15 methodological approaches used (Thies et al. 2003; Zaller et al. 2008b; Frank et al. 2010; Rusch et al. 2010), and the significance of biological control provided by parasitoids in the study region. According to a study carried out near Göttingen (Germany) with potted B. napus plants distributed in the landscape, the proportion of non-crop habitats were negatively related to herbivory by pollen beetles and positively related to parasitism (Thies et al. 2003). Similarly to Thies et al. (2003), field studies carried out in north-western France by Rusch et al. (2011) also confirmed that landscapes with more non-crop areas enhance the ecosystem service provided by the natural enemies of the pollen beetle. In contrast to these observations, field studies in Lower Austria showed that pollen beetle (and stem weevil) abundance (Zaller et al. 2008b), as well as pollen beetle (and brassica pod midge) damage (Zaller et al. 2008a), were negatively correlated with the proportion of OSR and positively related to non-crop habitats in the landscape. Although a possible reason for this contradictory finding is that in Lower Austria, parasitoids did not play a significant role in pollen beetle management during the study (Zaller et al. 2008b).

The infestation of OSR pods with C. obstrictus and its parasitism has not been studied before in a landscape context. The only landscape study mentioning C. obstrictus measured the incidence of different OSR pests, and that of the parasitoids possibly associated (based on available literature) with OSR pests, based on flight activity in southern Sweden (Berger et al. 2018). They did not find any C. obstrictus adults in the water trays; however, the abundance of one of its key parasitoids, S. gracilis, was higher in OSR fields further away from woody habitats, and was positively related to the proportion of OSR in the landscape.

16 3. HYPOTHESES AND AIM OF THE STUDY

The aim of the study was to provide a basis for environmentally safe and sustainable plant protection approaches to control C. obstrictus in OSR. The first approach was connected to the use of secondary plants in or near the main crop to enhance the occurrence of biocontrol agents through increased biodiversity in the cropping system; and the second approach involved the potential of landscape manipulation to manage this pest. The objectives of the study were as follows:

1. To determine how successful hymenopteran parasitoids are in keeping the abundance of C. obstrictus under control (Paper I, II, III, and IV).

Hypothesis: Ceutorhynchus obstrictus abundance is successfully controlled by its natural enemies, hymenopteran parasitoids.

2. To study if different host plant species affect the host finding success of C. obstrictus and its parasitoids (I, II).

Hypothesis: The infestation rate, parasitism rate, and parasitoid species composition depend on host plant species.

3. To study how land use affects the host finding success of C. obstrictus and its parasitoids at the field scale (III, IV), and at the landscape scale (IV).

Hypothesis: The host finding success of C. obstrictus and its parasitoids depends on the type of habitat at the field margin and the proportion of different habitats in the surrounding landscape.

17 4. MATERIALS AND METHODS

4.1 Study area and experimental design

4.1.1 Small plot studies with spring-sown host plants

To find secondary plants that can possibly protect OSR and/or to enhance the efficacy of biological control, small plot studies were conducted to determine the effect of different host plants on C. obstrictus infestation and its parasitoids between 2007–2012 (I, II) in the experimental fields of the Estonian University of Life Sciences in Tartu County, Estonia (I: latitude: 58.36°, longitude: 26.69°; II: latitude: 58.37°, longitude: 26.66°). The plot size was 1 × 5 m and there were three replicates of each plant species per year which were grown in a randomized complete block design. Crop management was uniform in all plots, and neither fertilizers nor pesticides were used.

Plant species grown over the years were as follows: OSR Brassica napus (L.) (spring variety, ‘Mascot’), turnip rape B. rapa (L.) (’Largo’), mustard greens B. juncea ([L.] Czern.) (‘Jadrjonaja’), black mustard B. nigra ([L.] W.D.J. Koch), rucola/arugula Eruca vesicaria subsp. sativa ([Mill.] Thell.) (= E. sativa, ‘Poker’), fodder radish Raphanus sativus (L.) var. oleiformis (Pers.) (= R. sativus, ‘Bille’, pale violet flowers), and white mustard Sinapis alba (L.) (‘Branco’) (Table 1).

Table 1. Plant species grown over the study period (2007–2012; I+II). 2007 2008 2010 2011 2012 B. napus * * * * * B. rapa * * * * B. juncea * * * * * B. nigra * * * E. sativa * * * R. sativus * * * S. alba * * * * *

18 4.1.2 Land use studies in and around winter oilseed rape ﬁelds

To estimate how field bordering habitats (III, IV), and the distance or proportion of habitats in the landscape (IV) influence C. obstrictus infestation and its parasitoids, field and landscape scale studies were carried out in and around winter OSR crops in 2013 (III) and 2014 (IV) in Tartu County, Estonia (latitude: 58.21° to 58.44°, longitude: 26.17° to 26.66°).

The study area selection, definition, and description of semi-natural habitats (III, IV), and mapping (IV) were based on standardized protocols across participating regions of the EU FP7 project QuESSA (QuESSA Europe 2013; Holland et al. 2014). The QuESSA project defined a semi- natural habitat as any habitat within or outside of the crop containing a community of non-crop plant species.

In 2013 and 2014, five types of semi-natural habitats were defined according to their dominant vegetation (herbaceous or woody) and shape (areal or linear): herbaceous areal (natural hayfields or abandoned fields that have not developed more than 30% shrub/tree canopy cover), woody areal (woodland), cover crop (in most cases clover), herbaceous linear (narrow grassy crop margin), and woody linear (hedge, line of trees). Herbaceous semi-natural habitats had less than 30% woody vegetation cover, otherwise they were considered woody. Linear habitats were less than 25 m wide, and areal habitats were a minimum of 25 m wide.

Sampling points in winter OSR crops were selected based on the type of habitat bordering the sampled side of the crop, within 18 non-overlapping landscape circles with a 1 km radius. In 2013 (III), the four bordering habitat types were all semi-natural habitats (herbaceous areal, herbaceous linear, woody areal, and woody linear). There were up to four sampling areas within a landscape circle, maximum one per bordering habitat type for 54 sampling areas in total. The number of replicates per bordering habitat type was unequal: 19 (herbaceous areal), 15 (herbaceous linear), 11 (woody areal), and 9 (woody linear).

In 2014 (IV), an improved and simplified study design was used. There were three bordering habitat types; two were semi-natural habitats (herbaceous linear, and woody linear) and one was a crop field. There was only one sampling area per landscape circle for 18 sampling areas in

19 total. The number of replicates per bordering habitat type was equal (six per bordering habitat type). The study area selection in 2014 was furthermore based on the proportion of semi-natural habitats around the sampling area within the 1 km radius.

Sampling within the fields was performed along a transect extending from the edge of the field bordering the habitat type, 20 m (III) and 75 m (IV) into its interior. Sampling areas were managed conventionally, however none of them were treated with insecticides against C. obstrictus.

In 2014 (IV), we extended the study carried out in 2013 (III) by studying the effect of habitats in the surrounding landscape in addition to the effect of the bordering habitat type. Land use was recorded around each sampling area (1 km radius) (IV), based on interactive maps (PRIA 2017; Estonian Land Board 2017) and detailed field inspections. Every habitat that was a minimum of 1.5 m wide and a minimum of 50 m long was mapped. The mapping was done using ArcGIS (Version 10.1; Esri, Redlands, CA, USA). Detailed information on the mapped habitat categories is given in Table S1, IV.

4.3 Insect sampling

The infestation rate of pods with C. obstrictus and the parasitism rate of C. obstrictus eggs and larvae were assessed at the ripening stage of the studied plant species (BBCH 80–85, Lancashire et al. 1991) (I, II, III, IV). Ten pods were collected per plant (five pods from the main and five from the third raceme from the top). During the plot experiments (2007–2012; I, II), 10 plants were sampled per each plot (300 pods per plant species). In 2013 (III) and 2014 (IV), five plants were sampled per sampling distance. In 2013 (III), plants were sampled at two distances (2 m and 20 m from the field edge, 50 pods per distance, and 100 pods per sampling area for 5400 pods in total). In 2014 (IV) plants were sampled at four distances (2, 25, 50, and 75 m from the field edge, 50 pods per distance, and 200 pods per field for 3600 pods in total).

The pods were kept at 20–23 °C in emergence boxes for 30 days as described by Veromann et al. (2006). Emerged weevil larvae and parasitoids were counted; all pods were examined and dissected to find exit holes and weevil or parasitoid remains. Adult parasitoids were preserved in

20 70% ethanol at –20 °C and identified using the keys of Graham (1969), Trjapitzin (1978), Gibson et al. (2006), Baur et al. (2007), and Muller et al. (2007) under an Olympus SZX-ZB9 stereo-microscope (Olym- pus Optical Co., Ltd., Tokyo, Japan) with a maximum magnification of 2289×, coupled to an Olympus Highlight 3100 light source (Olympus Optical Co., Europa, GmbH, Hamburg, Germany).

4.4 Statistical analysis

The proportion of infested pods with C. obstrictus of the collected pods and the proportion of infested pods that were parasitized were considered as dependent variables in the analyses (I, II, III, IV).

4.4.1 Analysis of the eﬀect of host plants

The effect of the plant species, year, and plant species × year interaction on the infestation rate was calculated using a generalized linear model (GLM) with a Poisson distribution and a log link function (I), and a generalized linear mixed effects model (GLMM) with a binomial distribution and logit link function, considering replicates as a random effect (II, I+II); followed by the pairwise comparison of the least square means corresponding to the studied factors (I, II, I+II), using the Tukey-Kramer method (I+II).

The effect of the plant species, year, and plant species × year interaction on the parasitism rate was calculated using a GLM (I), and a GLMM (II, I+II) with a binomial distribution and logit link function (I, II, I+II), considering replicates as a random effect (II, I+II); followed by the pairwise comparison of the least square means corresponding to the studied factors (I, II, I+II) using the Tukey-Kramer method (I+II).

The analysis of the species composition of the parasitoid community focused on two species: M. morys and T. perfectus. The remaining species were grouped as Mymaridae and ‘Other families’, due to their low abundance during the study. The relationship between plant species and parasitoid species composition was analysed using the two-way frequency table and Fisher’s exact test (II, I+II). Possible common patterns in parasitoid species composition and their differences between years and plant species were studied with principal component analysis (PCA; II).

21 The statistical analyses were conducted using the GENMOD procedure in SAS 8.02 (I) and the GLIMMIX procedure in SAS 9.4 (II, I+II) (SAS Institute, Inc., Cary, NC, USA). All results were considered statistically significant at P ≤ 0.05.

4.4.2 Analysis of the inﬂuence of land use

Analysis of the effects of bordering habitats and distance from the edge of the crop The effects that bordering element type (III, IV) and distance from the crop edge (IV) had on the infestation and parasitism rate were analysed using a GLM with a normal distribution and identity link function, followed by the analysis of the differences between treatments with the same analysis using a Bonferroni correction (III); and GLMM with a binomial distribution and logit link function, considering landscape circles as random effects (IV).

The relationship between the infestation rate and parasitism rate, and between the abundance of different parasitoid groups, was studied with the pairwise Pearson correlation analysis (IV).

The parasitoid species composition analysis focused on three species: T. perfectus, S. gracilis, and M. morys, grouping the remaining species as ‘Other parasitoid sp.’ due to their low abundance. The effect of bordering habitat type on parasitoid species composition was analysed using a GLMM with a binomial distribution and logit link function, including the bordering habitat type as a fixed effect and landscape circle as a random effect (IV).

Landscape level analyses Habitats that were present in at least 10 landscape circles were selected for the analysis. A spatially explicit method was used for weighing the effect of the selected habitats according to their distance to the sampling area. The effect of the habitats on C. obstrictus infestation and parasitism was analysed with a GLMM with a binomial distribution and a logit link function, using distance weighted landscape variables as regressors, where the distance weighting was obtained from a rotationally symmetric two- dimensional t-distribution kernel with parameters that were fitted to the data by maximizing the likelihood. With the use of the dredge function

22 (MuMIn in R; Bartoń 2016), a maximum of three habitats were selected that most strongly affected the infestation and parasitism rates (IV).

Statistical analyses were conducted in STATISTICA 12 (StatSoft Inc., Tulsa, OK, USA) (III), in SAS 9.4 (SAS Institute, Inc., Cary, NC, USA) using the GLIMMIX and CORR procedures (IV), and in R (R Founda- tion for Statistical Computing, Vienna, Austria) using the dredge function and a tailor-made script (IV). All results were considered statistically significant at P ≤ 0.05.

23 5. RESULTS

5.1 The eﬀect of host plants

5.1.1 Infestation rate of spring-sown host plant pods with Ceutorhynchus obstrictus

Ceutorhynchus obstrictus infestation was detected on all studied plant species over the study period (I, II), at different rates (Fig. 2); however, infestation stayed below the estimated threshold value of measurable damage (26–40%; Free and Williams 1978a, Lerin 1984, Buntin 1999) on all plant species. The plant species had a significant effect on the infestation

rate over the years (F6,79 = 15, P < 0.001). The highest average proportion of infested pods was recorded on B. napus (24%) in 2012. On average, 10.5% of B. napus pods were infested between 2007–2012, significantly more compared to those of B. nigra (P < 0.001), E. sativa (P < 0.001), and S. alba (P < 0.001). The overall difference between study years was

also significant (F4,79 = 5.23, P = 0.001). More pods were infested with C. obstrictus in 2012 compared to in 2010 (P = 0.003), and 2011 (P = 0.037).

� ��

��

Figure 2. Estimated proportion of Ceutorhynchus obstrictus-infested pods by plant species and year. Horizontal lines and corresponding values denote the overall proportions (±SE) over the study period (2007–2012; I+II). The ‘‘×’’ symbols denote that in 2010, B. rapa, in 2007 and 2008, B. nigra and R. sativus were not studied. Different letters indicate statistically significant differences between plant species (P < 0.05; Tukey–Kramer).

24 5.1.2 Parasitism rate of Ceutorhynchus obstrictus and the species composition of the parasitoid community

Plant species affected the parasitism significantly only when summing the data from 2007–2008 (I). During these two years, unlike the infestation, higher parasitism rates were recorded from the pods of B. juncea and B. rapa than from those of B. napus. The plant species had no influence on the parasitism rate of C. obstrictus between 2010–2012 (II), and over

the whole study period (F5,54 = 2.28, P = 0.06; Fig. 3). The average parasitism rate over the years was above 34% on all plant species, reaching more than 80% on R. sativus. There were no parasitoids found on E. sativa.

The overall difference between study years was significant (F4,54 = 6.88, P < 0.001). There were significantly higher parasitism rates recorded in 2010 than in 2011 (P < 0.001) and 2012 (P = 0.011). The highest average parasitism rate, a remarkable 95.5%, was recorded on R. sativus in 2010 (Fig. 3).

In total, 227 C. obstrictus parasitoid specimens were reared from pods between 2007–2012, mostly from B. napus (74 specimens, Table 2). In total, there were at least 17 different parasitoid species attacking

��

� ��

Figure 3. Estimated parasitism rate of Ceutorhynchus obstrictus by plant species and year. Horizontal lines and corresponding values denote the overall proportions (±SE) over the study period (2007–2012; I+II). The ‘‘×’’ symbols denote that in 2010, B. rapa, in 2007 and 2008, B. nigra and R. sativus were not studied. There were no statistically significant differences between plant species (P = 0.06).

25 C. obstrictus on different host plants. Mesopolobus morys was the dominant species over the whole study (46.3% of all parasitoids), followed by T. perfectus (22.9%). The species composition on R. sativus was different from that on the other plant species (P < 0.001; Fisher’s exact test; Fig. 4); on this plant species, C. obstrictus was mostly parasitized by seven different egg parasitoid species, all belonging to the Mymaridae family (Table 2).

Table 2. Total number of parasitoids reared from Ceutorhynchus obstrictus collected from different brassicaceous plant species over the study period (2007–2012; I+II).

Brassica napus Brassica juncea Brassica nigra Brassica rapa Raphanus sativus Sinapis alba Total Pteromalidae – total 62 45 6 46 1 2 162 Mesopolobus gemellus Baur & Muller 1 1 Mesopolobus incultus Walker 1 1 Mesopolobus morys Walker 34 34 3 34 105 Stenomalina gracilis Walker 1 1 Trichomalus bracteatus Walker 1 1 2 Trichomalus perfectus Walker 26 10 3 11 1 1 52 Mymaridae – total 9 5 4 40 1 59 Anaphes arenbergi Debauche 1 1 5 8 Anaphes brachygaster Debauche 1 1 Anaphes crassicornis group 7 7 Anaphes cultripennis Debauche 11 11 Anaphes fuscipennis Haliday 2 3 2 3 10 Anaphes regulus Walker 7 7 Anaphes tarsalis Mathot 1 2 3 1 7 unidentifiable Anaphes species 2 3 5 unidentifiable Anaphini 2 2 unidentifiable Mymaridae 1 1 Eulophidae – total 2 1 1 4 Euderus albitarsis Zetterstedt 1 1 Tetrastichus rebezae Kosjukov 1 1 Tetrastichus semidesertus Kosjukov 2 2 Eurytomidae – total 1 1 2 Eurytoma curculionum Mayr 1 1 unidentifiable Eurytoma species 1 1 Total 74 51 8 50 41 3 227

26 � � � � � � � ��

��

Figure 4. Species composition and total numbers of parasitoids reared from Ceutorhynchus obstrictus collected from different brassicaceous plant species over the study period (2007–2012; I+II). Different letters indicate statistically significant differences between plant species (P < 0.05; Fisher’s exact test).

5.2 The inﬂuence of land use

5.2.1 Infestation rate of winter-sown oilseed rape pods with Ceutorhynchus obstrictus

The proportion of infested pods per plant by C. obstrictus was between 6–12% on average per bordering habitat type between 2013–2014 (9.8% on average in 2013; 8% on average in 2014) (III, IV), which is below the threshold of considerable damage (26–40%; Free and Wil- liams 1978a, Lerin 1984, Buntin 1999). In 2014, infestation rates were higher in sampling areas bordered by herbaceous linear habitats, compared to infestation rates in areas bordered by woody linear or control habitats (IV). Although during the 2013 study, infestation rates were also somewhat higher next to herbaceous linear elements than next to woody linear elements (Fig. 1 in III), there was no significant difference between infestation rates in sampling areas bordered by different landscape elements (III).

27 At the landscape scale, the proportion of herbaceous areal semi-natural habitats, wheat and permanent grasslands within the 1 km landscape circles increased the infestation rate; however, the total proportion of semi-natural habitats (20.9–75.9%, Table 2 in IV) did not affect the infestation rate (IV).

5.2.2 Parasitism rate of Ceutorhynchus obstrictus and the species composition of the parasitoid community

Parasitism rates were between 41–62% on average per bordering habitat type between 2013–2014 (60% on average in 2013; 55% on average in 2014). The highest average parasitism rate (87%) was recorded in an OSR crop bordered by a woody linear habitat (Table 2 in IV). The bordering habitat type did not have any effect on parasitism rates (III, IV), and neither did the sampling distance from the field edge (IV). The total proportion of semi-natural habitats within the landscape circles had no effect on the parasitism rate; however, the closer proximity (within 750 m; Fig. S1 in IV) of herbaceous linear semi-natural habitats in the landscape increased parasitism rates, and the closer proximity of permanent grasslands in the landscapes decreased parasitism rates (Fig. 1 in IV).

In total, 165 C. obstrictus parasitoid specimens were reared from the pods in 2014 (IV). The bordering habitat type did not affect the species composition (Fig. 4 in IV). The most abundant species was T. perfectus (122 specimens) followed by S. gracilis (20 specimens), and M. morys (18 specimens). Furthermore, the following parasitoids were represented by one specimen each: T. lucidus, unidentifiable Trichomalus species, Anaphes fuscipennis, Eurytoma curculionum, and one unidentifiable parasitoid species.

Spring and winter OSR had different parasitoid species compositions (P < 0.001, Fisher’s exact test). Of the key parasitoid species, T. perfectus was dominant on winter OSR, while M. morys dominated on spring OSR (Fig. 5). Stenomalina gracilis did not parasitize C. obstrictus on spring OSR (Table 3).

28 Table 3. Occurrence of parasitoid species reared from Ceutorhynchus obstrictus collected from spring oilseed rape (SOSR, 2007–2012; I+II) and winter oilseed rape (WOSR, 2014; IV).

SOSR WOSR Pteromalidae Mesopolobus gemellus Baur & Muller * Mesopolobus incultus Walker * Mesopolobus morys Walker * * Stenomalina gracilis Walker * Trichomalus lucidus Walker * Trichomalus perfectus Walker * * Mymaridae Anaphes arenbergi Debauche * Anaphes fuscipennis Haliday * * Eulophidae Tetrastichus semidesertus Kosjukov * Eurytomidae Eurytoma curculionum Mayr * *

� ��

��

� ��

Figure 5. Species composition of parasitoids reared from Ceutorhynchus obstrictus collected from spring oilseed rape (SOSR, 2007–2012; I+II) and winter oilseed rape (WOSR, 2014; IV).

29 6. DISCUSSION

6.1 The eﬀect of host plants on Ceutorhynchus obstrictus infestation and its parasitoids

The infestation and parasitism rate as well as parasitoid communities of C. obstrictus were studied on spring-sown brassicaceous plant species to find potential secondary plants that increase the biological control of this pest (I, II).

The proportion of C. obstrictus-infested pods stayed below the threshold of economic damage of ~26% (Free and Williams 1978a, Lerin 1984, Buntin 1999) throughout the study years (I+II). The infestation rate differed significantly on different host plant species; unfortunately, none of the studied plant species were more infested than B. napus. Interestingly, B. rapa was also somewhat less infested on average than B. napus, which is in contrast with the findings of other North American studies (Brown et al. 1999; Cárcamo et al. 2007), which showed that B. rapa can be used as a trap crop by being more attractive to C. obstrictus than B. napus. Therefore, similar to the other tested plant species in this study, B. rapa is unlikely to be an effective trap crop in C. obstrictus management in Estonia.

The low infestation rates on S. alba (I, II) and B. nigra (II) in this study is in accordance with previous studies finding S. alba and B. nigra to be unattractive to C. obstrictus (Doucette 1947; Kalischuk and Dosdall 2004; Ulmer and Dosdall 2006a). This suggests that the C. obstrictus- resistant germplasm—based on S. alba crosses with B. napus—developed in Canada (Dosdall and Kott 2006; Tansey et al. 2010) should also work in Estonia. However, to date, C. obstrictus damage has been below the economic threshold in Estonia, hence the need for a weevil-resistant germplasm is not yet justified.

Despite the low weevil abundance, the parasitism rate of C. obstrictus larvae and eggs was above 34% on average on all plant species, exceeding the suggested level of an effective natural pest control (~32%; Hawkins and Cornell 1994) (I+II). Therefore, C. obstrictus parasitoids proved to be valuable biological control agents contributing to the reduction of the next generation of weevils.

30 Our findings extend the knowledge on parasitoid species attacking C. obstrictus (Williams 2003; Ulber et al. 2010b; Cárcamo and Brandt 2017) and show the parasitoid species composition on different host plant species. Mesopolobus morys, one of the most common larval ectoparasitoids of C. obstrictus (Murchie and Williams 1998), was the most abundant parasitoid species over the study period; it was dominant on B. rapa, B. juncea, and B. napus (I+II). Its high abundance on B. napus is in line with the results of Murchie (1996), who found M. morys to be more numerous on spring OSR than on winter OSR. In addition to larval parasitoids, an unexpectedly high proportion of parasitoids were egg parasitoids, belonging to the Mymaridae family. Although larval parasitism is important in reducing damage to the crop, egg parasitism eliminates seed consumption. A parasitized larva consumes less seeds than a healthy larva; however, when an egg is parasitized, it will not develop further, and hence no seeds will be consumed. Since mymarids dominated on R. sativus (II, I+II), the introduction of this plant species into the cropping system might lure more egg parasitoids into the landscape, possibly increasing their importance as biological control agents of seed weevils.

6.2 The inﬂuence of land use on Ceutorhynchus obstrictus infestation and its parasitoids

The effect of crop and non-crop habitats on C. obstrictus-infestation and its parasitoids in winter OSR were studied for the first time to gain knowledge on key habitats supporting biological control in the surrounding landscape (IV), including the field edge (III, IV).

Rather than the total proportion of habitats in the landscape, the type of habitat influenced the infestation levels and parasitism rate of this pest. Additionally, the spatial range of land use effects was greater for the pest than for its parasitoids. Habitats closer to the sampling area contributed more to the parasitism rate, while the infestation rate was affected by habitats across the entire landscape circle (IV). This is in line with studies stating that small, specialized parasitoid species experience their surrounding landscape at a smaller range than their herbivore hosts (Kruess and Tscharntke 1994; Kruess 2003; Tscharntke et al. 2005).

Areal herbaceous semi-natural habitats, permanent grasslands, and wheat fields within the landscape increased C. obstrictus infestation (IV). We can

31 hypothesize that herbaceous semi-natural habitats and permanent grasslands offer possible overwintering sites for C. obstrictus by being perennial habitats. In this study, wheat corresponds to the previous year’s OSR by following OSR in the crop rotation across the study area. A possible explanation for its positive effect on the infestation rate is that C. obstrictus found suitable overwintering sites close to the OSR crops it emerged from. This would be in line with findings about the pollen beetle, whose abundance after overwintering has been found to be positively related to the proximity of the previous year’s OSR (Rusch et al. 2012).

One habitat type, permanent grassland, had a different effect on C. obstrictus than on its parasitoids (IV). At the landscape scale, it increased infestation rates and decreased parasitism rates. Therefore, it acted as an ecosystem disservice provider in C. obstrictus management, promoting pest infestation and decreasing natural pest control.

The effect of herbaceous linear habitats varied at the field and landscape scale (IV). At the field scale, these habitats increased the infestation rate when they were bordering the studied field. However, at the landscape scale, herbaceous linear habitats increased the parasitism rate. Linear grass margins have been previously found to be important in pest control (Hol- land et al. 2012), mainly as overwintering habitats of predatory arthropods (Geiger et al. 2009; Treier et al. 2017). Although in the current study, the infestation rates were higher next to herbaceous linear habitats, they were still well under the economic threshold of ~26%, and at the same time parasitism rates were very high in these fields (62.5–81.2%). Therefore, it is likely, that the parasitoids of C. obstrictus used these habitats as overwintering sites.

The high total proportion of semi-natural habitats across the studied landscapes was most probably an important factor keeping parasitism rates high (IV). According to Thies and Tscharntke (2010), pollen beetle parasitism decreased below the value of effective biological control when the area of semi-natural habitats decreased to less than 20%. In this current study, the area of semi-natural habitats was more than 20% (up to 75.9%) in all studied landscape circles, indicating that the preservation of semi-natural habitats must be a priority in order to keep pest densities under control.

The low damage rate of OSR pods (III: 9.8%; IV: 8%) and high parasitism rate (up to 87%, IV) of C. obstrictus (III, IV) suggest that OSR growers

32 in Estonia should strictly avoid using insecticides against C. obstrictus, because naturally occurring parasitoid species play an important role in controlling seed weevil abundance. The plant protection scheme of OSR suggests applying insecticides against C. obstrictus at the end of the full flowering stage (BBCH 67–69). However, this exposes parasitoids to the detrimental effects of insecticides, because this is exactly when the parasitoids of C. obstrictus arrive to the crop to search for hosts and feed on nectar. The experiences of other countries have shown that insecticide use against C. obstrictus has the potential to decrease larval parasitism rates dramatically. For instance, in the UK the parasitism rate of the cabbage seed weevil was over 70% in the 1970s, but decreased to as low as 8% after using broad-spectrum insecticides for 10–20 years (Alford et al. 1996; Walters et al. 2003).

According to Büchs and Nuss (2000), Sunderland (2002), and Williams et al. (2010), generalist predators have an important role in pest control. Ceutorhynchus obstrictus, similar to most OSR pests (cabbage stem flea beetle, pollen beetle, cabbage stem weevil, and brassica pod midge), is exposed to predation when the mature larva drops to the ground to pupate in the soil. Therefore, it is likely that in addition to parasitoids, ground dwelling arthropods also contributed to the low infestation rates recorded in these studies.

The bordering habitat type did not affect the species composition of parasitoids (IV). Trichomalus perfectus, the most abundant parasitoid species in Europe (Murchie and Williams 1998), was the primary species in this study (IV). This finding is consistent with earlier studies (Murchie 1996; Ulber and Vidal 1998; Veromann et al. 2011), which showed that T. perfectus is the prevalent parasitoid of C. obstrictus in winter OSR. Mesopolobus morys was only moderately abundant on winter OSR (IV) in contrast to its dominance on spring OSR (I+II), which further proves that it favours to parasitize C. obstrictus found in pods of spring OSR. According to Haye et al. (2018), although there is an overlap in the abundance of T. perfectus and M. morys, M. morys might be more adapted to higher temperatures, which would explain its higher abundance later in the season (during the ripening stage of spring OSR). Stenomalina gracilis was as abundant as M. morys on winter OSR (IV) and completely absent on spring OSR (I+II), which suggests that its development is slightly more in sync with that of C. obstrictus on winter OSR.

33 6.3 Suggestions for future studies

To be able to draw further conclusions about insect preference, dual- and/ or multiple-choice assays should be carried out, followed by the testing of potential biocontrol plants under field conditions, to provide more reliable and inclusive results. Additionally, since winter OSR has gained pop- ularity in recent years, both spring and winter OSR should be included in the experiments. Land use studies, on the other hand, should include spring OSR, because it is still cultivated in a significant area, and its pest management requires a higher pesticide input. Landscape scale studies should also take other factors into consideration besides the proportion and location of habitats, such as farming practices (e.g. chemical inputs, tillage), weather (e.g. wind direction), and the vegetation composition of habitat types. Furthermore, most factors that contribute to pest management vary from year to year; hence, all experiments should be carried out as long-term (at least 3–4 years) studies.

34 7. CONCLUSIONS

This study was designed to measure if and how cabbage seed weevil infestation and its parasitoids are affected by different host plant species (I, II) and land use (III, IV), to be able to find alternative methods that control this pest, without using pesticides.

Low C. obstrictus infestation rates, paired with consistently high parasitism rates in the study, imply that the species rich parasitoid community played an important role in keeping C. obstrictus infestations under control (I, II, III, IV). This confirms the first hypothesis of this study. The finding that naturally occurring parasitoids can effectively control the abundance of the cabbage seed weevil is a very valuable and applicable information for OSR growers and policy makers, proving that there is no need to use synthetic insecticides against this pest.

Most results confirmed the second hypothesis. The host plant species did influence C. obstrictus infestation and parasitoid species composition; however, it did not affect the already high parasitism rate (I, II).

An implication of the novel finding that the parasitoid species composition of C. obstrictus depended on host plant species, is the possibility to increase the biodiversity in agricultural fields. Raphanus sativus provided a suitable habitat for less common C. obstrictus parasitoid species (Mymaridae) (II). Therefore, the inclusion of R. sativus in the cropping system would most likely contribute to a more diverse parasitoid community, which would potentially enhance the biocontrol service provided by parasitoids.

The third hypothesis was partially confirmed. Land use affected the infestation and parasitism rates; however, it did not have a significant effect on the species composition of parasitoids (IV).

To our knowledge, this is the first time that information about the key habitat types affecting C. obstrictus management is presented. The ecosystem service provided by the parasitoids of C. obstrictus was supported by herbaceous linear semi-natural habitats in the surrounding landscape. We propose that adding more grass margins to the landscape can increase parasitism rates. These habitats do not cover a large area in the landscape;

35 adding more of them to the cropping system is a realistic possibility even in areas under intensive crop production.

The results of this study offer useful and novel implications for modelling cropping systems that maintain or even increase the effectiveness of the biocontrol offered by the parasitoids of C. obstrictus.

36 REFERENCES

Alford DV, Nilsson C, Ulber B (2003) Insect pests of oilseed rape crops. In: Alford DV (ed) Biocontrol of oilseed rape pests. Blackwell Sci- ence Ltd, Oxford, UK, pp 9–42 Alford DV, Walters KFA, Williams IH, Murchie AK (1996) A commercially viable low-cost strategy for the management of oilseed weevil populations on winter oilseed rape in the UK. In: Proceedings of the Brighton crop protection conference – pests and diseases. pp 609–614 Barbosa P (1998) Agroecosystems and conservation biological control. In: Conservation biological control. Academic Press, San Diego, pp 39–54 Bartoń K (2016) Multi-Model Inference. R package version 1.15.6. Baur H, Muller FJ, Gibson GAP, Mason PG, Kuhlmann U (2007) A review of the species of Mesopolobus (Chalcidoidea: Pteromalidae) associated with Ceutorhynchus (Coleoptera: Curculionidae) host-species of European origin. Bull Entomol Res 97:387–397 Berger JS, Birkhofer K, Hanson HI, Hedlund K (2018) Landscape configuration affects herbivore–parasitoid communities in oilseed rape. J Pest Sci 91:1093–1105 Bianchi FJJA, Booij CJH, Tscharntke T (2006) Sustainable pest regulation in agricultural landscapes: a review on landscape composition, biodiversity and natural pest control. Proc R Soc B 273:1715–1727 Brown J, McCaffrey JP, Harmon BL, Davis JB, Brown AP, Erickson DA (1999) Effect of late season insect infestation on yield, yield components and oil quality of Brassica napus, B. rapa, B. juncea and Sinapis alba in the Pacific Northwest region of the United States. J Agric Sci 132:281–288 Büchi R (1993) Monitoring of parasitoids of the cabbage seed weevil, Ceutorhynchus assimilis during 1990 and 1991 in Switzerland. IOBC- WPRS Bull 16(9):145–149 Büchi R (1990) Investigations on the use of turnip rape as trap plant to control oilseed rape pests. IOBC-WPRS Bull 13(4):32–39 Büchs W, Nuss H (2000) First steps to assess the importance of epigaeic active polyphagous predators on oilseed rape insect pests with soil pupating larvae. IOBC-WPRS Bull 23(6):151–163

37 Buntin GD (1999) Damage loss assessment and control of the cabbage seedpod weevil (Coleoptera: Curculionidae) in winter canola using insecticides. J Econ Entomol 92:220–227 Butler D (2018) Scientists hail European ban on bee-harming pesticides. In: Nature. http://www.nature.com/articles/d41586-018-04987-4. Accessed 22 Jun 2018 Cárcamo HA, Brandt R (2017) Cabbage seedpod weevil management. In: Reddy GVP (ed) Integrated management of insect pests on canola and other Brassica oilseed crops. CAB International, Oxfordshire, UK, pp 77–87 Cárcamo HA, Dosdall L, Dolinski M, Olfert O, Byers JR (2001) The cabbage seedpod weevil, Ceutorhynchus obstrictus (Coleoptera: Cur- culionidae) - a review. J Entomol Soc Br Columbia 98:201–210 Cárcamo HA, Dunn R, Dosdall LM, Olfert O (2007) Managing cabbage seedpod weevil in canola using a trap crop - A commercial field scale study in western Canada. Crop Prot 26:1325–1334 Ceuppens B, Eeraerts M, Vleugels T, Cnops G, Roldan-Ruiz I, Smagghe G (2015) Effects of dietary lambda-cyhalothrin exposure on bumble- bee survival, reproduction, and foraging behavior in laboratory and greenhouse. J Pest Sci 88:777–783 Cook SM, Khan ZR, Pickett JA (2007a) The use of push-pull strategies in integrated pest management. Annu Rev Entomol 52:375–400 Cook SM, Rasmussen HB, Birkett MA, Murray DA, Pye BJ, Watts NP, Williams IH (2007b) Behavioural and chemical ecology underlying the success of turnip rape (Brassica rapa) trap crops in protecting oilseed rape (Brassica napus) from the pollen beetle (Meligethes aeneus). Arthropod-Plant Interact 1:57–67 Cook SM, Smart LE, Martin JL, Murray DA, Watts NP, Williams IH (2006) Exploitation of host plant preferences in pest management strategies for oilseed rape (Brassica napus). Entomol Exp Appl 119:221–229 Dmoch J (1965) The dynamics of a population of the cabbage seedpod weevil (Ceutorhynchus assimilis Payk.) and the development of winter rape. Part I. Ekol Pol–Pol J Ecol A 13: 249–287 Dmoch J (1975) Study on the parasites of the cabbage seed weevil (Ceu- thorrhynchus assimilis Payk.). I. Species composition and economic importance of the larval ectoparasites. Rocz Nauk Rol E 5:99–112

38 Dosdall LM, Kott LS (2006) Introgression of resistance to cabbage seedpod weevil to canola from yellow mustard. Crop Sci 46:2437 Dosdall LM, Mason PG (2010) Key pests and parasitoids of oilseed rape or canola in North America and the importance of parasitoids in integrated management. In: Williams IH (ed) Biocontrol-based integrated management of oilseed rape pests. Springer Netherlands, pp 167–213 Dosdall LM, Moisey DWA (2004) Developmental biology of the cabbage seedpod weevil, Ceutorhynchus obstrictus (Coleoptera: Curcu- lionidae), in spring canola, Brassica napus, in Western Canada. Ann Entomol Soc Am 97:458–465 Doucette CF (1947) Host plants of the cabbage seedpod weevil. J Econ Entomol 40:838–840 EFSA (2013a) Conclusion on the peer review of the pesticide risk assessment for bees for the active substance clothianidin. EFSA J 11:3066 EFSA (2013b) Conclusion on the peer review of the pesticide risk assessment for bees for the active substance thiamethoxam. EFSA J 11:3067 EFSA (2013c) Conclusion on the peer review of the pesticide risk assessment for bees for the active substance imidacloprid. EFSA J 11:3068 EFSA (2018) Evaluation of the data on clothianidin, imidacloprid and thiamethoxam for the updated risk assessment to bees for seed treatments and granules in the EU. EFSA Support Publ 15:1378E Estonian Land Board (2017) Web map of the Estonian Land Board. http://xgis.maaamet.ee/maps/XGis?app_id=UU82A&user_id=at& LANG=2&WIDTH=1304&HEIGHT=797&zlevel=0,552500,65 05000. Accessed 24 May 2107 European Commission (2013) Commission Implementing Regulation (EU) No 485/2013 OJ L 139, 12–26 European Commission (2009) Directive 2009/128/EC of the European Parliament and of the Council. OJ L 309, 71–86 Eurostat (2018) Crop production in EU standard humidity: oilseed rape and turnip rape seeds. http://ec.europa.eu/eurostat/data/database. Accessed 20 Jun 2018 Frank T, Drapela T, Moser D, Zaller JG (2010) Insect pests and spiders in oilseed rape and their response to site and landscape factors.

39 In: Williams IH (ed) Biocontrol-based integrated management of oilseed rape pests. Springer Netherlands, Dordrecht, pp 285–304 Free JB, Williams IH (1978a) A survey of the damage caused to crops of oil-seed rape (Brassica napus L.) by insect pests in south-central England and their effect on seed yield. J Agric Sci 90:417–424 Free JB, Williams IH (1979) The infestation of crops of oil-seed rape (Brassica napus L.) by insect pests. J Agric Sci 92:203–218 Free JB, Williams IH (1978b) The responses of the pollen beetle, Meligethes aeneus, and the seed weevil, Ceuthorhynchus assimilis, to oil-seed rape, Brassica napus, and other plants. J Appl Ecol 15:761–774 Geiger F, Wäckers FL, Bianchi FJJA (2009) Hibernation of predatory arthropods in semi-natural habitats. BioControl 54:529–535 Gibson GA, Gates MW, Buntin GD (2006) Parasitoids (Hymenoptera: Chalcidoidea) of the cabbage seedpod weevil (Coleoptera: Curculio- nidae) in Georgia, USA. J Hymenopt Res 15:187–207 Gill RJ, Ramos-Rodriguez O, Raine NE (2012) Combined pesticide exposure severely affects individual- and colony-level traits in bees. Nature 491:105–108 Godfray HCJ (1994) Introduction. In: Parasitoids: behavioral and evolutionary ecology. Princeton University Press, Princeton, New Jersey, pp 3–25 Graham MWR de V (1969) The Pteromalidae of North-Western Europe (Hymenoptera: Chalcidoidea). Bull Br Mus Nat Hist Entomol 16:1– 908 Gurr GM, van Emden HF, Wratten SD (1998) Chapter 9 - Habitat manipulation and natural enemy efficiency: Implications for the control of pests. In: Barbosa P (ed) Conservation biological control. Academic Press, San Diego, pp 155–183 Gurr GM, Wratten SD (1999) FORUM “Integrated biological control”: a proposal for enhancing success in biological control. Int J Pest Manag 45:81–84 Harmon BL, McCaffrey JP (1997) Laboratory bioassay to assess Brassica spp. germplasm for resistance to the cabbage seedpod weevil (Cole- optera: Curculionidae). J Econ Entomol 90:1392–1399 Hawkins BA, Cornell HV (1994) Maximum parasitism rates and successful biological control. Science 266:1886 Haye T, Olfert O, Weiss R, Mason PG, Gibson G, Gariepy TD, Gillespie DR (2018) Bioclimatic analyses of Trichomalus perfectus

40 and Mesopolobus morys (Hymenoptera: Pteromalidae) distributions, two potential biological control agents of the cabbage seedpod weevil in North America. Biol Control 124:30–39 Heimbach U, Müller A (2013) Incidence of pyrethroid-resistant oilseed rape pests in Germany. Pest Manag Sci 69:209–216 Hokkanen HMT (1991) Trap cropping in pest management. Annu Rev Entomol 36:119–138 Hokkanen HMT, Granlund H, Husberg G-B, Markkula M (1986) Trap crops used successfully to control Meligethes aeneus (Col., Nitiduli- dae), the rape blossom beetle. Ann Entomol Fenn 52:115–120 Holland JM, Bianchi FJ, Entling MH, Moonen AC, Smith BM, Jean- neret P (2016) Structure, function and management of semi-natural habitats for conservation biological control: a review of European studies. Pest Manag Sci 72:1638–1651 Holland JM, Douma JC, Crowley L, James L, Kor L, Stevenson DRW, Smith BM (2017) Semi-natural habitats support biological control, pollination and soil conservation in Europe. A review. Agron Sustain Dev 37:31 Holland JM, Jeanneret P, Herzog F, Moonen AC, Rossing W, van der Werf W, Kiss J, Helden VM, Paracchini ML, Cresswell J, Pointereau P, Heijne B, Veromann E, Antichi D, Entling M, Balázs B (2014) The QuESSA Project: quantification of ecological services for sustainable agriculture. IOBC-WPRS Bull 100:55–58 Holland JM, Oaten H, Moreby S, Birkett T, Simper J, Southway S, Smith BM (2012) Agri-environment scheme enhancing ecosystem services: A demonstration of improved biological control in cereal crops. Agric Ecosyst Environ 155:147–152 Idris AB, Grafius E (1996) Effects of wild and cultivated host plants on oviposition, survival, and development of diamondback moth (Lepi- doptera: Plutellidae) and its parasitoid Diadegma insulare (Hymenop- tera: Ichneumonidae). Environ Entomol 25:825–833 Jonsson M, Wratten SD, Landis DA, Gurr GM (2008) Recent advances in conservation biological control of arthropods by arthropods. Biol Control 45:172–175 Kaasik R, Kovács G, Kaart T, Metspalu L, Williams IH, Veromann E (2014a) Meligethes aeneus oviposition preferences, larval parasitism rate and species composition of parasitoids on Brassica nigra,

41 Raphanus sativus and Eruca sativa compared with on Brassica napus. Biol Control 69:65–71 Kaasik R, Kovács G, Toome M, Metspalu L, Veromann E (2014b) The relative attractiveness of Brassica napus, B. rapa, B. juncea and Sinapis alba to pollen beetles. BioControl 59:19–28 Kalischuk AR, Dosdall LM (2004) Susceptibilities of seven Brassicaceae species to infestation by the cabbage seedpod weevil (Coleoptera: Curculionidae). Can Entomol 136:265–276 Khan ZR, Pickett J (2004) The “push-pull” strategy for stem borer management: a case study in exploiting biodiversity and chemical ecology. In: Gurr GM, Wratten SD, Altieri MA (eds) Ecological engi- neering for pest management - Advances in habitat manipulation for arthropods. CSIRO Publishing, pp 155–164 Kozlowski MW, Lux S, Dmoch J (1983) Oviposition behaviour and pod marking in the cabbage seed weevil, Ceutorhynchus assimilis. Entomol Exp Appl 34:277–282 Kruess A (2003) Effects of landscape structure and habitat type on a plant-herbivore-parasitoid community. Ecography 26:283–290 Kruess A, Tscharntke T (1994) Habitat fragmentation, species loss, and biological control. Sci New Ser 264:1581–1584 Lancashire PD, Bleiholder H, Boom TVD, Langelüddeke P, Stauss R, Weber E, Witzenberger A (1991) A uniform decimal code for growth stages of crops and weeds. Ann Appl Biol 119:561–601 Landis DA, Wratten SD, Gurr GM (2000) Habitat management to conserve natural enemies of arthropod pests in agriculture. Annu Rev Entomol 45:175–201 Lerin J (1984) Assessment of yield loss in winter rape due to seedpod weevil (Ceutorhynchus assimilis Payk.) II. Yield loss in a cage experiment. Agronomie 4:147–154 Lõiveke H (2003) Kahjurid ja nende tõrje. In: Kaarli K (ed) Õlikultu- uride kasvataja käsiraamat. Eesti Maaviljeluse Instituut, Saku, Esto- nia, pp 76–81 McCaffrey JP, Harmon BL, Brown J, Brown AP, Davis JB (1999) Assess- ment of Sinapis alba, Brassica napus and S. alba x B. napus hybrids for resistance to cabbage seedpod weevil, Ceutorhynchus assimilis (Cole- optera: Curculionidae). J Agric Sci 132:289–295 Moreby SJ, Southway S, Barker A, Holland JM (2001) A comparison of the effect of new and established insecticides on nontarget inver-

42 tebrates of winter wheat fields. Environ Toxicol Chem 20:2243– 2254 Muller FJ, Baur H, Gibson GAP, Mason PG, Kuhlmann U (2007) Review of the species of Trichomalus (Chalcidoidea: Pteromalidae) associated with Ceutorhynchus (Coleoptera: Curculionidae) host species of European origin. Can Entomol 139:643–657 Murchie AK (1996) Parasitoids of cabbage seed weevil and brassica pod midge in oilseed rape. PhD Thesis, University of Keele Murchie AK, Williams IH (1998) A bibliography of the parasitoids of the cabbage seed weevil (Ceutorhynchus assimilis Payk.). IOBC-WPRS Bull 21:163–169 Murchie AK, Williams IH, Alford DV (1997) Effects of commercial insecticide treatments to winter oilseed rape on parasitism of Ceuto- rhynchus assimilis Paykull (Coleoptera: Curculionidae) by Trichoma- lus perfectus (Walker) (Hymenoptera: Pteromalidae). Crop Prot 16:199–202 Parolin P, Bresch C, Desneux N, Brun R, Bout A, Boll R, Poncet C (2012a) Secondary plants used in biological control: a review. Int J Pest Manag 58:91–100 Parolin P, Bresch C, Poncet C, Desneux N (2014) Introducing the term “Biocontrol Plants” for integrated pest management. Sci Agric 71:77–80 Parolin P, Bresch C, Poncet C, Desneux N (2012b) Functional characteristics of secondary plants for increased pest management. Int J Pest Manag 58:369–377 PRIA (2017) Interactive map of the Estonian Agricultural Registers and Information Board. https://kls.pria.ee/kaart/. Accessed 24 May 2017 QuESSA Europe (2013) Quantification of Ecological Services for Sustain- able Agriculture. http://www.quessa.eu/. Accessed 13 Nov 2017 Rusch A, Valantin-Morison M, Roger-Estrade J, Sarthou J-P (2012) Local and landscape determinants of pollen beetle abundance in overwintering habitats. Agric For Entomol 14:37–47 Rusch A, Valantin-Morison M, Sarthou JP, Roger-Estrade J (2010) Inte- grating crop and landscape management into new crop protection strategies to enhance biological control of oilseed rape insect pests. In: Williams IH (ed) Biocontrol-based integrated management of oilseed rape pests. Springer Netherlands, Dordrecht, pp 415–448

43 Rusch A, Valantin-Morison M, Sarthou J-P, Roger-Estrade J (2011) Multi-scale effects of landscape complexity and crop management on pollen beetle parasitism rate. Landsc Ecol 26:473–486 Shelton AM, Nault BA (2004) Dead-end trap cropping: a technique to improve management of the diamondback moth, Plutella xylostella (Lepidoptera: Plutellidae). Crop Prot 23:497–503 Shelton AM, Badenes-Perez FR (2006) Concepts and applications of trap cropping in pest management. Annu Rev Entomol 51:285–308 Sunderland KD (2002) Invertebrate pest control by carabids. In: Holland JM (ed) The agroecology of carabid beetles. Intercept, Andover, UK, pp 165–214 Tansey JA, Dosdall LM, Keddie A, Fletcher RS, Kott LS (2010) Antix- enosis and antibiosis resistance to Ceutorhynchus obstrictus in novel germplasm derived from Sinapis alba x Brassica napus. Can Entomol 142:212–221 Thies C, Steffan-Dewenter I, Tscharntke T (2003) Effects of landscape context on herbivory and parasitism at different spatial scales. Oikos 101:18–25 Thies C, Tscharntke T (2010) Biological rape pest control in spatio-temporally changing landscapes. In: Williams IH (ed) Biocontrol-based integrated management of oilseed rape pests. Springer Netherlands, pp 273–284 Treier K, Kovács G, Kaasik R, Männiste M, Veromann E (2017) The abundance of overwintered predatory arthropods in agricultural landscape elements. IOBC-WPRS Bull122:68–73 Trjapitzin VA (1978) Hymenoptera III, 2. Chalcidoidea 18. Mymaridae. In: Medvedev GS (ed) Keys to the insects of the European part of the USSR. Academy of Sciences of the USSR, Institute of Zoology, Nauka, Leningrad, USSR, pp 516–537 Tscharntke T, Bommarco R, Clough Y, Crist TO, Kleijn D, Rand TA, Tylianakis JM, Nouhuys Sv, Vidal S (2007) Conservation biological control and enemy diversity on a landscape scale. Biol Control 43:294–309 Tscharntke T, Klein AM, Kruess A, Steffan-Dewenter I, Thies C (2005) Landscape perspectives on agricultural intensification and biodiversity – ecosystem service management. Ecol Lett 8:857–874 Ulber B, Klukowski Z, Williams IH (2010a) Impact of insecticides on parasitoids of oilseed rape pests. In: Williams IH (ed) Biocontrol-

44 based integrated management of oilseed rape pests. Springer Neth- erlands, pp 337–355 Ulber B, Vidal S (1998) Influence of host density and host distribution on parasitism of Ceutorhynchus assimilis by Trichomalus perfectus. IOBC-WPRS Bull 21(5):185–195 Ulber B, Williams IH, Klukowski Z, Luik A, Nilsson C (2010b) Para- sitoids of oilseed rape pests in Europe: key species for conservation biocontrol. In: Williams IH (ed) Biocontrol- based integrated management of oilseed rape pests. Springer Netherlands, pp 45–76 Ulmer BJ, Dosdall LM (2006a) Glucosinolate profile and oviposition behavior in relation to the susceptibilities of Brassicaceae to the cabbage seedpod weevil. Entomol Exp Appl 121:203–213 Ulmer BJ, Dosdall LM (2006b) Spring emergence biology of the cabbage seedpod weevil (Coleoptera: Curculionidae). Ann Entomol Soc Am 99:64–69 USDA (2018a) Oilseeds and products annual. USDA-FAS GAIN report AU1803. Available at: https://gain.fas.usda.gov/Recent%20GAIN% 20Publications/Oilseeds%20and%20Products%20Annual_Vienna_ EU-28_3-29-2018.pdf Accessed 20 June 2018 USDA (2018b) Oilseeds: World markets and trade. USDA-FAS June 2018. Available at: https://apps.fas.usda.gov/psdonline/circulars/ oilseeds.pdf Accessed 20 June 2018 Veromann E, Kaasik R, Kovács G, Metspalu L, Williams IH, Mänd M (2014) Fatal attraction: search for a dead-end trap crop for the pollen beetle (Meligethes aeneus). Arthropod-Plant Interact 8:373–381 Veromann E, Luik E, Metspalu L, Williams I (2006) Key pests and their parasitoids on spring and winter oilseed rape in Estonia. Entomol Fenn 17:400–404 Veromann E, Metspalu L, Williams IH, Hiiesaar K, Mand M, Kaasik R, Kovács G, Jogar K, Svilponis E, Kivimagi I, Ploomi A, Luik A (2012) Relative attractiveness of Brassica napus, Brassica nigra, Eruca sativa and Raphanus sativus for pollen beetle (Meligethes aeneus) and their potential for use in trap cropping. Arthropod-Plant Interact 6:385–394 Veromann E, Toome M, Kännaste A, Kaasik R, Copolovici L, Flink J, Kovács G, Narits L, Luik A, Niinemets Ü (2013) Effects of nitrogen fertilization on insect pests, their parasitoids, plant diseases and volatile organic compounds in Brassica napus. Crop Prot 43:79–88

45 Veromann E, Williams IH, Kaasik R, Luik A (2011) Potential of parasitoids to control populations of the weevil Ceutorhynchus obstrictus (Marsham) on winter oilseed rape. Int J Pest Manag 57:85–92 Virk JS, Brar KS, Sohi AS (2004) The role of trap crops in increasing parasitisation efficiency of Trichogramma chilonis Ishii in cotton. J Biol Control 18:61–64 Walters KFA, Young JEB, Kromp B, Cox PD (2003) Management of oilseed rape pests. In: Alford DV (ed) Biocontrol of oilseed rape pests. Blackwell Science Ltd, Oxford, UK, pp 43–71 Williams IH (2010) The major insect pests of oilseed rape in Europe and their management: an overview. In: Williams IH (ed) Biocontrol- based integrated management of oilseed rape pests. Springer Neth- erlands, Dordrecht, pp 1–43 Williams IH (2003) Parasitoids of cabbage seed weevil. In: Alford DV (ed) Biocontrol of oilseed rape pests. Blackwell Science Ltd, Oxford, UK, pp 97–112 Williams IH, Ferguson AW (2010) Spatio-temporal distributions of pests and their parasitoids on the oilseed rape crop. In: Williams IH (ed) Biocontrol-based integrated management of oilseed rape pests. Springer Netherlands, pp 245–271 Williams IH, Ferguson AW, Kruus M, Veromann E, Warner DJ (2010) Ground beetles as predators of oilseed rape pests: incidence, spatio-temporal distributions and feeding. In: Williams IH (ed) Bio- control-based integrated management of oilseed rape pests. Springer Netherlands, pp 115–149 Zaller JG, Moser D, Drapela T, Schmöger C, Frank T (2008a) Effect of within-field and landscape factors on insect damage in winter oilseed rape. Agric Ecosyst Environ 123:233–238 Zaller JG, Moser D, Drapela T, Schmöger C, Frank T (2008b) Insect pests in winter oilseed rape affected by field and landscape characteristics. Basic Appl Ecol 9:682–690 Zimmer CT, Köhler H, Nauen R (2014) Baseline susceptibility and insecticide resistance monitoring in European populations of Meligethes aeneus and Ceutorhynchus assimilis collected in winter oilseed rape. Entomol Exp Appl 150:279–288

46 SUMMARY IN ESTONIAN

PEREMEESTAIME JA MAAKASUTUSE MÕJU KÕDRA-PEITKÄRSAKA KAHJUSTUSE JA PARASITEERITUSE TASEMELE

Raps (Brassica napus L., Capparales: Brassicaceae) on üks tähtsamaid parasvöötmes kasvatatavaid ristõielisi kultuure ja peamisi õlikultuure Euroopa Liidus (USDA 2018). Ohtlik rapsi kahjur on kõdra-peit- kärsakas (Ceutorhynchus obstrictus Marsham, Coleoptera: Curculionidae) (Alford et al. 2003), kelle vastsed toituvad ristõieliste kõtrades arene- vatest seemnetest. Seetõttu võivad need univoltiinsed mardikad teki- tada olulist kahju saagile nii Euroopa (Alford et al. 2003; Williams 2010) kui ka Põhja-Ameerika rapsikasvatuspiirkondades (Kalischuk, Dosdall 2004; Ulmer, Dosdall 2006). Euroopa rapsikasvatajad kasuta- vad putukkahjurite arvukuse piiramiseks peamiselt sünteetilisi pestit- siide (neonikotinoide ja püretroide), kuid need mõjuvad surmavalt ka sihtrühma mittekuuluvatele organismidele, sealhulgas kasulikele lüli- jalgsetele (nt tolmeldajad, röövtoidulised lülijalgsed ja parasitoidid). Peale selle võib pestitsiidide sagedase kasutamise tagajärjel kujuneda kahjuritel resistentsus toimeainete suhtes. Neonikotinoidide kahjuliku mõju tõttu tolmeldajatele keelas Euroopa Komisjon 2013. a kolme toi- meaine (clothianidin, thiamethoxam ja imidacloprid) kasutamise tolmeldajatele atraktiivsetel põllukultuuridel. Seepärast on alternatiivsete tõrjemeetmete väljatöötamine aktuaalne nii teadlastele kui ka põllu- meestele. Alates 2014. aastast on Euroopa Liidu ühtseks põllumajan- duspoliitikaks integreeritud taimekaitse, mis seab esikohale mitmekesise külvikorra kasutamise, minimeeritud maaharimispraktika, vähendatud sekkumise agroökosüsteemidesse, bioloogilise mitmekesisuse suuren- damise põllumajandusmaastikul ja kahjurite looduslike vaenlaste esi- nemise soodustamise (European Commission 2009).

Kiletiivalised parasitoidid (Hymenoptera: Parasitica) aitavad looduslikult vähendada kõdra-peitkärsaka arvukust ja hoiavad selle rapsipõldudel allpool majandusliku tõrjekriteeriumi lävendit (Williams 2003; Dosdall, Mason 2010; Ulber et al. 2010; Veromann et al. 2011). Suurendades bioloogilist mitmekesisust ja põllumajandusmaastiku heterogeensust, saab soodustada ja säilitada bioloogilise tõrje agentide esinemist ja arvukust, et sel moel vähendada sünteetiliste taimekaitsevahendite kasutamist

47 (Barbosa 1998; Gurr et al. 1998; Jonsson et al. 2008). Kultuurpõllu tasandil saab elurikkuse suurendamiseks lisada põllule näiteks biotõrje- taimi (atraktantsed ja/või repellentsed kahjuritele või atraktantsed parasitoididele jt kasuritele) (Parolin et al. 2012, 2014). Maastiku tasandil saab manipuleerida mitmesuguste maastikuelementide kui kasurite elu-, toitumis- ja talvitumispaikadega, muutes nende proportsioone ja asu- kohta kultuurpõldude suhtes ning vähendades põllumajanduskeemia potentsiaalset triivi keskkonda. Selleks et neid võtteid edukalt rakendada, on vaja teada saada, millised on kõige tõhusamad biotõrjetaimed ja millised põllumajandusmaastikul esinevad elupaigatüübid toetavad kõige paremini bioloogilist tõrjet.

Rapsi teise olulise kahjuri, naeri-hiilamardika (Brassicogethes aeneus Fabricius syn. Meligethes aeneus Fabricius, Coleoptera: Nitidulidae) toitumis- ja paljunemiseelistuste kohta peremeestaimedel ning ka tema parasitoidide eelistuste kohta on Euroopas suhteliselt palju uuringuid tehtud. Samas on kõdra-peitkärsaka peremeestaime-eelistust uuritud peamiselt vaid Põhja-Ameerikas ja parasitoidide eelistuste alased uurin- gud peaaegu puuduvad.

Suurem osa maastikuskaalal tehtud uuringuid on pärit Lääne-Euroopast, seevastu Ida-Euroopa kohta on kirjandusandmeid väga vähe (Holland et al. 2017). Veelgi enam, kõdra-peitkärsakat pole maastiku tasandil üldse uuritud. Siinne uuring täidabki seda lünka ja käsitleb peremeestaimede, kõdra-peitkärsaka ja parasitoidide tritroofilisi suhteid ning põllumajan- dusmaastiku elementide mõju kõdra-peitkärsaka kahjustuse ja parasiteerituse tasemele.

Töö peamiseks eesmärgiks oli leida keskkonnasõbralikke ja jätkusuutlikke võimalusi kõdra-peitkärsaka kui olulise rapsikahjuri tõrjeks.

Sellest tulenevalt olid tööülesanneteks välja selgitada kas ja kuidas mõju- tavad eri taimeliigid (I, II) ning põllumajandusmaa kasutus (III, IV) kõdra-peitkärsaka ja tema parasitoidide peremeestaimede otsingu edukust.

48 Käesoleva uurimistöö hüpoteesid olid järgmised:

1. Kiletiivalised parasitoidid piiravad edukalt kõdra-peitkärsaka arvukust. 2. Kõdra-peitkärsaka kahjustuse ja parasiteerituse taset ning parasitoidide liigilist koosseisu mõjutab peremeestaime liik 3. Kõdra-peitkärsaka ja tema parasitoidide peremeesorganismide otsingu edukust mõjutavad rapsipõlluga piirnevad ja seda ümbritse- vad maastikuelemendid.

Esiteks, selleks et leida potentsiaalseid biotõrjetaimi, mis aitaksid suuren- dada kõdra-peitkärsaka parasiteerituse taset, hinnati väikesemahulistes lapikatsetes aastatel 2007–2012 kahjuri arvukust, tema parasiteerituse taset ja parasitoidide liigilist koosseisu nii suvirapsil kui ka teistel ristõie- listel taimedel (I, II). Eri aastatel katsetes kasutatud taimeliigid olid järg- mised: suviraps (B. napus), kapsasrohi (B. juncea), must rõigas (B. nigra), suvirüps (B. rapa), rukola (Eruca vesicaria subsp. sativa = E. sativa), õli- rõigas (Raphanus sativus var. oleiformis = R. sativus) ja valge sinep (Sinapis alba).

Teiseks, selleks et leida maastikutüüp, mis pärsiks kõdra-peitkärsaka levi- kut, uuriti kahjustuse ja parasiteerituse taset ning ka parasitoidide liigilist koosseisu erinevate maastikuelementidega piirnevatel talirapsipõldudel (III, IV) ja erinevate maastikuelementide osakaalu mõju eelnimetatud näitajatele (IV), näiteks seda, kuidas kultuurpõldude, mitteharitavate alade, metsa jne osakaal maastikus mõjutab kahjustuse suurust talirapsi põllul. Maastikuskaalal tehtud uuringute analüüsil kasutati uudset, ruu- miliselt täpset metoodikat, et kaaluda erinevate maastikul asuvate elu- paikade toimet kahjustuse ja parasiteerituse tasemele olenevalt nende kaugusest katsetes uuritud talirapsipõldudest (IV).

Katsematerjal koguti taimede kõtrade küpsemise staadiumis (BBCH 80–85; Lancashire et al. 1991) (I, II, III, IV). Kõtru hoiti väljakasva- tuspüünistes toatemperatuuril 30 päeva, pärast seda loendati kõtradest väljunud vastsed ja parasitoidid. Ühtlasi avati kõik kõdrad ja kontrolliti hoolikalt üle; kõik augud ning vastsete ja parasitoidide jäänused tuvastati ja loendati. Parasitoidi valmikud määrati liigini. Statistilises analüüsis loeti sõltuvateks tunnusteks kahjustuse ja parasiteerituse tase (I, II, III, IV).

49 Töö tulemused võib kokku võtta järgmiselt.

Uuringu praktiliselt rakendatavad ja uudsed tulemused aitavad oluliselt kaasa kõdra-peitkärsaka (C. obstrictus) arvukuse reguleerimisele keskkon- nasõbralike looduslike bioloogilise tõrje mehhanismide kaudu.

Uuringust selgus, et looduslikult esinevad parasitoidid on võimelised edukalt piirama kõdra-peitkärsaka arvukust ja hoidma seda allpool majanduslikku tõrjekriteeriumi hoolimata peremeesorganismi vähesusest (I, II, III, IV). Seega leidis töö esimene hüpotees kinnitust. Tõsiasi, et looduslikult esinevad parasitoidid piiravad tõhusalt kõdra-peitkärsaka arvukust, on väga väärtuslik ja oluline informatsioon rapsikasvatajatele, tõestades, et selle kahjuri vastu ei ole rapsipõldudel vaja kasutada sünteetilisi taime- kaitsevahendeid.

Suurem osa tulemusi kinnitas ka teist hüpoteesi. Peremeestaimede liigiline kuuluvus mõjutas nii kõdra-peitkärsaka kahjustuse suurust kui ka parasitoidide liigilist koosseisu. Samas ei mõjutanud see parasiteerituse taset (I, II). Peale selle mõjutas põllumajandusmaastiku kasutus nii kahjustuse kui ka parasiteerituse määra, kuid ei mõjutanud parasitoidide liigilist koosseisu (IV).

Analüüsides viie aasta andmeid koos (I+II), selgus, et rapsil (B. napus) oli kahjustatud kõtrade osakaal kõige suurem, keskmiselt 10,5%, jäädes siiski oluliselt allapoole majanduslikku kahju tekitavat tõrjekriteeriumi lävendit (26–40%; Free, Williams 1978; Lerin 1984; Buntin 1999). Hoolimata sellest et kahjuri arvukus oli väike, leiti parasiteeritud vastseid kõikide kahjustatud peremeestaimede kõtradest välja arvatud rukolalt (E. sativa). Seejuures oli parasiteerituse tase märkimisväärselt kõrge, ulatudes kõiki- del liikidel üle 34% ja õlirõikal (R. sativus) üle 80%. Peale erakordselt kõrge parasiteerituse taseme leiti õlirõikal ka teistest erinev parasitoidide liigiline koosseis. Nimelt parasiteerisid seal kõdra-peitkärsakat ka muna- parasitoidid (Mymaridae), samal ajal kui teistel taimeliikidel dominee- risid vastseparasitoidid sugukonnast Pteromalidae. Seega võib öelda, et suurendades ristõieliste taimede mitmekesisust kultuurpõldudel, saab potentsiaalselt mitmekesistada parasitoidide liigilist koosseisu, pakkudes elupaiku ja peremeesorganisme ka vähem arvukatele parasitoidiliikidele (II), ning seega tõhustada bioloogilist tõrje.

50 Esmakordselt leiti seos kõdra-peitkärsaka arvukuse ja põllumajandus- maastiku elementide vahel (IV). Maastikuskaalal leiti, et ühekilomeetrise raadiusega maastikuringi keskel oleval rapsipõllul mõjutasid positiivselt kõtrade kahjustuse taset poollooduslikud rohumaad, nisupõllud (mis külvikorras järgnevad tavaliselt talirapsile) ja püsirohumaad. Kui püsirohu- maad asetsesid mõnesaja meetri kaugusel katses kasutatud rapsipõllust, mõjutasid nad ka kõdra-peitkärsaka parasiteerituse taset, kuid mõju oli negatiivne. Parasiteerituse tase oli kõrgem nendel talirapsipõldudel, mille läheduses (~750 m) asetsesid rohtsed põlluservad. Samas leiti ka posi- tiivne seos kahjustatud kõtrade ja rohtsete põlluservade vahel, kui need vahetult piirnesid uurimisaluse põlluga. Seega võib maastikuelementide mõju varieeruda olenevalt skaalast ja mõni element võib soodustada nii kahjurite kui ka tema looduslike vaenlaste esinemist. Saadud tulemused viitavad sellele, et maastikuskaalal mängivad poollooduslikud elupaigad olulist rolli nii kahjurite kui ka kasurite elutsüklis. Ent kuigi rohtsed põllu- servad suurendasid kahjustatud kõtrade arvukust rapsipõllul, oli nendega piirnenud põldudel ka väga kõrge parasiteerituse tase. Peale selle soo- dustasid rohtsed põlluservad parasiteerituse taset kogu maastikuskaalal ja seega ei tohi nende lihtsalt loodavate ja kergesti hooldatavate maastikuelementide tähtsust ökosüsteemi teenuste pakkujana alahinnata.

Uuringu tulemused täiendavad oluliselt meie teadmisi nii kõdra-peit- kärsaka kui ka tema parasitoidide leviku bioloogiast ning neid teadmisi saab rakendada ka külvikorra planeerimisel, et hoida kõdra-peitkärsaka arvukus kontrolli all.

51 ACKNOWLEDGEMENTS

An unforgettable journey has come to an end. I am forever grateful to all of those who helped me get through the difficulties and made this thesis possible.

I would like to express my deepest and heartfelt gratitude to associate professor Eve Veromann and professor emeritus Anne Luik, my supervisors, for their encouragement and confidence in me throughout my studies. Under your supervision I overcame many obstacles and learned a lot. Eve, thank you for your continuous support, guidance, and friendship during these years. Anne, thank you for your unstoppable enthusiasm and help.

Special thanks to my workmate Riina Kaasik for her constant support through these years. Riina, your friendship has helped me get through many-many-many moments of crisis. Thank you for the truckload of good times!

Sincere thanks to all my dear fellow doctoral students and colleagues (current and former) for the warm welcome, and for their friendly and caring attitudes during my study years at the chair of plant health. Without the help provided by many of you, and numerous dedicated undergraduate and master’s students during field and lab work, this thesis would not have been possible. The work was demanding but you made it easier and fun. I thank you all!

I am deeply indebted to the Veromann family: Peeter, thank you for being always available to help with design both on paper and in the field; Linda- Liisa and Martin, and Kai-Riin and Andi, thanks for helping in the field and in the lab. Many thanks also to Merje and Greg for their kindness and support.

My heartfelt gratitude goes to my dear co-authors through the years for sharing their invaluable knowledge and insight, and for devoting their precious time to help make the papers happen! I am particularly grateful to Tanel Kaart for providing statistical assistance.

52 I would like to thank Eda Tursk and Lilian Ariva for the academic and administrative support during my study years.

I sincerely thank Triinu Remmel and Jaan Liira for carefully reviewing previous versions of this thesis and giving constructive feedback. I am grateful to Kalle Hein for editing the Estonian summary. I also wish to thank Eesti Loodusfoto for greatly improving the layout of this thesis.

I could not have completed my studies without the encouragement of my closest and fondest to whom I dedicate this thesis: all members of my family and all my friends. I wish to especially thank my parents and my brother for their continuous and unconditional love, care, and support throughout my life. I am extremely grateful to my relatives in Tartu, especially to my dear late grandfather Heino, for welcoming me and keeping me grounded. I also thank my friends, near and far, for making the world a brighter and happier place, and reminding me of what is important. Ivar, your support and patience has helped me get through the worst. I can’t thank you enough.

Köszönöm nektek! Aitäh teile kõigile!

This study has received funding from the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 311879 (QuESSA). This study was also supported by the Institutional Research Funding project no. IUT36- 2 of the Estonian Research Council, the grant no. 8895 of the Estonian Science Foundation, the Capacity-building project no. P9003PKPK of the Estonian University of Life Sciences, the Targeted Financing project no. SF0170057s09 of the Estonian Ministry of Education and Research, the European Social Fund’s Doctoral Studies and Internationalisation Programme DoRa, carried out by Foundation Archimedes.

ORIGINAL PUBLICATIONS Kovács, G.; Kaasik, R.; Metspalu, L.; Williams, I.H.; Luik, A.; Veromann E. 2013. Could Brassica rapa, Brassica juncea and Sinapis alba facilitate the control of the cabbage seed weevil in oilseed rape crops? Biological Control, 65: 124–129 Biological Control 65 (2013) 124–129

Contents lists available at SciVerse ScienceDirect

Biological Control

journal homepage: www.elsevier.com/locate/ybcon

Could Brassica rapa, Brassica juncea and Sinapis alba facilitate the control of the cabbage seed weevil in oilseed rape crops?

Gabriella Kovács ⇑, Riina Kaasik, Luule Metspalu, Ingrid H. Williams, Anne Luik, Eve Veromann

Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Kreutzwaldi 1, Tartu 51014, Estonia h i g h l i g h t s graphical abstract

" Plant species inﬂuenced the species composition of parasitoids. " The studied plant species did not increase the pest infestation. " The parasitism level reached up to 85.4% on Brassica juncea.

article info a b s t r a c t

Article history: A two-year small plot experiment was performed to test whether (1) other plants in the cruciferous fam- Received 27 July 2012 ily: Brassica juncea (L.) Czern. and Sinapis alba L. are more attractive to the cabbage seed weevil (Ceu- Accepted 30 January 2013 torhynchus obstrictus (Marsham)) than turnip rape (Brassica rapa Metzg.) and oilseed rape (Brassica Available online 9 February 2013 napus L.); (2) the host plant species affects the oviposition preferences of parasitoid species attacking the cabbage seed weevil. According to our results oilseed rape had the highest pest infestation. The par- Keywords: asitism rate was highest on B. juncea and B. rapa. Thus, we can assume that wild relatives of oilseed rape Ceutorhynchus obstrictus plants can support the presence of hymenopteran natural enemies of pests in the ﬁeld. The parasitism Parasitoids B juncea C obstrictus Trichomalus perfectus rate was surprisingly high, up to 85.4% on . . In addition to the key parasitoids of . , spe- Mesopolobus morys cies from the family Mymaridae were also found. The species composition of parasitoids differed between Host plant attractiveness investigated plant species, suggesting that the physical or volatile characteristics of closely related plant Anaphes sp. species affect the behavioural responses of C. obstrictus parasitoids and therefore could be used to enhance the natural enemy complex. � 2013 Elsevier Inc. All rights reserved.

1. Introduction methods. However, several recent studies have shown that growing more attractive plants for pests at the edges of the main crop Plant species diversity may facilitate natural pest control in an- can reduce its infestation rate (Cook et al., 2006, 2007a,b; Vero- nual cropping systems (Zehnder et al., 2007). However, current mann et al., 2012). For example, growing of Brassica rapa Metzg. agricultural practice tends to lead to crop monocultures without at the edges of an oilseed rape (Brassica napus L.) crop noticeably any other plant species within or adjacent to the crops. The most reduces damage caused by Meligethes aeneus (Fab.) (Cook et al., undesirable weeds for farmers are those belonging to the same 2007a,b). However, commercial management of oilseed rape pests family as the main crop because they require complicated control still relies on a heavy input of pesticides despite their harmful side–effects on functional agro–biodiversity through killing beneﬁ-

Corresponding author. Fax: +372 7313511. cial arthropods as well as pests (Corbet et al., 1991; Miranda et al., ⇑ E-mail address: [email protected] (G. Kovács). 2003; Nilsson, 1994; Nitzsche, 1998; Veromann et al., 2008).

57 G. Kovács et al. / Biological Control 65 (2013) 124–129 125

Therefore, more sustainable and environment friendly crop protec- Our objectives were to compare the cabbage seed weevil’s ovi- tion approaches are needed. To amplify the effectiveness of natu- position preferences for B. rapa, B. juncea, S. alba and B. napus, and rally occurring biological control agents it is possible to employ to study if and how the host finding success of parasitoids and their the plant community to support the delivery of these ecosystem species composition is influenced by the host plant. services. Several pests of oilseed rape are known. Ceutorhynchus obstric- 2. Materials and methods tus (Marsham) (syn. Ceutorhynchus assimilis Paykull) (Coleoptera: Curculionidae) is one of the serious ones and is attacked by parasit- 2.1. Experimental field and set-up oids at all of its life stages (Murchie and Williams, 1998), in total by 34 species (Williams, 2003). The most common and best studied The study was carried out in an experimental field of the Esto- ones are three larval parasitoids belonging to the family Pteromal- nian University of Life Sciences in 2007 and 2008 in Tartu County, idae (Hymenoptera: Chalcidoidea): Stenomalina gracilis (Walker), Estonia. The plants were grown in a randomized complete block Mesopolobus morys (Walker) and Trichomalus perfectus (Walker). design with the plot size of 1 5 m and three replicates of each � The parasitism rate of C. obstrictus can be as high as 90% in Estonia plant species: oilseed rape (B. napus L.), turnip rape (B. rapa Met- (Veromann et al., 2011, 2013) and over 50% in several other Euro- zg.), mustard greens (B. juncea L.) and white mustard (S. alba L.). pean countries, e.g. in Germany (Nissen, 1997), Switzerland (Büchi, In order to minimize inter–plot interactions there was a 1 m wide 1993) and the UK (Murchie, 1996). buffer zone of bare soil around each plot. All plots were sown on 5 The host searching behaviour of hymenopteran parasitoids and May in 2007 and on 15 May in 2008, with the density of 250 seeds herbivorous insects is mainly based on the same signals. Firstly, per m2; the seeds used were obtained from the seed collection of volatile compounds emitted by host plants (De Moraes and the Estonian University of Life Sciences. Crop management was Mescher, 1998; Dicke, 1999; McCall et al., 1993; Turlings et al., uniform in all plots. Plant growth stage (BBCH) was assessed every 1991), although to locate a host within its habitat chemical, visual, 3–4 days in both years using the decimal code system of Lancashire acoustical and/or vibrational stimuli can also be used (Dorn et al., et al. (1991). 1999; van Alphen and Jervis, 1996). Furthermore, semiochemicals released specifically during herbivore feeding attract parasitoids to 2.2. Sampling and quantification of C. obstrictus and its parasitoids attack the herbivore (Bottrell et al., 1998; Dicke et al., 1990). Para- sitoids are influenced by several plant characteristics such as plant To determine the abundance of C. obstrictus, the damage it size, shape, pubescence, epicuticular waxes etc. (Bottrell et al., caused and its parasitism level, 10 pods were collected from 10 1998; Price et al., 1980). In addition to direct impact, indirect inter- randomly chosen plants of each plot at the mature pod stage (BBCH actions between plants and parasitoids also occur via the plants’ 83–85) in both years (Table 1). Pods were collected from the main influence on herbivorous host quality. Similarly, parasitoid and raceme and the third branch, placed into emergence traps (de- herbivore fitness are influenced by plant quality (Awmack and scribed in detail in Veromann et al., 2011) and incubated for four Leather, 2002; Stiling and Moon, 2005; Price et al., 1980). weeks under laboratory conditions. Emerged weevil larvae and In our study, we compared the relative attractiveness of closely adult parasitoids were then counted and identified. The adult par- related cruciferous plant species: B. rapa, Brassica juncea (L.) and asitoids were identified with an Olympus SZ-CTV binocular (Olyp- Sinapis alba L. with that of B. napus to the seed weevil for oviposi- mus Optical CO, LTD, Japan) with a 12.5 objective using the key of tion as well as to its parasitoids. B. rapa has previously been � Trjapitzin (1978), Gibson et al. (2006a), Baur et al. (2007) and Mul- reported to be a more attractive plant than oilseed rape to M. aen- ler et al. (2007). Pods were examined externally and any exit-holes eus, C. obstrictus (Buechi, 1990; Cook et al., 2002; Hokkanen et al., of larvae and of parasitoids were counted. All pods were then dis- 1986) and Ceutorhynchus pallidactylus (Barari et al., 2005) hence it sected and any non-exited weevil larvae or parasitoid pupae were has potential for use as a trap crop in central Europe and the UK. noted. Finally, the mean number of emerged parasitoids, C. obstric- However, its possible attractiveness to C. obstrictus under our more tus-infested pods and the percentage of infested weevils were cal- severe climatic conditions still remains unknown. B. juncea (L.) culated per plant. The parasitism level was calculated for each plot Czernajew is an annual species with yellow flowers, cultivated as as follows:

n n parasitoids parasitoids 1 1 % Parasitism n 100 100 ¼ Xavailable hosts � ¼ healthy dead X parasitized alone � 1 larvae þ larvae þ larvae þ parasitoid þ holes X X X X X X

n n an oil crop, leaf vegetable and for preparation of mustard powder Where 1parasitoids and 1available hosts are respectively: worldwide (Afonin et al., 2008; Downey, 2003). It is an annual total numbers of parasitoids (alive or dead, including pupae of P P weed in North America (Frankton and Mulligan, 1970; Mulligan parasitoids) and hosts attainable for parasitoids per plot;

and Bailey, 1975). S. alba is an annual plant with yellow ﬂowers, larvaehealthy and holes are numbers of healthy host larvae grown for its seed used as animal feed or as a catch crop. It is con- and exit holes on pods; dead and parasitized are P P larvae larvae sidered to be an alien species in Northern Europe and is featured in the numbers of dead larvae with traces of host feeding and of P P the Estonian alien species database (2012), the Lithuanian invasive parasitized larvae, respectively, and parasitoidalone is the number species database (2012) and the NOBANIS (2012) and DAISIE of parasitoids that emerged from the host. P (2012) databases.

58 126 G. Kovács et al. / Biological Control 65 (2013) 124–129

Table 1 v2 = 22.31, df = 3, P < 0.0001; 2008: v2 = 12020.7, df = 2, The sampling dates of pods for estimation of oviposition activity of Ceutorhynchus P < 0.0001; and summing the two years: v2 = 42.91, df = 3, obstrictus and its parasitoids in 2007 and 2008, in Tartu County, Estonia. P < 0.0001; Table 2). In 2007, the greatest number of damaged pods 2007 2008 was recorded on B. napus (2 pods per plant), significantly greater Brassica napus 8.08 14.08 than on B. juncea (P = 0.043), B. rapa (P = 0.0009) and S. alba Brassica rapa 2.08 31.07 (P = 0.0004). In 2008, the infestation rate stayed below 1% on each Brassica juncea 2.08 31.07 plant species; B. napus, B. rapa and B. juncea had similar mean Sinapis alba 2.08 24.07 numbers of damaged pods, S. alba had no damaged pods at all. Over the two years of the study, the number of damaged pods was great- Table 2 est on B. napus, 1.40 (14%) damaged pods per plant, which signifi- Mean (±SE) number of damaged pods by Ceutorhynchus obstrictus per plant on cantly differed from B. juncea (P = 0.002) where the mean number different host plant species in 2007, 2008 and over these two years in Tartu County, of damaged pods were 0.80, B. rapa (P = 0.0005) with 0.73 damaged Estonia. Different letters indicate significant differences between plant species (GENMOD, Differences of Least Squares Means test, SAS). pods and S. alba (P < 0.0001) where the number of damaged pods were 0.05 (Fig. 1). The lowest number of damaged pods were found 2007 2008 2007–2008 on S. alba, 0.5% of pods were damaged per plant, significantly less Mean ± SE Mean ± SE Mean ± SE than that on B. napus (P < 0.0001), B. rapa (P < 0.0001), and B. juncea Brassica napus 2.00 ± 0.28 a 0.80 ± 0.10 a 1.40 ± 0.17 a (P < 0.0001). Brassica rapa 0.53 ± 0.17 bc 0.93 ± 0.08 a 0.73 ± 0.13 b Brassica juncea 1.07 ± 0.36 b 0.53 ± 0.02 a 0.80 ± 0.20 b Sinapis alba 0.10 ± 0.07 c 0.00 ± 0.00⁄ 0.05 ± 0.04 c 3.2. Parasitism level

⁄The analysis gave no data to compare the mean of B. rapa and S. alba. The parasitism rate of C. obstrictus was also influenced by plant species over the two years of study ( 2 = 22.38, df = 3, P < 0.0001), All pod samplings were carried out at the same growth stage of v however the years had no significant effect (P = 0.19). The highest each plant species (Table 1). rate over these two years was 85.4% recorded on B. juncea, followed 2.3. Statistical analysis by 70.5% on B. rapa and 33.3% S. alba. However, concerning S. alba, only a single parasitoid constituted the percentage of parasitism. The effect of plant species on the mean number of C. obstrictus The lowest parasitism rate was 29.8% on B. napus where signifi- larvae was calculated by using Wald statistic Type III empirical cantly fewer C. obstrictus larvae were parasitized than on B. juncea standard error analysis with the Poisson distribution and the log (P < 0.0001) and on B. rapa (P < 0.0039). The parasitism was low on link function. Differences of means between the tested plant spe- S. alba, but due to large variation between replicates, no significant cies were made using the GENMOD procedure Differences of Least differences were found compared to other plant species (B. napus: Squares Means test. Comparisons of the number of C. obstrictus P = 0.82; B. juncea: P < 0.16; B. rapa: P < 0.47). parasitized larvae were made using the same analysis, but with Binomial distribution and logit link function and the response var- 3.3. Species composition of parasitoids iable was damaged pod. The scale parameter was estimated by Pearson Chi-Square divided by the degrees of freedom to account In total, 79 specimens of C. obstrictus parasitoids were reared for the model over dispersion. Statistical analyses were carried from pods over the two years, all belonging to the superfamily out using GLM and GENMOD procedure in SAS 8.02 (SAS Institute, Chalcidoidea. The greatest number of parasitoids hatched from Inc., Cary, NC, USA). pods of B. rapa with 28 specimens, followed by B. juncea with 27 and B. napus with 23 specimens, only 1 specimen was found from 3. Results S. alba (Fig. 2). The dominant species was M. morys (59.5% of all reared C. obstrictus parasitoids), followed by T. perfectus (25.3%) 3.1. Plant attractiveness to C. obstrictus (family Pteromalidae); 15.2% of C. obstrictus parasitoids belonged to the family Mymaridae. Damage caused by C. obstrictus was influenced by year The predominant species found from B. rapa and B. juncea was (v2 = 10.13, df = 1, P < 0.0015) and by plant species (in 2007: M. morys, 21 and 19 specimens, respectively, 7 specimens were

Fig. 1. Mean (±SE) number of pods damaged by Ceutorhynchus obstrictus per plant and its parasitism rate on different host plant species over the two years of the study period (2007 and 2008) in Tartu County, Estonia. Asterisk indicates a possibly misleading data (1 parasitoid per 3 larvae). Different letters indicate signiﬁcant differences between plant species (capital letters: parasitism rate, small letters: damaged pods; GENMOD, Differences of Least Squares Means test, SAS).

59 G. Kovács et al. / Biological Control 65 (2013) 124–129 127

Fig. 2. Total numbers of Ceutorhynchus obstrictus larval endoparasitoids and their species composition on different host plant species over the two years of the study period (2007 and 2008) in Tartu County, Estonia.

found from B. napus, and none from S. alba. The percentage parasit- by Brown et al. (1999), who found that both C. obstrictus adults ized by M. morys on different plant species was: 75.0% on B. rapa, and larvae were less abundant on S. alba plants than on B. napus, 70.4% on B. juncea and 30.4% on B. napus. B. rapa and B. juncea. Several studies have shown S. alba to be resis- The second most prevalent species was T. perfectus, which was tant to feeding damage (Bodnaryk and Lamb, 1991; Brown et al., the most common parasitoid species on B. napus with 14 speci- 1999; Lamb, 1980) suspecting the reason to be the trichomes of mens, equally represented with 3 specimens on B. rapa and B. jun- the pods. The preference of C. obstrictus for B. napus suggests that cea, and by none on S. alba. On B. napus, it attained 60.9% of the the proximity of other studied plant species would not increase total parasitism of C. obstrictus, on B. rapa 10.7% and on B. juncea the damage caused by this pest to the crop. 11.1%. The parasitism rates of C. obstrictus commonly exceed 50% In addition to the most common parasitoid species, parasitoids (Alford et al., 1996; Ulber and Vidal, 1998; Ulber et al., 2010; Vero- from the family Mymaridae also occurred: 5 specimens were mann et al., 2011, 2013). Therefore, the parasitoids of C. obstrictus reared from pods of B. juncea, 4 from B. rapa, 2 from B. napus. On can be highly efficient and can control their host’s population size S. alba a single specimen was found, the only parasitoid species since the effective pest control by parasitoids requires at least 32% associated with C. obstrictus. The proportions of parasitism by of total parasitism (Hawkins and Cornell, 1994), and the only plant Mymaridae were 18.5% on B. juncea, 14.3% on B. rapa and 8.7% on where it stayed under this threshold level was on B. napus with B. napus. All reared Mymaridae belong to the genus Anaphes: Ana- only 29.8% of larvae parasitized. Two of the C. obstrictus parasitoid phes fuscipennis Haliday (4 specimens), Anaphes tarsalis Mathof (4 species M. morys and T. perfectus reared in this study are the most specimens) and Anaphes arenbergi Debauche (2 specimens), all common and widespread throughout Europe (Williams, 2003), of these were detected for the first time in Estonia. In this experi- which T. perfectus has been considered widely distributed and par- ment, A. tarsalis and A. arenbergi were associated with C. obstrictus ticularly important (Williams, 2003). In total, the most abundant for the first time. parasitoid species was M. morys – comprising 59.5% from all parasitoids. The majority of M. morys specimens were found on B. rapa and B. juncea and only a few specimens were found on B. napus 4. Discussion where the dominant species was T. perfectus. These results are con- trary to studies in the UK where M. morys was an important para- Over the two years, in comparison to B. napus, all other tested sitoid species attacking C. obstrictus larvae on spring oilseed rape plant species were less attractive for oviposition to C. obstrictus (Murchie, 1996) while T. perfectus is generally the more important and therefore their presence in or near oilseed rape crops should on winter oilseed rape (Buntin, 1998; Kevväi et al., 2006; Vero- not increase the population size of this pest. However, the parasit- mann et al., 2006a,b,c, 2011). Greater numbers and parasitism ism rate was significantly greater on B. juncea and on B. rapa, which rates of M. morys on B. rapa and B. juncea than on B. napus indicate suggests that they could enhance parasitoid abundance. that their host seeking efficiency was higher on the alternative host In our study, the damage caused by C. obstrictus stayed well be- plants. Growing of B. rapa and B. juncea near B. napus can contrib- low 26%, which is considered to be the infestation rate above ute to help to build up the populations of parasitoid species by which yield loss will occur (Buntin, 1999; Free and Williams, lowering interspecific competition and therefore enhancing the 1978; Lerin, 1984). The greatest number of damaged pods was capacity of natural control. On crops of oilseed rape T. perfectus is found on B. napus (14% of pods damaged per plant), followed by usually the dominant species and may be able to outcompete other the similarly infested B. rapa and B. juncea, where damage rates parasitoid species. This might be supported by the fact that over were about half of that on B. napus. The least suitable host plant the years, parasitism by M. morys declined and concurringly para- for egg laying was S. alba, with less than 1% of pods damaged, sig- sitism by T. perfectus increased on winter oilseed rape in Estonia nificantly less than on all other plant species. Adults of C. obstrictus (Veromann et al., 2011). The diversification of potential host plant clearly preferred to oviposit on B. napus and not to oviposit on S. species of pests in agricultural land may offer more choices for dif- alba, despite the fact that on S. alba pods of a suitable size were ferent parasitoid species and can more efficiently utilise functional available for longer than on the other plant species. Adults of C. agro–biodiversity services. obstrictus oviposit into 20–40 mm pods (Free and Williams, Interestingly, the third most widely distributed parasitoid spe- 1978); it is likely that the abundance of suitable pods were coun- cies attacking C. obstrictus, S. gracilis, was not found in this study teracted by other factors. Supporting results have been reported although previous studies by Veromann et al. (2006a,b, 2011)

60 128 G. Kovács et al. / Biological Control 65 (2013) 124–129 and Kevväi et al. (2006) have shown it to have greater importance ported by the grant 8895 of the Estonian Science Foundation, and in Estonia than in other European countries and it also appeared to Estonian Ministry of Education and Research targeted financing be more viable compared to the other two common species (Vero- project SF 0170057s09 and the capacity–building project mann et al., 2011). When the most important European parasitoid P9003PKPK. species M. morys, T. perfectus and S. gracilis were introduced to Can- ada in the 1940s (Gibson et al., 2006b; McLeod, 1962) only S. gracilis established permanently (Gillespie et al., 2006) which also References indicates the greater flexibility and adaptation of this species. Fur- thermore, S. gracilis has a broader host range than other commonly Afonin, A.N., Greene, S.L., Dzyubenko, N.I., Frolov, A.N. (Eds.), 2008. Interactive occurring C. obstrictus parasitoids; several weevils and flies are agricultural ecological atlas of Russia and neighboring countries. Economic Plants and Their Diseases, Pests and Weeds, [Online] suitable for its development (Gibson et al., 2006b; Klander, 2001; (accessed 25.04.12). Noyes, 2002). The non–appearance of S. gracilis in our study might Alford, D.V., Walters, K.F.A., Williams, I.H., Murchie, A.K., 1996. A commercially be caused by unfavourable habitat conditions, low abundance of C. viable low-cost strategy for the management of oilseed weevil populations on winter oilseed rape in the UK. In: Proceedings of the Brighton Crop Protection obstrictus as a host and thereby higher competitiveness of other Conference – Pests and Diseases, vol. 2. pp. 609–614. parasitoid species; additionally they might prefer some other po- Awmack, C.S., Leather, S.R., 2002. Host plant quality and fecundity in herbivorous tential and available host species over C. obstrictus. insects. Annual Review Entomology 47, 817–844. Barari, H., Cook, S.M., Clark, S.J., Williams, I.H., 2005. Effect of a turnip rape (Brassica In addition to the key parasitoids, a surprisingly high proportion rapa) trap crop on stem-mining pests and their parasitoids in winter oilseed of C. obstrictus parasitoids reared from pods belonged to the family rape (Brassica napus). Biological Control 50, 69–86. Mymaridae. Although four species of this family are known to par- Baur, H., Muller, F.J., Gibson, G.A., Mason, P.G., Kuhlmann, U., 2007. A review of the species of Mesopolobus (Chalcidoidea: Pteromalidae) associated with asitize eggs of C. obstrictus, their importance in controlling the Ceutorhynchus (Coleoptera: Curculionidae) host-species of European origin. abundance of the pest has been considered negligible (Williams, Bulletin Entomological Research 97, 387–397. 2003). This study may reveal their greater relevance: more than Bodnaryk, R.P., Lamb, R.J., 1991. Influence of seed size in canola, Brassica napus L. and mustard, Sinapisalbis L., on seedling resistance against flea beetles, 15% of reared parasitoids in our study were Mymarids. Further- Phyllotreta cruciferae (Goeze). Canadian Journal Plant Science 71, 397–404. more, two species: A. tarsalis and A. arenbergi have never been re- Bottrell, D.G., Barbosa, P., Gould, F., 1998. Manipulating natural enemies by plant ported to attack C. obstrictus before. Still, the abundance of variety selection and modification: a realistic strategy? Annual Review Mymaridae was too low to draw any conclusion about their prefer- Entomology 43, 347–367. Brown, J., McCaffrey, J.P., Harmon, B.L., Davis, J.B., Erickson, D.A., Brown, A.P., ences over C. obstrictus’ food plants. Erickson, D.A., 1999. Effect of late season insect infestation on yield, yield To summarize, the hymenopteran parasitoids associated with C. components and oil quality of Brassica napus, B. rapa, B. juncea and Sinapis alba obstrictus were more abundant on B. juncea and B. rapa despite the in the Pacific Northwest region of the United States. Journal of Agricultural Science 132, 281–288. more concentrated host abundance on B. napus. This demonstrates Buechi, R., 1990. Investigations on the use of turnip rape as trap plant to control the parasitoids’ excellent host finding efficiency on the alternative oilseed rape pests. Bulletin IOBC/WPRS 13, pp. 32–pp. 39. host plant species. Even though B. juncea and B. rapa may attract Buntin, G.D., 1998. Cabbage seedpod weevil (Ceutorhynchus assimilis, Paykull) management by trap cropping and its effect on parasitism by Trichomalus the parasitoids away from B. napus to a certain extent, they could perfectus (Walker) in oilseed rape. Crop Protection 17, 299–305. contribute to strengthen the parasitoid populations which are sup- Buntin, G.D., 1999. Cabbage seedpod weevil control and damage loss assessment in pressed by dominating species on B. napus as the species composi- winter canola using insecticides. In: Proceedings of the 10th International Rapeseed Congress, Canberra, Australia [Online]. (accessed 17.07.12). Without additional study we cannot draw far-reaching conclusions Büchi, R., 1993. Monitoring of parasitoids of the cabbage seed weevil, Ceutorhynchus from the parasitism rate on S. alba as only one parasitoid reared assimilis during 1990 and 1991 in Switzerland. Bulletin IOBC/WPRS 16, 145– 149. during the two years of study. The unpopularity of this plant Cook, S.M., Khan, Z.R., Pickett, J.A., 2007a. The use of push–pull strategies in among both pests and parasitoids may be caused by several fac- integrated pest management. Annual Review Entomology 52, 375–400. tors; its buds and pods are hairy, which prevents oviposition by Cook, S.M., Rasmussen, H.B., Birkett, M.A., Murray, D.A., Pye, B.J., Watts, N.P., pests (Lamb, 1980) and is probably also a prohibitive factor for par- Williams, I.H., 2007b. Behavioural and chemical ecology underlying the success of turnip rape (Brassica rapa) trap crops in protecting oilseed rape (Brassica asitoids. Additionally, the low oviposition activity by pests also af- napus) from the pollen beetle (Meligethes aeneus). Arthropod–Plant Interact 1, fects the detection of suitable hosts by parasitoids and therefore 57–67. may decrease parasitism rate. Cook, S.M., Smart, L.E., Martin, J.L., Murray, D.A., Watts, N.P., Williams, I.H., 2006. Exploitation of host plant preferences in pest management strategies for oilseed rape (Brassica napus). Entomologia Experimentalis et Applications 119, 221– 229. 5. Conclusion Cook, S.M., Smart, L.E., Potting, R.J.P., Bartlet, E., Martin, J.L., Murray, D.A., Watts, N.P., Williams, I.H., 2002. Turnip rape (Brassica rapa) as a trap crop to protect B. rapa, B. juncea and S. alba did not attract more seed weevils to oilseed rape (Brassica napus) from infestation by insect pests: potential and mechanisms of action. Proceedings of the BCPC Conference: Pests and Diseases, the oilseed rape fields as they preferred to oviposit on B. napus. 2, 2002. British Crop Protection Council, Farnham, UK. These plant species have a great potential in supporting the pres- Corbet, S.A., Williams, I.H., Osborne, J.L., 1991. Bees and the pollination of crops and ence of beneficial insects. B. juncea and B. rapa has possible poten- wild flowers in the European community. Bee World 72, 47–59. DAISIE (Delivering alien invasive species inventories for Europe) database, 2012. tial for exploitation as an upholder of ecosystem services as the Sinapis alba [Online]. (accessed 25.04.12). B. rapa and B. juncea promoted the presence and abundance of De Moraes, C.M., Mescher, M.C., 1998. Plant–parasitoid interaction in tritrophic systems: a review of recent work with implications for biological control. the parasitoid M. morys which is usually suppressed by the domi- Journal of Entomological Science 34, 31–39. nant species, T. perfectus, on oilseed rape crops. For this reason, Dicke, M., 1999. Are herbivore-induced plant volatiles reliable indicators of these plant species might have potential as biological control herbivore identity to foraging carnivorous arthropods? Entomologia Experimentalis et Applications 91, 131–142. agents, diversifying the species composition of naturally occurring Dicke, M., Sabelis, M.W., Takabayashi, J., Bruin, J., Posthumus, M.A., 1990. Plant enemies of pests on oilseed rape crops. strategies of manipulating predator–prey interactions through allelochemicals: prospects for application in pest control. Journal of Chemical Ecology 16, 3091– 3118. Acknowledgments Dorn, S., Schumacher, P., Abivardi, C., Meyhöfer, R., 1999. Global and regional pest insects and their antagonists in orchards: spatial dynamics. Agriculture Ecosystems & Environment 73, 111–118. We wish to thank Kadri Pääsukene for her assistance with Downey, R.K., 2003. Mustard. In: Katz, S.H., Weaver, W.W. (Eds.), Encyclopedia of the statistical analysis used in this paper. This study was sup- Food and Culture. Gale Virtual Reference Library, New York, Scribner.

61 G. Kovács et al. / Biological Control 65 (2013) 124–129 129

Estonian alien species database, 2012. Sinapis alba [Online]. (In Estonian) (accessed 30.04.12). Christian–Albrechts–Universitätzu Kiel, Germany. Frankton, C., Mulligan, G.A., 1970. Weeds of Canada, Publication 948. Canada Nitzsche, O., 1998. Auftreten und Effizienz von Parasitoidenalsnatürliche Department of Agriculture, Ottawa. Gegenspieler von SchadinseketenimWinterrapsunterbesondererBerücksichtig Free, J.B., Williams, I.H., 1978. The responses of the pollen beetle, Meligethes aeneus, ungunterschiedlicherBodenbearbeitungsmabnahmen nachWinterraps[PhD and the seed weevil, Ceutorhynchus assimilis, to oilseed rape, Brassica napus and thesis].Georg–August–University of Göttingen, Germany. other plants. Journal of Applied Ecology 15, 761–774. NOBANIS (European network on invasive alien species) database, 2012. Sinapis alba, Gibson, G.A.P., Gates, M.W., Buntin, G.D., 2006a. Parasitoids (Hymenoptera: [Online] (accessed 26.04.12). Chalcidoidea) of the cabbage seedpod weevil (Coleoptera: Curculionidae) in Noyes, J.S., 2002. Interactive Catalogue of World Chalcidoidea, second ed. [CD– Georgia USA. Journal of Hymenoptera Research 15, 187–207. ROM]. Taxapad, Vancouver, British Columbia, The Natural History Museum, Gibson, G.A.P., Gillespie, D.R., Dosdall, L., 2006b. The species of Chalcidoidea London. (Hymenoptera) introduced to North America for biological control of the Price, P.W., Bouton, C.E., Gross, P., McPheron, B.A., Thompson, J.N., Weis, A.E., 1980. cabbage seedpod weevil, and the first recovery of Stenomalinagracilis Interactions among three trophic levels: influence of plants on interactions (Chalcidoidea: Pteromalidae). Canadian Entomological 138, 285–291. between insect herbivores and natural enemies. Annual Review of Ecology and Gillespie, D.R., Mason, P.G., Dosdall, L.M., Bouchard, P., Gibson, G.A.P., 2006. Systematics 11, 41–65. Importance of long-term research in classical biological control: an analytical Stiling, P., Moon, D.C., 2005. Quality or quantity: the direct and indirect effects of review of a release against the cabbage seedpod weevil in North America. host plants on herbivores and their natural enemies. Oecologia 142, 413–420. Journal of Applied Entomology 130, 401–409. Trjapitzin, V.A., 1978. Hymenoptera II. Chalcidoidea 18. Mymaridae, In: Medvedev, Hawkins, B.A., Cornell, H.V., 1994. Maximum parasitism rates and successful G.S. (Ed.), Keys to the Insects of the European Part of the USSR. Volume III: biological control. Science 266, 1886. Hymenoptera. Part II. Academy of Sciences of the USSR, Institute of Zoology, Hokkanen, H.H., Granlund, H., Husberg, G.B., Markkula, M., 1986. Trap crops used Nauka, Leningrad, USSR, p. 758 (in Russian). successfully to control Meligethes aeneus (Col., Nitidulidae), the rape blossom Turlings, T.C.J., Tumlinson, J.H., Heath, R.R., Proveaux, A.T., Doolittle, R.E., 1991. beetle. Annales Entomologici Fennici 52, 115–120. Isolation and identification of allelochemicals that attract the larval parasitoid, Kevväi, R., Veromann, E., Luik, A., Saarniit, M., 2006. Cabbage seed weevil Cotesia marginiventris (Cresson), to the microhabitat of one of its hosts. Journal (Ceutorhynchus assimilis Payk.) and its parasitoids in oilseed rape crops in of Chemical Ecology 17, 2235–2251. Estonia. Agronomy Research 4, 227–230. Ulber, B., Vidal, S., 1998. Influence of host density and host distribution on Klander, B., 2001. Die Rüsselkäfer der Unterfamilie Ceutorhynchinae und deren parasitism of Ceutorhynchus assimilis by Trichomalus perfectus. Bulletin IOBC/ Parasitoide auf Winterraps und begleiten Unkräutern in Schleswig–Holstein WPRS 21, 185–195. [M.Sc. thesis]. Christian–Albrechts–UniversitätzuKiel, Germany. Ulber, B., Klukowski, Z., Williams, I.H., 2010. Impact of insecticides on parasitoids of Lamb, R.J., 1980. Hairs protect pods of mustard (Brassica hirta ‘Gisilba’) from flea oilseed rape pests. In: Williams, I.H. (Ed.), Biocontrol-based Integrated beetle feeding damage. Canadian Journal Plant Science 60, 1439–1440. Management of Oilseed Rape. Springer, London, New York, pp. 337–355. Lancashire, P.D., Bleiholder, H., van den Boom, T., Langelüddeke, P., Strauss, R., van Alphen, J.J.M., Jervis, M.A., 1996. Foraging behaviour, In: Jervis, M., Kidd, N. Weber, E., Witzenberger, A., 1991. A uniform decimal code for the growth stages (Eds.), Insect Natural Enemies, vol. 4. Chapman Hall, London, p. 91. of crops and weeds. Annals of Applied Biology 119, 561–601. Veromann, E., Kevväi, R., Luik, A., Williams, I.H., 2008. Do cropping system and Lerin, J., 1984. Assessment of yield loss in winter rape due to seedpod weevil insecticide use in spring oilseed rape affect the abundance of pollen beetles (Ceutorhynchus assimilis Payk.) II. Yield loss in a cage experiment. Agronomie 4, (Meligethes aeneus Fab.) on the crop? International Journal of Pest Management 147–154. 54, 1–4. Lithuanian invasive species database, 2012. Sinapis alba [Online]. (accessed 26.04.12). Estonia. Bull. IOBC/wprs. International Workshop of Integrated control on McCall, P.J., Turlings, T.C.J., Lewis, W.J., Tumlinson, J.H., 1993. Role of plant volatiles oilseed crops, 11–12 October, 2005. Poznan, Poland, pp. 165–172. in host location by the specialist parasitoid Microplitis croceipes Cresson Veromann, E., Luik, A., Metspalu, L., Williams, I., 2006b. Key pests and their (Braconidae, Hymenoptera). Journal of Insect Behavior 6, 625–639. parasitoids on spring and winter oilseed rape in Estonia. Entomologic Fennici McLeod, J.H., 1962. Cabbage seedpod weevil – Ceutorhynchus assimilis (Payk.) 17, 400–404. Curculionidae. In: McLeod, J.H., McGugan, B.M., Coppel, H.C. (Eds.), A Review of Veromann, E., Metspalu, L., Williams, I.H., Hiiesaar, K., Mand, M., Kaasik, R., Kovacs, the Biological Control Attempts Against Insects and Weeds in Canada. Common G., Jogar, K., Svilponis, E., Kivimagi, I., Ploomi, A., Luik, A., 2012. Relative Wealth Agricultural Bureau (CAB), Farnham Royal, Bucks, UK. attractiveness of Brassica napus, Brassica nigra, Eruca sativa and Raphanus sativus Miranda, J.E., Navickiene, H.M.D., Nogueira–Couto, R.H., De Bortoli, S.A., Kato, M.J., for pollen beetle (Meligethes aeneus) and their potential for use in trap cropping. Bolzani, V. da S., Furlan, M., 2003. Susceptibility of Apis mellifera (Hymenoptera: Arthropod–Plant Interact 6, 385–394. Apidae) to pellitorine, an amide isolated from Piper tuberculatum (Piperaceae). Veromann, E., Tarang, T., Kevväi, R., Luik, A., 2006c. Insect pests and their natural Apidologie 34, 409–415. enemies on spring oilseed rape in Estonia: impact of cropping systems. Muller, F.J., Baur, H., Gibson, G.A.P., Mason, P.G., Kuhlmann, U., 2007. Review of the Agricultural and Food Science 15, 61–72. species of Trichomalus (Chalcidoidea: Pteromalidae) associated with Veromann, E., Toome, M., Kännaste, A., Kaasik, R., Copolovici, L., Flink, J., Kovács, G., Ceutorhynchus (Coleoptera: Curculionidae) host species of European origin. Narits, L., Luik, A., Niinemets, Ü., 2013. Effects of nitrogen fertilization on insect Canadian Entomologist 139, 643–657. pests, their parasitoids, plant diseases and volatile organic compounds in Mulligan, G.A., Bailey, L.G., 1975. The biology of Canadian weeds. 8. Sinapis arvensis Brassica napus. Crop Protection 43, 79–88. L.. Canadian Journal Plant Science 55, 171–183. Veromann, E., Williams, I.H., Kaasik, R., Luik, A., 2011. Potential of parasitoids to Murchie, A.K., 1996. Parasitoids of Cabbage Seed Weevil and Brassica Pod Midge in control populations of the weevil Ceutorhynchus obstrictus (Marsham) on winter Oilseed Rape [Ph.D. thesis]. University of Keele, UK. oilseed rape. International Journal of Pest Management 57, 85–92. Murchie, A.K., Williams, I.H., 1998. A bibliography of the parasitoids of the cabbage Williams, I.H., 2003. Parasitoids of cabbage seed weevil. In: Alford, D.V. (Ed.), seed weevil (Ceutorhynchus assimilis Paykull). Bulletin IOBC/WPRS 21, 163–169. Biocontrol of Oilseed Rape Pests. Blackwell Publishing, Oxford UK, pp. 97–112. Nilsson, C., 1994. Pollen beetles (Meligethes spp.) in oil seed rape crops (Brassica Zehnder, G., Gurr, G.M., Kühne, S., Wade, M.R., Wratten, S.D., Wyss, E., 2007. napus L.): biological interactions and crop losses. Department of Plant Arthropod pest management in organic crops. Annual Review Entomology 52, Protection Sciences, SLU. Dissertations 1, p. 39. 57–80.

62 II Kovács, G.; Kaasik, R.; Kaart, T.; Metspalu, L.; Luik, A.; Veromann E. 2017. In search of secondary plants to enhance the efficiency of cabbage seed weevil management. BioControl, 62: 29–38

64 BioControl (2017) 62:29–38 DOI 10.1007/s10526-016-9765-9

In search of secondary plants to enhance the efﬁciency of cabbage seed weevil management

Gabriella Kova´cs . Riina Kaasik . Tanel Kaart . Luule Metspalu . Anne Luik . Eve Veromann

Received: 25 April 2016 / Accepted: 26 September 2016 / Published online: 1 October 2016 � International Organization for Biological Control (IOBC) 2016

Abstract The infestation rate and parasitoid commu- These findings are important for planning new sustain- nities of Ceutorhynchus obstrictus (Marsham) (Coleop- able pest management approaches. tera: Curculionidae) were assessed on seven spring sown brassicaceous plant species to find potential Keywords Ceutorhynchus obstrictus Mesopolobus secondary plants that might help increase the parasitism morys Trichomalus perfectus Hymenopteran� � � rates of this serious oilseed pest. Over the three-year parasitoids Host plant preference Tritrophic � � study, the average infestation rate of pods by C. obstric- interactions tus remained below 10 % for each plant species. Despite the low pest abundance, C. obstrictus was parasitized by hymenopteran parasitoids on all plant species, except on Introduction Eruca vesicaria subsp. sativa ((Mill.) Thell.). Parasitism rates were remarkably high: between 33.7 and 70.8 % The cabbage seed weevil Ceutorhynchus obstrictus on average and peaked at 94.7 % on Raphanus sativus (Marsham) (Coleoptera: Curculionidae) is a wide- (L.) var. oleiformis (Pers.). Not only was the parasitism spread and significant pest of cultivated brassicaceous rate high on R. sativus, but it also had a different seed crops in Europe (Alford et al. 2003) and in North parasitoid species composition consisting mainly of egg America (Kalischuk and Dosdall 2004; Ulmer and parasitoids (Mymaridae), while on the other plant Dosdall 2006). Following their overwintering period species larval parasitoids (Pteromalidae) dominated. adult weevils feed mainly on buds, flowers, and developing pods of Brassicaceae species before females start to oviposit into pods that are more than 2 mm in diameter (Doucette 1947; Dmoch 1965) or Handling Editor: Stefano Colazza approximately 40–60 mm long (Dosdall and Moisey G. Kovaćs (&) R. Kaasik L. Metspalu 2004). Usually only one egg per pod is laid through a A. Luik E. Veromann� � � hole pierced by the mouthparts (Doucette 1947) and � Institute of Agricultural and Environmental Sciences, after that the pod is marked with an oviposition- Estonian University of Life Sciences, Fr. R. Kreutzwaldi 5, 51014 Tartu, Estonia deterring pheromone to avoid further oviposition into e-mail: [email protected] the same pod (Kozlowski et al. 1983). Adults puncture flower buds and pods while feeding and ovipositing T. Kaart but the main damage to the plant is caused by the larva Institute of Veterinary Medicine and Animal Sciences, Estonian University of Life Sciences, Fr. R. Kreutzwaldi feeding on developing seeds inside the pod. Mature 5, 51014 Tartu, Estonia larvae exit the pod, drop to the ground and pupate in

123 65 30 G. Kovaćs et al. the soil. Severe infestations lead to reduced yields: their populations and keep them in the fields (Gurr and each larvae consumes about five to six seeds in a pod, Wratten 1999; Landis et al. 2000; Parolin et al. 2012). which reduces the yield per pod by approximately In Estonia, certain oilseed brassicaceous plants have 18 % (Free and Williams 1978). As a secondary the potential to increase the biological control of an damage, puncture wounds on the pods made by C. important pest of oilseed rape, the pollen beetle obstrictus adults might make plants more accessible Meligethes aeneus (Fabricius) (Coleoptera: Nitiduli- for the brassica pod midge Dasineura brassicae dae). For example, according to Kaasik et al. (2014b) (Winnertz) (Diptera: Cecidomyiidae) to infest and and Veromann et al. (2012), Brassica nigra (L.) W. damage oilseed rape plants via feeding on the pod wall D. J. Koch and Sinapis alba (L.) were more attractive (Alford et al. 2003). In addition, the exit holes of to adult pollen beetles than B. napus (L.). Raphanus C. obstrictus larvae can also provide good pre- sativus (L.) var. oleiformis (Pers.) was equally attrac- condition for infestation by fungal pathogens (Dosdall tive as B. napus for oviposition to pollen beetles, but et al. 2001). their larvae could not complete their development in In Europe, one adult weevil per plant can cause 4 % the pods of R. sativus (Veromann et al. 2014). yield reduction on winter oilseed rape (Williams B. juncea (L.) Czern. and B. nigra supported the 2010). Oilseed rape can compensate for seed loss parasitoids of pollen beetles (Kaasik et al. 2014a, b). linked to C. obstrictus infestation of up to 26 % Hence, introducing secondary plant species to the infested pods without significant yield loss (Buntin cropping system might be a good alternative method 1999). In Estonia, this economic threshold for C. ob- for cabbage seed weevil management as well. To be strictus in winter oilseed rape was surpassed for the able to test this hypothesis we aimed to find suit- first time in 2007 (Veromann et al. 2011). Despite able secondary plant species through determining occurring in both spring and winter varieties, C. ob- whether the cabbage seed weevil and its parasitoids strictus seems to cause more damages in winter have any host plant preferences. Levels of infestation oilseed rape in Estonia, due likely to earlier feeding by C. obstrictus and the percent of parasitism of and oviposition opportunity. C. obstrictus on different brassicaceous species had When it comes to plant protection, the only time to been previously investigated in Estonia during a small effectively spray against C. obstrictus is during their scale pilot study (Kovaćs et al. 2013). However, the full invasion at early to mid-flowering stage (Williams species composition of parasitoids on different plant 2010). Unfortunately, the chemical insect control can species has not been quantitatively analysed. The also be detrimental to beneficial insects, especially to current study was designed to examine (1) the damage pollinators and parasitoids, and therefore it is more made by C. obstrictus (2) the parasitism rate and (3) and more limited in use. Parasitoids are valuable the parasitoid communities of C. obstrictus on seven natural resources for pest control since they can reduce different brassicaceous plant species. the number of individuals of the next generation of pests (Godfray 1994). In order to preserve parasitoids, crop management practices should aim to enhance the Materials and methods resources that support their survival. Hymenopteran parasitoids are able to attack C. obstrictus at all stages Study area and set-up of its life cycle (Murchie and Williams 1998). The parasitism rates reported over the years were mostly The study was carried out in an experimental field of substantial in central and northern Europe, often rising the Estonian University of Life Sciences in 2010, 2011 above 30 % (Bu¨chi 1993; Williams 2003) or even and 2012 in Tartu County, Estonia (58�21059.0300N, higher in Estonia, with a report of 90 % (Veromann 26�39�54.0500E). The plants were grown in a random- et al. 2011, 2013). A key factor in maintaining ized complete block design with the plot size of successful pest suppression by parasitoids is to provide 1 m 9 5 m, similar to the plot size described by a favourable environment for their adults. Secondary Kalischuk and Dosdall (2004). Each plant species was plants grown together with the primary crop can replicated three times: oilseed rape Brassica napus provide additional food, alternative and/or more (L.) (spring variety, ‘Mascot’), mustard greens suitable hosts for parasitoids, which can help to grow B. juncea (L.) Czern. (‘Jadrjonaja’), black mustard

123 66 In search of secondary plants to enhance the efﬁciency of cabbage seed weevil management 31

B. nigra (L.) W.D.J. Koch, turnip rape B. rapa (L.) Statistical analysis (’Largo’; not cultivated in 2010), rucola/arugula Eruca vesicaria subsp. sativa ((Mill.) Thell.) The rate of pods infested with C. obstrictus was (=E. sativa, ‘Poker’), fodder radish Raphanus sativus calculated for each plant by dividing the total number (L.) var. oleiformis (Pers.) (=R. sativus, ‘Bille’, pale of infested pods by ten (number of pods collected per violet flowers) and white mustard Sinapis alba (L.) plant). The parasitism rate was calculated for each (‘Branco’). In order to minimize inter-plot interac- infested plant by dividing the total number of tions, there was a one meter wide buffer zone of bare parasitoids by the total number of infested pods (see soil around each plot. Plots were sown on 12 May in the detailed formula in Kovaćs et al. 2013). The effects 2010, on 9 May in 2011 and on 15 May in 2012, with of plant species, year and plant species 9 year the density of 250 seeds per m2. The seeds were interaction on infestation and parasitism rate were obtained from the seed collection of the Estonian tested with a generalized linear mixed model with University of Life Sciences. Crop management was logit link function considering also the random effect uniform in all plots, no fertilizers or pesticides were of replicate, and followed by pairwise comparison of used. The growth stage of plants (BBCH) was assessed least square means of the studied factors. The twice a week every year using the decimal code system relationship between parasitoid species composition of Lancashire et al. (1991). and plant species was studied using the two-way frequency table and Fisher exact test. To discover the common patterns in parasitoid species composition Insect sampling and their differences between years and plant species, the principal component analysis was performed. In order to estimate the weevil infestation rates and Since there were no parasitoids found on E. sativa parasitism levels, ten pods were collected per plant over the three years, this plant species was excluded (five pods from the main and five pods from the third from the analysis considering the parasitism rate and raceme from the top) from ten randomly chosen plants parasitoid species composition. Due to the low in each plot at the beginning of the ripening stage abundance of some parasitoid species, only two (BBCH 80–81). As described in detail by Dosdall and separate species—M. morys and T. perfectus—were Kott (2006), Dosdall et al. (2006), Veromann et al. used in the parasitoid species composition analyses. (2011), Kovaćs et al. (2013), pods were placed in The remaining species were grouped as Mymaridae, cardboard emergence boxes and kept at room temper- other Pteromalidae, and other families. ature for 30 days. Emerged larvae and parasitoids The statistical analyses were conducted with SAS were counted, and all pods were carefully examined 9.4 software (SAS Institute Inc., Cary, NC, USA), for and dissected to detect parasitoid or weevil remains modelling and least square means comparison the and exit holes. Parasitoid adults were preserved in GLIMMIX procedure was used. All results were Eppendorf tubes containing 70 % ethanol, at -20 �C considered statistically significant at P B 0.05. and identified—within two weeks after the dissection of pods—using an Olympus SZX-ZB9 stereo-microscope (Olympus Optical Co, Ltd, Japan) with a Results 2289 maximum magnification, coupled to an Olym- pus Highlight 3100 light source (Olympus Optical Co, Pods infested with C. obstrictus Europa, GMBH, Hamburg) using the keys of Graham (1969), Trjapitzin (1978), Gibson et al. (2006), Baur Overall, the rate of C. obstrictus-infested pods was et al. (2007), Muller et al. (2007). In case of low. The effect of the host plant on the proportion of Mesopolobus morys (Walker) (Hymenoptera: Ptero- infested pods per plant species was statistically malidae), Trichomalus perfectus (Walker) (Hy- significant throughout the study period (F6,51 = 8.58, menoptera: Pteromalidae) and Stenomalina gracilis P \ 0.001). Infested pods were most frequently found (Walker) (Hymenoptera: Pteromalidae) voucher spec- on B. napus, where 8.8 % of pods were infested on imens—from Rothamsted Research, UK—were used average, followed by R. sativus, B. juncea and B. rapa as an addition to the keys to confirm the identification. (6.0, 5.5 and 5.5 %, respectively; Fig. 1). When

123 67 32 G. Kova´cs et al.

Fig. 1 Proportions of Ceutorhynchus obstrictus-infested pods studied. Different letters indicate statistically significant differ- by plant species and years. Numbers inside/under the bars ences between plant species (P \ 0.05; pairwise comparison of denote the number of infested pods per 300 pods studied for each least square means according to the generalized linear mixed plant species-year combination. Horizontal lines and corre- model with logit link function considering fixed effects of sponding values denote the overall proportions (±SE) over the culture and year and non-zero covariance between observations years. The symbol ‘‘9’’ denotes that in 2010 B. rapa was not corresponding to the same replicate) compared with other species, the proportion of Parasitoid species composition infested pods was significantly lower on B. nigra (2.3 %), and the lowest on S. alba and E. sativa (0.6 All 148 parasitoid specimens reared over the study and 0.1 %, respectively). period were determined to belong to the Chalcidoidea The overall difference between the study years was superfamily. Four families were represented: Ptero- also significant (F2,51 = 6.76, P = 0.003), with the malidae with six species (95 specimens in total), highest infestation frequency of pods in 2012. Infes- Mymaridae with eleven species (47 specimens in tation rates within the same plant species differed total), Eulophidae with three species (four specimens among the study years (F11,40 = 4.48, P \ 0.001). in total) and Eurytomidae with two species (two This is most evident when comparing R. sativus specimens total; Table 1). Mesopolobus morys was infestation rates with other plant species: in 2010 and the most numerous parasitoid species followed by 2011 R. sativus had the greatest number of infested T. perfectus. pods, in 2012 the other plant species (except S. alba) The most diverse species composition was found were more infested (Fig. 1). from the pods of B. napus (11 different parasitoid species), followed by R. sativus (ten different parasitoid species). Brassica napus was the only plant Parasitism rate species from where all four parasitoid families were found. During the three years of study, R. sativus had a Plant species had no influence on the parasitism rate of significantly different species composition when com- C. obstrictus (2010–2012: F5,38 = 1.03, P = 0.41; pared to the other plant species (P \ 0.001, Fisher excluding E. sativa). The highest mean parasitism rate exact test) (Table 1). Ceutorhynchus obstrictus on was recorded on R. sativus (70.8 %) over the three R. sativus was mainly attacked by parasitoids belong- years. Sinapis alba had the lowest percentage of ing to the Mymaridae family (97.6 %), while in case of parasitized pests (33.7 %), while there were no the four Brassica species the majority of parasitoids parasitoids found on E. sativa. In 2010, the parasitism were pteromalids. Only two parasitoid species (Ptero- rates were particularly high on B. napus (88.2 %), malidae) were found from the pods of S. alba. B. juncea (88.9 %), B. nigra (80.0 %) and R. sativus The principal component analysis of the five groups (94.7 %) (Fig. 2). of parasitoid species—M. morys, T. perfectus, other

123 68 In search of secondary plants to enhance the efﬁciency of cabbage seed weevil management 33

Fig. 2 Parasitism rate by plant species and years. Numbers statistically significant difference between plant species inside the bars denote the number of parasitoids per number of (P = 0.41; generalized linear mixed model with logit link infested pods. Horizontal lines and corresponding values denote function considering fixed effects of culture and year and non- the overall proportions (±SE) over the years. The symbol ‘‘9’’ zero covariance between observations corresponding to the denotes that in 2010 B. rapa was not studied. There was no same replicate) Pteromalidae, Mymaridae, other families (Eulophi- B. juncea than on B. napus and B. rapa. However, in the dae, Eurytomidae)—revealed that M. morys and present study the preference towards B. napus over T. perfectus were more often found from the pods of B. juncea, B. rapa and R. sativus was not statistically the same plant species compared with other para- significant. To further investigate insect preference, sitoids, and the presence of Eulophidae and Euryto- dual-choice experiments should be performed. Our midae (other families) did not depend on the presence finding that S. alba had extremely low infestation or absence of Mymaridae and Pteromalidae (Fig. 3). levels over the study period is also supported by Pairing the parasitoid species patterns with host plants previous studies by Brown et al. (1999), Ulmer and and study years reveals that T. perfectus and M. morys Dosdall (2006), and Kovaćs et al. (2013), who found were mainly found in 2012, and mostly from B. rapa, S. alba to be unattractive to C. obstrictus and even B. napus and B. juncea, while mymarids were mainly considered the plant to be resistant to feeding damage related to R. sativus and were more numerous in 2010 (Bodnaryk and Lamb 1991; Brown et al. 1999; and 2011 (Fig. 3). Kalischuk and Dosdall 2004). These resistant traits of S. alba are used in developing cabbage seed weevil- resistant oilseed rape by introgression (Dosdall and Discussion Kott 2006; Tansey et al. 2010a). Although B. nigra has been shown to be very attractive to the pollen beetle During the three-year study, the infestation by C. ob- and accordingly a promising trap crop (Veromann et al. strictus was low on all studied plant species, remaining 2012; Kaasik et al. 2014a), this plant species was below the economic threshold (26 %). In this study, the significantly less attractive to C. obstrictus compared number of infested pods fluctuated greatly between with all other studied plant species, except S. alba. This plant species as well as years, with the highest number finding is in accordance with the study of Kalischuk of infested pods recorded on B. napus (24 %) in 2012. and Dosdall (2004) who also found S. alba and B. nigra Summing the three years, B. napus had the highest to be unattractive to C. obstrictus. Several plant proportion of infested pods, which is in accordance characteristics such as volatiles emitted from brassi- with earlier observations of Kovaćs et al. (2013). caceous plants’ flowers and green leaves, pod’s wall Similarly, Dosdall et al. (2006), Ulmer and Dosdall thickness, colour, size, etc., may play a role in host- (2006), and Ca´rcamo et al. (2007) found that C. ob- plant selection for oviposition. Previous studies have strictus females deposited significantly fewer eggs on shown that C. obstrictus does not distinguish host

123 69 34 G. Kova´cs et al.

Table 1 Total number of parasitoids reared from C. obstrictus collected from different brassicaceous plant species in Tartu County, Estonia, 2010, 2011, 2012 2010–2012 2010 2011 2012 Total B. napus B. juncea B. nigra B. rapa R. sativus S. alba

Pteromalidae—total 41 23 6 22 1 2 16 9 70 95 Mesopolobus gemellus Baur & 1 1 1 Muller Mesopolobus incultus Walker 1 1 1 Mesopolobus morys Walker 27 15 3 13 10 7 41 58 Stenomalina gracilis Walker 1 1 1 Trichomalus bracteatus Walker 1 1 1 1 2 Trichomalus perfectus Walker 12 7 3 8 1 1 5 27 32 Mymaridae—total 7 40 23 17 7 47 Anaphes arenbergi Debauche 1 5 1 5 6 Anaphes brachygaster Debauche 1 1 1 Anaphes crassicornis group 7 4 3 7 Anaphes cultripennis Debauche 11 7 4 11 Anaphes fuscipennis Haliday 1 3 3 1 4 Anaphes regulus Walker 7 6 1 7 Anaphes tarsalis Mathot 3 3 3 Other Anaphes species 2 3 5 5 Other Anaphini 2 2 2 Other Mymaridae 1 1 1 Eulophidae—total 2 1 1 4 4 Euderus albitarsis Zetterstedt 1 1 1 Tetrastichus rebezae Kosjukov 1 1 1 Tetrastichus semidesertus 2 2 2 Kosjukov Eurytomidae—total 1 1 2 2 Eurytoma curculionum Mayr 1 1 1 Other Eurytoma species 1 1 1 Total number of parasitoids 51 24 8 22 41 2 45 26 77 148 plants only by the profiles or levels of glucosinolates control by parasitoids (Hawkins and Cornell 1994). In (Ulmer and Dosdall 2006; Tansey et al. 2010b, c; Blake case of R. sativus, 70 % of the infested pods were et al. 2014), but probably the synergy of different parasitized on average, peaking at 94.74 % in 2010. volatiles together with other plant characteristics may High parasitism rates in this study are consistent with affect the weevils’ oviposition behaviour. It is also earlier European studies that found C. obstrictus to be possible that the population that lays eggs on the spring highly parasitized (Alford et al. 1996; Ulber and Vidal varieties of oilseed plants (i.e., population that realizes 1998; Ulber et al. 2010; Veromann et al. 2011, 2013; its oviposition potential later than the majority of Kovaćs et al. 2013). One unanticipated finding for the weevils that deposit eggs on winter oilseed rape) is current study was that no parasitoids were recorded on adapted to a wider range of host plants. B. rapa in 2011, even though hosts were available, and The overall parasitism levels in the present study despite the low overall parasitoid abundance that year, were noteworthy. Over the three years, the parasitism parasitoids were present on all other plant species, rate of the studied pest exceeded 32 % on several plant except on E. sativa. This result suggests a need for species, which is the suggested level for effective pest follow-up studies.

123 70 In search of secondary plants to enhance the efﬁciency of cabbage seed weevil management 35

Interestingly, in addition to the common larval ectoparasitoid species, a high proportion of egg parasitoids (Mymaridae) reared from the pods, pri- marily from hosts on R. sativus. The reasons for the egg and larval parasitoids to prefer hosts on different plant species might be associated with relying on different cues while locating their hosts. For orienta- tion, egg parasitoids are believed to be limited to rely on plant volatiles induced by oviposition (Hilker and Fig. 3 Principal component analysis of parasitoid species and Meiners 2002; Colazza et al. 2004a, b), which are families (arrows, parasitoid names and families in boldface). released in lower amounts than cues from larval White rhombus and black circles denote the average scores feeding (Kaiser et al. 1989; Hilker and McNeil 2008). corresponding to different host plant species and study years, respectively However, egg deposition occurred on all studied plant species, suggesting that the volatile composition and/ or amount emitted by R. sativus as an indirect defence response to oviposition might be slightly different The most numerous parasitoid species—M. morys from that of the other plant species. Parasitized larvae and T. perfectus—found in this and the authors’ consume fewer seeds than healthy larvae (Murchie previous study (Kovaćs et al. 2013) are two of the 1996). However parasitized eggs will not develop three most cited larval ectoparasitoids of C. obstrictus further. Therefore no seed consumption will occur. (Williams 2003). Mesopolobus morys and T. perfectus Hence, the presence of egg parasitoids in the fields were more likely found from pods of the same plant likely further reduce the damage made to the crop. In species, which can be explained by their similar addition to the potential of supporting egg parasitoids, biology. Both M. morys and T. perfectus are solitary in previous studies, R. sativus also showed potential to ectoparasitoids, attacking concealed larvae (Bu¨chi control the pollen beetle, by not supporting the 1993; Murchie and Williams 1998). To date, T. per- survival of pollen beetle eggs and larvae in its flowers fectus has been the predominant parasitoid species on (Veromann et al. 2012), and suggesting a potential to B. napus in Estonia and tended to outcompete other be a dead-end trap crop of M. aeneus (Veromann et al. parasitoid species, while M. morys has been more 2014). The potential of R. sativus to provide habitat for abundant on other plant species (Veromann et al. non-prevalent parasitoid species is also documented 2011; Kovaćs et al. 2013). In the current study, by Kaasik et al. (2014a). however, M. morys was dominating on most of the Previous Estonian studies (Veromann et al. 2012; studied plant species, including B. napus. This result Kaasik et al. 2014a, b; Veromann et al. 2014) have seems to be in line with those obtained by Murchie suggested that growing certain brassicaceous plant (1996), who found that M. morys may be relatively species in or nearby the oilseed rape field might be more important on spring than on winter oilseed rape. useful in controlling the abundance of pollen beetles. Another common pteromalid parasitoid of C. obstric- In the current study we found that these plant species tus in Europe, S. gracilis, was recorded only once significantly limited the growth of the next generation during our study period, on one of the least infested of seed weevils as well, due to their very high plant species, S. alba. The almost complete absence of parasitism rate. Additionally, R. sativus was found to S. gracilis in our study was surprising because be attractive for oviposition to C. obstrictus, and also Veromann et al. (2011) showed that this parasitoid supported egg parasitoids over larval parasitoids. species was relatively more important in Estonia than Therefore, the presence of R. sativus in or near other in other European countries, and even more abundant oilseed crops might introduce more egg parasitoids than M. morys, on winter oilseed rape. Thus, we can into the growing area and increase their so far speculate that the life cycle of S. gracilis can be more negligible importance in the biological control of synchronized with the larval development of C. ob- C. obstrictus. strictus on winter oilseed rape, which would explain Though, the results discussed above may suggest its low abundance in our study. useful implications in plant protection strategies

123 71 36 G. Kovaćs et al. which are based on multitrophic interactions, we are Ca´rcamo H, Olfert O, Dosdall LM, Herle C, Beres B, Soroka J convinced that the current study is only a preliminary (2007) Resistance to cabbage seedpod weevil among selected Brassicaceae germplasm. Can Entomol one. In order to provide more reliable results, we think 139:658–669 that field scale experiments are necessary both on Colazza S, Fucarino A, Peri E, Salerno G, Conti E, Bin F winter and spring oilseed crops. (2004a) Insect oviposition induces volatile emission in herbaceous plants that attracts egg parasitoids. J Exp Biol Acknowledgments We would like to thank Dr. Merje Toome- 207:47–53 Heller and the anonymous reviewers for their valuable and Colazza S, McElfresh JS, Millar JG (2004b) Identification of detailed comments on an earlier version of this manuscript. We volatile synomones, induced by Nezara viridula feeding would also like to thank the editors for their helpful comments and oviposition on bean spp., that attract the egg parasitoid and support during the review process. This study was supported Trissolcus basalis. J Chem Ecol 30:945–964 by the grant no. 8895 of the Estonian Science Foundation, the Dmoch J (1965) The dynamics of a population of the cabbage capacity-building Project no. P9003PKPK of the Estonian seedpod weevil (Ceutorhynchus assimilis Payk.) and the University of Life Sciences, the targeted financing Project no. development of winter rape. Part I. Ekol Pol Ser A SF0170057s09 of the Estonian Ministry of Education and 4:249–287 Research, the institutional research Funding no. IUT36-2 of the Dosdall LM, Kott LS (2006) Introgression of resistance to Estonian Research Council, and the European Social Fund’s cabbage seedpod weevil to canola from yellow mustard. Doctoral Studies and Internationalisation Programme DoRa Crop Sci 46:2437–2445 (which was carried out by Foundation Archimedes). Dosdall LM, Moisey DWA (2004) Developmental biology of the cabbage seedpod weevil, Ceutorhynchus obstrictus (Coleoptera: Curculionidae), in spring canola, Brassica napus, in Western Canada. Ann Entomol Soc Am 97:458–465 References Dosdall LM, Moisey D, Ca´rcamo HA, Dunn R (2001) Cabbage seedpod weevil.pdf. In: Agri-Facts, practical information Alford DV, Walters KFA, Williams IH, Murchie AK (1996) A for Alberta’s agriculture industry. Revised January 2014, commercially viable low-cost strategy for the management Agdex 622-21. http://www1.agric.gov.ab.ca/$department/ of oilseed weevil populations on winter oilseed rape in the deptdocs.nsf/all/agdex2538/$file/622-21.pdf. Accessed 18 UK. In: Proceedings of the Brighton Crop Protection Apr 2016 Conference—Pests and Diseases. Brighton, pp 609–614 Dosdall LM, Ulmer BJ, Gibson GAP, Ca´rcamo HA (2006) The Alford DV, Nilsson C, Ulber B (2003) Insect pests of oilseed spatio-temporal distribution dynamics of the cabbage rape crops. In: Alford DV (ed) Biocontrol of oilseed rape seedpod weevil, Ceutorhynchus obstrictus (Coleoptera: pests. Blackwell Science Ltd, Oxford, pp 9–42 Curculionidae), and its larval parasitoids in canola in Baur H, Muller FJ, Gibson GAP, Mason PG, Kuhlmann U Western Canada. Biocontrol Sci Technol 16:987–1006 (2007) A review of the species of Mesopolobus (Chalci- Doucette CF (1947) Host plants of the cabbage seedpod weevil. doidea: Pteromalidae) associated with Ceutorhynchus J Econ Entomol 40:838–840 (Coleoptera: Curculionidae) host-species of European Free JB, Williams IH (1978) A survey of the damage caused to origin. B Entomol Res 97:387–397 crops of oil-seed rape (Brassica napus L.) by insect pests in Blake AJ, Dosdall LM, Tansey JA (2014) Nutritional effects on south-central England and their effect on seed yield. the appearance of canola and its attractiveness to the cab- J Agric Sci 90:417–424 bage seedpod weevil. J Insect Behav 27:759–775 Gibson GA, Gates MW, Buntin GD (2006) Parasitoids (Hy- Bodnaryk RP, Lamb RJ (1991) Influence of seed size in canola, menoptera: Chalcidoidea) of the cabbage seedpod weevil Brassica napus L. and mustard, Sinapis alba L., on seed- (Coleoptera: Curculionidae) in Georgia, USA. J Hy- ling resistance against flea beetles, Phyllotreta cruciferae menoptera Res 15:187–207 (Goeze). Can J Plant Sci 71:397–404 Godfray HCJ (eds) (1994) Introduction. In: Parasitoids: Brown J, McCaffrey JP, Harmon BL, Davis JB, Brown AP, behavioral and evolutionary ecology. Princeton University Erickson DA (1999) Effect of late season insect infestation Press, Princeton, pp 3–25 on yield, yield components and oil quality of Brassica Graham MWR de V (1969) The Pteromalidae of North-Western napus, B. rapa, B. juncea and Sinapis alba in the Pacific Europe (Hymenoptera: Chalcidoidea). Bulletin of the Northwest region of the United States. J Agric Sci British Museum (natural history). Entomology 16:1–908 132:281–288 Gurr GM, Wratten SD (1999) FORUM ‘‘Integrated biological Bu¨chi R (1993) Monitoring of parasitoids of the cabbage seed control’’: A proposal for enhancing success in biological weevil, Ceutorhynchus assimilis during 1990 and 1991 in control. Int J Pest Manag 45:81–84 Switzerland. In: Paul VH, Williams IH, Gladders P (eds) Hawkins BA, Cornell HV (1994) Maximum parasitism rates and Proceedings of the Working group ‘‘Integrated control in successful biological control. Science 266:1886 oilseed crops’’. Le Rheu, pp 145–149 Hilker M, McNeil J (2008) Chemical and behavioral ecology in Buntin GD (1999) Damage loss assessment and control of the insect parasitoids: How to behave optimally in a complex cabbage seedpod weevil (Coleoptera: Curculionidae) in odorous environment. In: Wajnberg E´ , Bernstein C, van winter canola using insecticides. J Econ Entomol Alphen J (eds) Behavioral ecology of insect parasitoids. 92:220–227 Blackwell Publishing Ltd, Oxford, pp 92–112

123 72 In search of secondary plants to enhance the efﬁciency of cabbage seed weevil management 37

Hilker M, Meiners T (2002) Induction of plant responses to associated with novel genotypes developed by Sinapis alba oviposition and feeding by herbivorous arthropods: a L. 9 Brassica napus L. Arthropod-Plant Interact 4:95–106 comparison. Entomol Exp Appl 104:181–192 Tansey JA, Dosdall LM, Keddie BA, Noble SD (2010c) Con- Kaasik R, Kovaćs G, Kaart T, Metspalu L, Williams IH, tributions of visual cues to cabbage seedpod weevil, Ceu- Veromann E (2014a) Meligethes aeneus oviposition pref- torhynchus obstrictus (Marsham) (Coleoptera: erences, larval parasitism rate and species composition of Curculionidae), resistance in novel host genotypes. Crop parasitoids on Brassica nigra, Raphanus sativus and Eruca Prot 29:476–481 sativa compared with on Brassica napus. Biol Control Trjapitzin VA (1978) Hymenoptera III, 2. Chalcidoidea 18. 69:65–71 Mymaridae. In: Medvedev GS (ed) Keys to the insects of Kaasik R, Kovaćs G, Toome M, Metspalu L, Veromann E the European part of the USSR. Academy of Sciences of (2014b) The relative attractiveness of Brassica napus, the USSR, Institute of Zoology, Nauka, Leningrad, B. rapa, B. juncea and Sinapis alba to pollen beetles. pp 516–537 BioControl 59:19–28 Ulber B, Vidal S (1998) Influence of host density and host Kaiser L, Pham-Delegue MH, Bakchine E, Masson C (1989) distribution on parasitism of Ceutorhynchus assimilis by Olfactory responses of Trichogramma maidis Pint. et Trichomalus perfectus. Bulletin OILB/SROP Voeg.: effects of chemical cues and behavioral plasticity. 21(5):185–195 J Insect Behav 2:701–712 Ulber B, Klukowski Z, Williams IH (2010) Impact of insecti- Kalischuk AR, Dosdall LM (2004) Susceptibilities of seven cides on parasitoids of oilseed rape pests. In: Williams IH Brassicaceae species to infestation by the cabbage seedpod (ed) Biocontrol-based integrated management of oilseed weevil (Coleoptera: Curculionidae). Can Entomol rape pests. Springer, Netherlands, pp 337–355 136:265–276 Ulmer BJ, Dosdall LM (2006) Glucosinolate profile and Kovaćs G, Kaasik R, Metspalu L, Williams IH, Luik A, Vero- oviposition behavior in relation to the susceptibilities of mann E (2013) Could Brassica rapa, Brassica juncea and Brassicaceae to the cabbage seedpod weevil. Entomol Exp Sinapis alba facilitate the control of the cabbage seed Appl 121:203–213 weevil in oilseed rape crops? Biol Control 65:124–129 Veromann E, Williams IH, Kaasik R, Luik A (2011) Potential of Kozlowski MW, Lux S, Dmoch J (1983) Oviposition behaviour parasitoids to control populations of the weevil Ceu- and pod marking in the cabbage seed weevil, Ceu- torhynchus obstrictus (Marsham) on winter oilseed rape. torhynchus assimilis. Entomol Exp Appl 34:277–282 Int J Pest Manag 57:85–92 Lancashire PD, Bleiholder H, Boom TVD, Langelu¨ddeke P, Veromann E, Metspalu L, Williams IH, Hiiesaar K, Mand M, Strauss R, Weber E, Witzenberger A (1991) A uniform Kaasik R, Kovaćs G, Jogar K, Svilponis E, Kivimagi I, decimal code for growth stages of crops and weeds. Ann Ploomi A, Luik A (2012) Relative attractiveness of Bras- Appl Biol 119:561–601 sica napus, Brassica nigra, Eruca sativa and Raphanus Landis DA, Wratten SD, Gurr GM (2000) Habitat management sativus for pollen beetle (Meligethes aeneus) and their to conserve natural enemies of arthropod pests in agricul- potential for use in trap cropping. Arthropod-Plant Interact ture. Ann Rev Entomol 45:175–201 6:385–394 Muller FJ, Baur H, Gibson GAP, Mason PG, Kuhlmann U Veromann E, Toome M, Ka¨nnaste A, Kaasik R, Copolovici L, (2007) Review of the species of Trichomalus (Chalci- Flink J, Kovaćs G, Narits L, Luik A, Niinemets U¨ (2013) doidea: Pteromalidae) associated with Ceutorhynchus Effects of nitrogen fertilization on insect pests, their par- (Coleoptera: Curculionidae) host species of European ori- asitoids, plant diseases and volatile organic compounds in gin. Can Entomol 139:643–657 Brassica napus. Crop Prot 43:79–88 Murchie AK (1996) Parasitoids of cabbage seed weevil and Veromann E, Kaasik R, Kovaćs G, Metspalu L, Williams IH, brassica pod midge in oilseed rape. PhD Thesis, University Ma¨nd M (2014) Fatal attraction: search for a dead-end trap of Keele crop for the pollen beetle (Meligethes aeneus). Arthropod- Murchie AK, Williams IH (1998) A bibliography of the para- Plant Interact 8:373–381 sitoids of the cabbage seed weevil (Ceutorhynchus assim- Williams IH (2003) Parasitoids of cabbage seed weevil. In: ilis Payk.). In: Proceedings of the Working group Alford DV (ed) Biocontrol of oilseed rape pests. Blackwell ‘‘Integrated control in oilseed crops’’. Bulletin. Poznan, Science Ltd, Oxford, pp 97–112 pp 163–169 Williams IH (2010) The major insect pests of oilseed rape in Parolin P, Bresch C, Desneux N, Brun R, Bout A, Boll R, Poncet Europe and their management: an overview. In: Williams C (2012) Secondary plants used in biological control: a IH (ed) Biocontrol-based integrated management of oil- review. Int J Pest Manag 58:91–100 seed rape pests. Springer, Netherlands, pp 1–43 Tansey JA, Dosdall LM, Keddie A, Fletcher RS, Kott LS (2010a) Antixenosis and antibiosis resistance to Ceu- torhynchus obstrictus in novel germplasm derived from Gabriella Kovaćs is a PhD student studying the cabbage seed Sinapis alba 9 Brassica napus. Can Entomol 142:212–221 weevil Ceutorhynchus obstrictus and its parasitoids. Tansey JA, Dosdall LM, Keddie A, Fletcher RS, Kott LS (2010b) Responses of Ceutorhynchus obstrictus (Mar- Riina Kaasik is interested in natural enemies of insect pest as sham) (Coleoptera: Curculionidae) to olfactory cues ecosystem service providers.

123 73 38 G. Kova´cs et al.

Tanel Kaart specialises in data analysis for life sciences, Anne Luik is an entomologist devoted to sustainable plant especially (population) genetics, plant and animal science, protection and organic cropping systems. (veterinary) medicine, ecology, psychology. Eve Veromann is dealing with insects as ecosystem service Luule Metspalu is interested in the physiology and toxicology providers and novel biosafe IPM strategies. of insects.

123 74 III Kovács, G.; Kaasik, R.; Treier, K.; Luik, A.; Veromann E. 2016. Do different field bordering elements affect cabbage seed weevil damage and its parasitism rate differently in winter oilseed rape? IOBC-WPRS Bulletin, 116: 75-80

76 Integrated Control in Oilseed Crops IOBC-WPRS Bulletin Vol. 116, 2016 pp. 75-80

Do different field bordering elements affect cabbage seed weevil damage and its parasitism rate differently in winter oilseed rape?

Gabriella Kovács, Riina Kaasik, Kaia Treier, Anne Luik & Eve Veromann Estonian University of Life Sciences, Institute of Agricultural and Environmental Sciences, Department of Plant Protection, Fr. R. Kreutzwaldi 1, 51014 Tartu, Estonia

Abstract: The cabbage seed weevil, Ceutorhynchus obstrictus, is an important oilseed rape crop pest in Europe. Its abundance is usually managed by synthetic insecticides that can be harmful to neutral and beneficial organisms, including parasitoids, occurring in the agricultural fields. Parasitoids can play an important role in the control of the population size of seed weevils. This experiment was conducted to see if and how different field bordering element types affect cabbage seed weevil infestation and parasitism rate in conventionally grown winter oilseed rape crops. The percentage of damaged pods was low (between 8.5% and 10.9%), but even with such low pest abundance the parasitism rate was sufficient for efficient biocontrol; varying between 55.5% and 68%.

Key words: Ceutorhynchus obstrictus, field margins, conservation biological control

Introduction

The cabbage seed weevil Ceutorhynchus obstrictus Marsham (Coleoptera: Curculionidae) is a significant brassicaceous oilseed crop pest in Europe and North America (Dosdall et al., 2001; Alford et al., 2003). Adult weevils feed on oilseed rape plants on two separate occasions during the season, but the main yield loss is caused by larval feeding inside the pods on developing seeds (Doucette, 1947; Free et al., 1983; Buntin, 1999). In the spring, emerged adults feed on buds and flowers of early flowering brassicaceous plants (Doucette, 1947) and invade winter oilseed rape during the bud and/or early flowering stage (Williams & Free, 1978), where they continue feeding, then mate and females lay their eggs, one or two at a time inside pods that are at least 2 mm in diameter (Doucette, 1947; Dmoch, 1965). In Europe parasitoids play an important role in the population regulation of the cabbage seed weevil (Williams, 2003; Veromann et al., 2011; Kovács et al., 2013), however in most oilseed rape growing areas, control measures currently rely on synthetic pyrethroid insecticides applied as foliar sprays (Williams, 2003), which are effective but at the same time create a harmful environment for beneficial organisms. The over use of pesticides threatens the efficacy of biological control by killing natural enemies of pests. When natural regulatory processes are reduced, insecticide applications have to be increased in order to keep pests under control (Pickett et al., 1995), which might lessen the economic competitiveness of the crop. Additionally, the resistance of C. obstrictus to tau-fluvalinate, etofenproxand and lambda-cyhalothrin recorded in Germany (Heimbach & Müller, 2013) shows the importance of avoiding unjustified pesticide use to lessen the likelihood of pesticide resistance development. Hence, alternative management strategies should be favoured. Creating and maintaining alternative habitats for beneficial insects has to be a key task in pest management.

77 75 76

Non-cultivated field margins serve as hibernation sites and offer food sources for adult weevils, supporting their survival, especially if the vegetation is botanically related to the main crop (Altieri, 1999). On the other hand, these sites also provide regulating ecosystem services by maintaining a habitat for the natural enemies of many pests via providing a modulate microclimate (Rahim et al., 1991), overwintering habitat, food resources (Jervis et al., 1993), source of alternative prey, all of which results in higher densities of predators and parasitoids (Lys et al., 1994; Sutherland et al., 2001). The aim of this study was to measure and compare cabbage seed weevil infestation and parasitism rate in oilseed rape fields bordered by different non-crop habitat types.

Material and methods

Study area The study was carried out on commercial winter oilseed rape crops in 2013, Tartu County, Estonia. Four different commonly occurring landscape element types were selected: herbaceous areal, woody areal, herbaceous linear and woody linear. The dimensions of areal elements were at least 60 x 60 meters while the width of linear elements did not exceed 12.5 meters and the length was at least 150 meters. At the sampling point the study crop was directly bordered by one landscape element and the minimum distance to a different element was at least 200 meters. There were 54 sampling points; 19 on winter oilseed rape crops bordered by herbaceous areal elements, 11 by woody areal, 15 by herbaceous linear and 9 by woody linear elements.

Sampling method The cabbage seed weevil damage and parasitism rate was assessed at the ripening stage of oilseed rape (BBCH 83-85) (Lancashire et al., 1991). Pod samples were collected from two different distances per sampling point, 2 and 20 meters from the field margin. At each distance 10 pods (5 pods from the main raceme and 5 pods from the third raceme) from 5 randomly selected plants were collected, 100 pods per sampling point. Collected pods were incubated for four weeks in emergence traps (described in detail in Dosdall et al., 2006; Dosdall & Kott, 2006; Veromann et al., 2011; Kovács et al., 2013) after which time the emerged larvae and parasitoids were counted. Pods were examined and dissected and all exit- holes and weevil or parasitoid remains were noted. The rate of damaged pods and parasitism rate was calculated per plant.

Statistical analyses The effect of field bordering elements on the damage and parasitism rate were analysed using a generalized linear model with normal distribution and identity-link function. The differences between treatments were studied using the same analysis but with Bonferroni correction. Statistical analyses were carried out using STATISTICA 12 (Statsoft Inc. USA, 2014).

Results and discussion

Generally, the damage rate caused by seed weevils was low in all studied winter oilseed rape fields and approximated 10%. In our study, the bordering landscape element type of the oilseed rape crop had no effect on the damage rate caused by C. obstrictus (χ2 = 2.40, df = 3, p = 0.49). The damage rate was similar on all crops regardless their bordering landscape

78 77

elements (Figure 1) probably due to the low average damage rate (9.9% ± 0.6%). However, despite the low damage rate, parasitism rate was high in all studied fields and exceeded 50%, although the crop bordering element type had no effect on the parasitism rate (χ2 = 3.25, df = 3, p = 0.35). The highest parasitism rate was recorded from crops bordered with herbaceous linear elements (68% ± 5.3%) and lowest on locations where oilseed rape was bordered with woody elements: 55.7% (± 7.4%) of C. obstrictus was parasitized next to woody linear elements and 55.5% (± 5.9%) next to woody areal elements (Figure 1). Although not statistically significantly different, a somewhat lower parasitism rate in oilseed rape crops adjacent to woody landscape elements might indicate difficulties in the host finding of parasitoids.

30 80 70 25

60 20 50 15 40 30 10 Damage rate, % Damage 20 % rate, Parasitism 5 10 0 0 Herbaceous areal Woody areal Herbaceous linear Woody linear Damage rate, % Parasitism rate, %

Figure 1. The mean (± SE) percentage of damaged winter oilseed rape pods caused by Ceutorhynchus obstrictus and its larval parasitism rates on crops adjacent to different landscape elements in Tartu County, Estonia, 2013.

Like other parasitoids, species attacking C. obstrictus use volatile cues emitted by host plants (Turlings et al., 1991; McCall et al., 1993; De Moraes & Mescher, 1998; Dicke, 1999); parasitoids of brassicaceous pests may use in addition isiothiocyanates, specific volatile organic compounds emitted by plants of this family, to locate and recognize host plants (Bartlet et al., 1993; Alford et al., 2003; Schiestl, 2010; Williams & Cook, 2010). Both oilseed rape pests and their parasitoids fly upwind towards the scent of host plants (Williams et al., 2007). As the dispersal of volatile compounds is directly linked to physical barriers, woody elements (forests and hedgerows) suppress wind and therefore decrease the dispersal of odours indirectly affecting the host finding success of parasitoids. The odour plumes and their dispersal is known to differ in forests and open landscapes (Murlis et al., 2000) and likely affect the dispersal of both pests and parasitoids. As pests only need to locate the crop they are in a better position compared to parasitoids, which first need to locate the crop and then the host within that habitat (Vinson, 1998) which might explain the equal damage rate between different crop bordering elements. Despite the lack of significant differences among

79 78

parasitism rate, the average parasitism rate per sampling point type varied up to 5%, which suggests that parasitoids might be more affected by landscape composition than pests. However, the parasitism rate in all studied fields surpassed the level of effective pest control (32% according to Hawkins & Cornell, 1994) and we can conclude that in Estonia they can effectively control the population size of seed weevils. In future studies it would be important to determine if and how much different field margins affect the damage rate and parasitism rate of this pest by comparing fields bordered by non-crop habitats with fields bordered directly by another crop field.

Acknowledgements

This study was supported by the project QuESSA that is funded by the European Commission through the Seventh Framework Programme, grant no. 8895 of the Estonian Science Foundation, the capacity-building project no. P9003PKPK of the Estonian University of Life Sciences, the institutional research funding no. IUT36-2 of the Estonian Research Council, and the European Social Fund’s Doctoral Studies and Internationalisation Programme DoRa (which was carried out by Foundation Archimedes).

References

Alford, D. V., Nilsson, C. & Ulber, B. 2003: Insect pests of oilseed rape crops. In: Biocontrol of Oilseed Rape Pests (ed. Alford, D. V.): 9-42. Blackwell Science Ltd., Oxford, UK. Altieri, M. A. 1999: The ecological role of biodiversity in agroecosystems. Agric. Ecosyst. Environ. 74: 19-31. Bartlet, E., Blight, M. M., Hick, A. J. & Williams, I. H. 1993: The responses of the cabbage seed weevil (Ceutorhynchus assimilis) to the odour of oilseed rape (Brassica napus) and to some volatile isothiocyanates. Entomol. Exp. Appl. 68: 295-302. Buntin, G. D. 1999: Damage loss assessment and control of the cabbage seedpod weevil (Coleoptera: Curculionidae) in winter canola using insecticides. J. Econ. Entomol. 92: 220-227. De Moraes, C. M. & Mescher, M. C. 1998: Plant-parasitoid interactions in tritrophic systems: a review of recent work with implications for biological control. J. Entomol. Sci. 34: 31-39. Dicke, M. 1999: Are herbivore-induced plant volatiles reliable indicators of herbivore identity to foraging carnivorous arthropods? Entomol. Exp. Appl. 91: 131-142. Dmoch, J. 1965: The dynamics of a population of the cabbage seedpod weevil (Ceutorhynchus assimilis Payk.) and the development of winter rape. Part I. Ekol. Pol. 249-287. Dosdall, L. M. & Kott, L. S. 2006: Introgression of resistance to cabbage seedpod weevil to canola from yellow mustard. Crop Sci. 46: 2437. Dosdall, L. M., Moisey, D., Cárcamo, H. & Dunn, R. 2001: Cabbage seedpod weevil.pdf [WWW Document]. Agri-Facts Pract. Inf. Alberta’s Agric. Ind. Revis. January 2014 Agdex 622-21. URL: http://www1.agric.gov.ab.ca/$department/deptdocs.nsf/all/agdex2538/$file/622-21.pdf (accessed 4.18.16).

80 79

Dosdall, L. M., Ulmer, B. J., Gibson, G. A. P. & Cárcamo, H. A. 2006: The spatio-temporal distribution dynamics of the cabbage seedpod weevil, Ceutorhynchus obstrictus (Coleoptera: Curculionidae), and its larval parasitoids in canola in western Canada. Biocontrol Sci. Technol. 16: 987-1006. Doucette, C. F. 1947: Host plants of the cabbage seedpod weevil. J. Econ. Entomol. 40: 838-840. Free, J. B., Ferguson, A. W. & Winfield, S. 1983: Effect of various levels of infestation by the seed weevil (Ceutorhynchus assimilis Payk.) on the seed yield of oil-seed rape (Brassica napus L.). J. Agric. Sci. 101: 589-596. Hawkins, B. A. & Cornell, H. V. 1994: Maximum Parasitism Rates and Successful Biological Control. Science 266: 1886-1886. Heimbach, U. & Müller, A. 2013: Incidence of pyrethroid-resistant oilseed rape pests in Germany. Pest Manag. Sci. 69: 209-216. Jervis, M. A., Kidd, N. A. C., Fitton, M. G., Huddleston, T. & Dawah, H. A. 1993: Flower- visiting by hymenopteran parasitoids. J. Nat. Hist. 27: 67-105. Kovács, G., Kaasik, R., Metspalu, L., Williams, I. H., Luik, A. & Veromann, E. 2013: Could Brassica rapa, Brassica juncea and Sinapis alba facilitate the control of the cabbage seed weevil in oilseed rape crops? Biol. Control 65: 124-129. Lancashire, P. D., Bleiholder, H., Boom, T. V. D., Langelüddeke, P., Stauss, R., Weber, E. & Witzenberger, A. 1991: A uniform decimal code for growth stages of crops and weeds. Ann. Appl. Biol. 119: 561-601. Lys, J.-A., Zimmermann, M. & Nentwig, W. 1994: Increase in activity density and species number of carabid beetles in cereals as a result of strip-management. Entomol. Exp. Appl. 73: 1-9. McCall, P. J., Turlings, T. C. J., Lewis, W. J. & Tumlinson, J. H. 1993: Role of plant volatiles in host location by the specialist parasitoid Microplitis croceipes Cresson (Braconidae: Hymenoptera). J. Insect Behav. 6: 625-639. Murlis, J., Willis, M. A. & Cardé, R. T. 2000: Spatial and temporal structures of pheromone plumes in fields and forests. Physiol. Entomol. 25: 211-222. Pickett, J. A., Butt, T. M., Doughty, K. J., Wallsgrove, R. M. & Williams, I. H. 1995: Minimising pesticide input in oilseed rape by exploiting natural regulatory processes. Proceedings of the 9. International Rapeseed Congress, Cambridge, UK, 4-7 July 1995: 565-571. Rahim, A., Hashmi, A. A. & Khan, N. A. 1991: Effects of temperature and relative humidity on longevity and development of Ooencyrtus papilionis Ashmead (Hymenoptera: Eulophidae), a parasite of the sugarcane pest, Pyrilla perpusilla Walker (Homoptera: Cicadellidae). Environ. Entomol. 20: 774-775. Schiestl, F. P. 2010: The evolution of floral scent and insect chemical communication. Ecol. Lett. 13: 643-656. Statsoft Inc. USA 2014: STATISTICA 12. Sutherland, J. P., Sullivan, M. S. & Poppy, G. M. 2001: Distribution and abundance of aphidophagous hoverflies (Diptera: Syrphidae) in wildflower patches and field margin habitats. Agric. For. Entomol. 3: 57-64. Turlings, T. C., Tumlinson, J. H., Heath, R. R., Proveaux, A. T. & Doolittle, R. E. 1991: Isolation and identification of allelochemicals that attract the larval parasitoid, Cotesia marginiventris (Cresson), to the microhabitat of one of its hosts. J. Chem. Ecol. 17: 2235-2251.

81 80

Veromann, E., Williams, I. H., Kaasik, R. & Luik, A. 2011: Potential of parasitoids to control populations of the weevil Ceutorhynchus obstrictus (Marsham) on winter oilseed rape. Int. J. Pest Manag. 57: 85-92. Vinson, S. B. 1998: The general host selection behavior of parasitoid Hymenoptera and a comparison of initial strategies utilized by larvaphagous and oophagous species. Biol. Control 11: 79-96. Williams, I. H. 2003: Parasitoids of cabbage seed weevil. In: Biocontrol of Oilseed Rape Pests (ed. Alford, D.V.): 97-112. Blackwell Science Ltd. Williams, I. H., & Cook, S. M. 2010: Crop location by oilseed rape pests and host location by their parasitoids. In: Biocontrol-Based Integrated Management of Oilseed Rape Pests (ed. Wiiliams, I. H.): 215-244). Springer, Netherlands. Williams, I. H., Frearson, D. J. T., Barari, H. & McCartney, A. 2007: First field evidence that parasitoids use upwind anemotaxis for host-habitat location. Entomol. Exp. Appl. 123: 299-307. Williams, I. H. & Free, J. B. 1978: The Journal of Agricultural Science – The feeding and mating behaviour of pollen beetles (Meligethes aeneus Fab.) and seed weevils (Ceutorhynchus assimilis Payk.) on oil-seed rape (Brassica napus L.). Cambridge Journals Online. J. Agric. Sci. 91: 453-459.

82 IV Kovács, G.; Kaasik, R.; Lof, M.; van der Werf, W.; Kaart, T.; Holland, J.H.; Luik, A.; Veromann, E. 2018. Effects of land use on infestation and parasitism rates of cabbage seed weevil in oilseed rape. Pest Management Science DOI: https://doi.org/10.1002/ps.5161

84 Research Article

Received: 22 March 2018 Revised: 27 July 2018 Accepted article published: 2 August 2018 Published online in Wiley Online Library:

(wileyonlinelibrary.com) DOI 10.1002/ps.5161 Eﬀects of land use on infestation and parasitism rates of cabbage seed weevil in oilseed rape Gabriella Kovács,a* Riina Kaasik,a Marjolein E Lof,b Wopke van der Werf,b Tanel Kaart,c John M Holland,d Anne Luika and Eve Veromanna

Abstract

BACKGROUND: This study investigated how infestation rates of an important oilseed rape pest, the cabbage seed weevil (Ceutorhynchus obstrictus) and rates of parasitization by its parasitoids are aﬀected by land use, up to 1000 m from 18 focal ﬁelds.

RESULTS: The mean proportion of C. obstrictus-infested pods per plant was 8% (2–19.5%). Infestation rates were higher if the adjacent habitat was a herbaceous semi-natural habitat than if it was either another crop or a woody habitat. Infestation rates were positively related to the area of herbaceous semi-natural vegetation, permanent grassland and wheat (which followed oilseed rape in the crop rotation) at a spatial scale of at least 1 km. The mean parasitism rate of C. obstrictus larvae was 55% (8.3–87%), suﬃcient to provide eﬃcient biocontrol. Parasitism rates were unrelated to adjacent habitats, however, they were positively related to the presence of herbaceous linear elements in the landscape and negatively related to permanent grasslands at a spatial scale of 200 m.

CONCLUSION: Proximity of herbaceous elements increased both infestation rates and parasitism, while infestation was also related to landscape factors at larger distances. The ﬁndings provide an empirical basis for designing landscapes that suppress C. obstrictus, at both ﬁeld and landscape scales. © 2018 Society of Chemical Industry

Supporting information may be found in the online version of this article.

Keywords: Ceutorhynchus obstrictus; conservation biological control; semi-natural habitats; landscape ecology; hymenopteran parasitoids

1 INTRODUCTION on approximately five seeds inside the pod, resulting in an 18–19% 12 Sustainable approaches are needed to manage agricultural land- loss in seed weight per pod. In the case of infestations higher scapes in a way that enhances regulating ecosystem services such than 26–40% pods per plant, larval feeding can significantly 12–14 as pest control by natural enemies. Natural enemies and other ben- reduce yield. After the larva has finished its development eficial insects such as pollinators often utilize semi-natural habi- inside the pod, it chews an exit hole through the pod wall and tats (SNHs) in agricultural landscapes for foraging or shelter.1–4 drops to the ground to pupate in the upper layer of the soil. There is substantial variation in the literature on the contribution Next-generation adults emerge later in the summer, feed on that different SNHs within agricultural landscapes make towards various brassicaceous plant species, and overwinter in perennial 7,15 supporting pest control services and these contributions are vegetation and leaf litter. context-dependent.5,6 Furthermore, the majority of studies on the subject are from a limited number of Western European countries.4 Thus, to clarify and expand our current understanding of the role ∗ Correspondence to: G Kovács, Institute of Agricultural and Environmental of different semi-natural and crop habitats in ecosystem service Sciences, Estonian University of Life Sciences, Fr. R. Kreutzwaldi 5, 51006 Tartu, Estonia. E-mail: [email protected] provisioning across Europe, more research is needed, especially in areas that are poorly represented in the literature (e.g. Eastern a Institute of Agricultural and Environmental Sciences, Estonian University of Europe). Life Sciences, Tartu, Estonia The cabbage seed weevil, Ceutorhynchus obstrictus (Marsham) b Centre for Crop Systems Analysis, Department of Plant Sciences, Wageningen (Coleoptera: Curculionidae) is an important univoltine insect pest University and Research, Wageningen, The Netherlands of Brassica species in Europe7,8 and North America.9,10 The cabbage seed weevil migrates from overwintering habitats to brassicaceous c Institute of Veterinary Medicine and Animal Sciences, Estonian University of Life Sciences, Tartu, Estonia plant species, including winter oilseed rape (OSR), during early flowering.8,11 Usually, one egg is laid per pod, and one larva feeds d Game & Wildlife Conservation Trust, Fordingbridge, UK

85 www.soci.org G Kovács et al.

In most OSR-growing areas in Europe and North America,16,17 2 MATERIALS AND METHODS control of C. obstrictus relies on prophylactic pyrethroid insec- 2.1 Study area and site selection ticide applications.8,18,19 Prophylactic pesticide applications can The field work was conducted in eastern Estonia, Tartu County, interfere with biological control, reducing the abundance of not between Lake Võrtsjärv and Tartu (latitude: 58.21∘N to 58.44∘N, only the target pest, but also beneficial insects including natural longitude: 26.17∘E to 26.66∘E; elevation: 32–83.5 m a.s.l.) in 2014. enemies and pollinators.20–23 The non-target effect of agrochem- In this area, in 2014 the average annual air temperature was 6.8 ∘C ical inputs should be taken into consideration when pest man- (30-year average: 5.8 ∘C) and the precipitation was 701.1 mm agement decisions are made, because natural enemies, mainly (30-year average: 680 mm).53,54 Eighteen non-overlapping land- hymenopteran ectoparasitoids, provide an important ecosystem scape circles of 1-km radius were selected in the overall study area service by regulating the population of C. obstrictus.16,24–28 In to create a gradient of land use and the proportion of SNHs around Europe, the key parasitoid species of C. obstrictus are solitary idio- the winter OSR crops to be studied (focal field). Selection of land- biont larval ectoparasitoids belonging to the hymenopteran family scape circles was furthermore based on the type of habitat bor- Pteromalidae: Trichomalus perfectus (Walker), Mesopolobus morys dering the focal field of each circle, such that we would have six (Walker) and Stenomalina gracilis (Walker).24,25,29,30 Of these three replicate landscape circles for each of three types of adjacent habi- species, T. perfectus is the most abundant and therefore the most tat: crop field, linear herbaceous SNH or linear woody SNH. The studied.24,30 Overwintered T. perfectus females arrive to OSR crops study site selection (Section 2.1), landscape description and clas- a few weeks after their host by the time seed weevil larvae have sification of SNHs (Section 2.2), and experimental design (Section reached their second/third instar within the pods.31,32 They lay one 2.3) were based on standardized protocols of the EU FP7 project egg on the host larva, which hatches after a maximum 4 days and QuESSA across participating regions and crops.55–60 feeds on its host externally for ∼ 10 days. The larva then pupates within the pod for 8–15 days after which the adult chews through 2.2 Landscape description and classification of semi-natural the pod wall to exit.24,33 They overwinter in protected places, for habitats example, in evergreen vegetation.32,34 SNHs were classified according to their dominant vegetation SNHs such as field boundaries, extensive grassland and wood- (herbaceous or woody) and shape (areal or linear). Woody SNHs land can provide natural enemies of insect pests with refuge, over- had at least 30% woody vegetation cover, otherwise they were wintering sites and alternative food resources,1,3,35 because they considered as herbaceous. Five types of SNHs were defined: herba- are less disturbed habitats than crops. Both C. obstrictus and its ceous areal (natural hayfields or abandoned fields that have not parasitoids depend on SNHs for feeding before and after winter developed 30% shrub/tree canopy cover), woody areal (wood- hibernation, and for overwintering. Adult parasitoids need nec- > land), cover crop (in most cases clover), herbaceous linear (narrow tar for both survival and fecundity,36,37 and SNHs near agricul- grassy crop margin) and woody linear (hedge, line of trees). The tural fields can provide wild plants with easily accessible nectar minimum size of each mapped habitat was 75 m2 (with a minimum resources. width of 1.5 m and a minimum length of 50 m). The width of linear Ceutorhynchus obstrictus has not previously been studied in a SNHs was between 1.5 and 25 m, and minimum size of areal landscape context. Several landscape-scale studies on biological ≥ < SNHs was 1250 m2 (at least 25 m wide and 50 m long) to guarantee control in OSR focus on another important OSR pest, the pollen an impact on ecosystem service delivery. beetle Brassicogethes aeneus (Fabricius) (Coleoptera: Nitidulidae; Based on interactive maps61,62 and thorough field inspections, syn. Meligethes aeneus (Fabricius)),38–44 or on a pest complex land use was recorded in a 1-km radius around each focal field. excluding C. obstrictus.45,46 Our previous work on C. obstrictus27 Land use was mapped using ArcGIS (version 10.1, Esri, Redlands, addressed only the effects of marginal habitats and did not con- CA, USA) and the resulting GIS data were used to calculate the sider the wider landscape. Here, we extend earlier work by assess- area and proportion of each habitat type. The habitat types ing the effect of bordering habitat on not only C. obstrictus, but also mapped were: all the above-mentioned SNHs (five categories); its parasitoids. Moreover, we assess the effects of the surround- crops (annual herbaceous crops: 16 categories, perennial herba- ing landscape on both the pest and its parasitoids. We hypothe- ceous crops: one category, perennial woody crops: three cate- size that infestation rates of the pest are not strongly related to gories); urban areas (four categories – based on the proportion of semi-natural elements in the landscape, because OSR as a crop is green area within the habitat); water courses (one category); and already an important reservoir for the pest. However, we hypoth- non-habitats (one category, roads, lakes, etc.). Detailed informa- esize that semi-natural landscape elements, which may be rich in tion on the mapped habitat categories is given in Table S1. Based flower resources and are not affected by pesticides, may be impor- on the maps, percentages of each habitat type were calculated for tant resource habitats for parasitoids. Therefore, we expect that the total 1-km radius area around the sampled portion of the focal parasitism rates will be more strongly related to the total pro- field (see below). portion of SNHs in the surrounding landscape than pest infestation rates are. We further hypothesize that the distance of SNHs from the field affects their importance in the delivery of ecosys- 2.3 Experimental design tem services to crop fields. It has been suggested that interacting The sampled side of the focal field was adjacent to either another species experience the landscape at different special scales based crop field (six replicate landscapes; control, CT), a linear herba- on body size – smaller species disperse more locally than larger ceous SNH (HL) (six replicates), or a linear woody SNH (WL) (also species – 47 or the combinations of body size, trophic level and spe- six replicates) for a total of 18 landscape replicates, each with one cialization – specialized parasitoids have often smaller dispersal focal field. Sampling within the focal field was done on a transect ranges than their (herbivore) hosts.48–50 As C. obstrictus is a strong extending from the edge of the field bordering the SNH (or a crop flier,8,51,52 possibly stronger than its parasitoids, we expect any field) 75 m into its interior. Of the six adjacent crop fields, two were landscape effects on pest infestation rates to extend over larger winter wheat, two were spring OSR, one was spring barley and one spatial scales than landscape effects on parasitism. was potato. All focal fields were managed conventionally; however, wileyonlinelibrary.com/journal/ps © 2018 Society of Chemical Industry Pest Manag Sci (2018)

86 Landscape effects on Ceutorhynchus obstrictus infestation and parasitism rates www.soci.org none was treated with insecticides against C. obstrictus. Informa- The model fitting was conducted in R using a loop to find the best tion about crop management is summarized in Table S2. scale parameter of the kernel model, whereas automated model selection with the dredge function in R was used to find the best 2.4 Insect sampling fitting linear mixed effects model for each kernel scale. Both for Within each of the 18 focal fields, pod samples were taken at parasitism of C. obstrictus larvae and for infestation of OSR we the beginning of the ripening stage of the crop (BBCH 80–81)63 ran dredge for 11 values for the spatial scales parameter u = 50, to measure cabbage seed weevil infestation rates and parasitism 100, 200, 300, 400, 500, 600, 700, 800, 900 and 1000 m. For each levels. Samples were collected along a transect perpendicular to spatial scale the key habitats, the source strengths and the model the adjacent field boundary at distances of 2, 25, 50 and 75 m statistics [P-values, Akaike’s and Bayesian information criteria (AIC from the edge in each field. Each sample consisted of 10 pods and BIC) and corrected Akaike’s information criterion (AICc)] were (five from the main and five from the third raceme from the top) stored. Final model selection across all kernel-scale parameters per plant. Five randomly chosen plants were sampled at each was performed by ranking models based on Akaike’s information distance, giving 50 pods per distance in each field, and a total of criterion for small sample sizes (AICc). The kernel approach was 200 pods (20 samples) per field. Pods were placed in cardboard applied using a tailor-made R script (available on request). emergence boxes and kept at room temperature for 30 days as To test the contribution of distance-weighting to the explanatory described by Veromann et al.25 and Kovács et al.26; similar to the value of the model, we also fitted models in which the weight method used by Dosdall et al.64 Emerged larvae and parasitoids was set to one across all distances. The best model was selected were counted, and all pods were examined and dissected in order with AIC. to detect weevil or parasitoid remains and exit holes. Parasitoid adults were preserved in 70% ethanol, at −20 ∘C and identified 2.5.2 Analysis of the effects of adjacent habitats and distance from using an Olympus SZX-ZB9 stereo-microscope (Olympus Optical the edge of the field Co, Ltd, Japan) with a 2289 maximum magnification, coupled to an The effects of adjacent habitats and distance from the edge of the Olympus Highlight 3100 light source (Olympus Optical Co, Europa, field on the incidence of C. obstrictus (number of pods infested GMBH, Hamburg) using the keys of Graham,34 Trjapitzin,65 Gibson of 50 per location per field) and the incidence of parasitism in et al.,66 Baur et al.67 and Muller et al.68 Voucher specimens from the infested pods were analysed by fitting generalized linear Rothamsted Research, UK were used to confirm the identification of T. perfectus, M. morys and S. gracilis. mixed effects models. The regression model included the adjacent habitat (categorical with three levels) and distance from the edge (categorical with four levels) and their interaction as fixed effects 2.5 Statistical analysis and a random intercept to represent variability between the 18 2.5.1 Landscape level analyses sites. We used a logit link and binomial error model. The strength of association between the infestation rate and par- We focused analysis of the species composition of the parasitoid asitism rate, and between the abundance of different parasitoid community on three species, T. perfectus, S. gracilis and M. morys. groups, was studied with pairwise Pearson correlation analysis The remaining species were grouped as ‘other parasitoid species’ using procedure CORR in SAS 9.4. The effect of the mapped habi- because of their low abundance. The effect of adjacent habitat tats on C. obstrictus and its parasitoids was analysed using a sta- type on the proportion of parasitoid species was analysed with a tistical model in which contributions of habitats across the land- logistic regression model that included the adjacent habitat (cat- scape were summed over the whole landscape using weighting egorical with three levels) as a fixed effect and landscape circles that decreased with distance between the focal field and the habi- as random effect. The logistic models were fitted using proce- tat according to a statistical kernel model based on the rotationally dure GLIMMIX in SAS 9.4. All results were considered statistically symmetric two-dimensional t-distribution.69–72 For this analysis, significant if P 0.05. the maps were converted to a finely grained raster with cells of ≤ 2.5 × 2.5 m to preserve the smaller landscape features. The analysis focused on habitats that were present in at least 10 landscape circles: six SNHs (herbaceous areal, cover crop, herbaceous 3 RESULTS linear, woody linear, woody areal edges and woody areal interior); 3.1 Landscape-level analyses three crop species (wheat, barley and OSR); permanent grasslands; 3.1.1 Infestation with Ceutorhynchus obstrictus urban areas; water courses; and non-habitats (roads, lakes, etc.). The proportion of C. obstrictus-infested pods per plant varied from As dependent variables in this analysis, we considered the pro- 2% to 19.5% infested pods per field (Table 2), which is below the portion of damaged pods in a sample of 200 pods from each threshold of considerable damage (26–40%).12–14 The total pro- focal field and the proportion of the damaged pods that were portion of SNHs in landscape circles ranged from 20.9% to 75.9% parasitized. (Table 2), and did not have a significant effect on the infesta- The effect of landscape on infestation rate and parasitism rate tion rate (P = 0.412) according to the kernel model. A combina- was estimated with a generalized linear mixed effects model73 tion of herbaceous areal SNHs, wheat and permanent grasslands with a binomial distribution and a logit link, using the distance in the surrounding landscape best explained the infestation rate weighted landscape variables as regressors, where the distance (Table 1). The AIC of the optimal kernel model with a scale parame- weighting was obtained from the t-distribution kernel with param- ter of 600 m was less than two points better than a model that sim- eters that were fitted to the data by maximizing the likelihood.71,72 ply included the fraction of these three habitats within the 1-km We used the dredge function from the package MuMIn in R74,75 to radius, without weighting for distance. Both models thus have sim- select a set of habitats that most strongly affected the empirically ilar support from the data, and both indicate that habitats across measured parasitism and infestation rates. Considering the limited the entire 1-km landscape circle affected pest infestation. In both number of 18 independent data points (landscape circles and focal models, herbaceous areal SNHs, wheat and permanent grasslands fields), we included at most three explanatory variables (Table 1). within the 1-km radius significantly increased the infestation rate.

87 www.soci.org G Kovács et al.

Table 1. Key habitats that aﬀected infestation rate by and parasitism rate of Ceutorhynchus obstrictus. Habitats in bold increased infestation rate (disservice) and parasitism rate (service), underlined habitats decreased parasitism rate (disservice)

a b c Model u (m) AICc AIC 훽0 훽1 훽2 훽3

Infestation rate – fraction – 129.4 124.4 Intercept Herbaceous areal Wheat Permanent grassland in landscape Estimate SE −4.04 ± 0.38 6.69 ± 2.36 5.19 ± 1.29 3.00 ± 1.08 P-value <0.0001 0.0047 <0.0001 0.0055 Infestation rate – distance 600 128.1 123.1 Intercept Herbaceous areal Wheat Permanent grassland weighted Estimate SE −3.94 ± 0.34 10.63 ± 3.26 6.81 ± 1.60 3.42 ± 1.39 P-value <0.0001 0.0011 <0.0001 0.0142 Parasitism rate – fraction – 90.1 88.4 Intercept Non-habitat (Roads) in landscape Estimate SE 0.41 ± 0.21 −9.55 ± 4.70 P-value 0.0545 0.0420 Parasitism rate – distance 200 80.3 77.2 Intercept Herbaceous linear Permanent grassland – weighted Estimate SE −0.43 ± 0.23 90.65 ± 19.80 −5.70 ± 1.69 P-value 0.057 <0.0001 <0.001 a Spatial scale parameter in meters. b Akaike’s information criterion for small sample sizes. c Akaike’s information criterion.

Table 2. Mean (± SE) proportion of Ceutorhynchus obstrictus-infested oilseed rape (OSR) pods and parasitized C. obstrictus larvae per focal ﬁeld, and the total proportion of all semi-natural habitats (SNHs) around the focal ﬁelds

Adjacent habitat OSR variety Infestation(%) Parasitism(%) Total SNH(%)

Herbaceous linear SNH Abakus 7.0 ± 1.8 75.0 ± 13.4 22.9 Herbaceous linear SNH Visby 11.5 ± 2.8 25.0 ± 15.7 33.7 Herbaceous linear SNH Sherpa 6.0 ± 1.8 62.5 ± 15.7 28.0 Herbaceous linear SNH Thorin 18.5 ± 3.6 81.2 ± 8.0 34.2 Herbaceous linear SNH Thorin 19.5 ± 3.6 66.1 ± 9.6 46.6 Herbaceous linear SNH Thorin 6.5 ± 2.0 66.7 ± 16.7 37.0 Woody linear SNH Visby 6.0 ± 2.6 58.3 ± 15.7 24.0 Woody linear SNH Rohan 8.5 ± 2.0 40.3 ± 12.0 30.7 Woody linear SNH Rohan 10.0 ± 3.6 37.7 ± 15.0 35.5 Woody linear SNH Rohan 8.5 ± 2.3 87.0 ± 8.7 39.3 Woody linear SNH Rohan 3.5 ± 1.3 8.3 ± 8.3 40.0 Woody linear SNH Rohan 5.0 ± 1.4 66.7 ± 16.7 75.9 Another crop Thorin 14.0 ± 2.8 64.2 ± 10.5 43.0 Another crop Rohan 2.0 ± 1.2 33.3 ± 33.3 68.7 Another crop Rohan 3.5 ± 1.3 41.7 ± 20.1 26.1 Another crop Visby 3.0 ± 1.5 50.0 ± 20.4 23.4 Another crop Rohan 4.0 ± 1.5 58.3 ± 20.1 47.7 Another crop Rohan 10.5 ± 2.9 58.5 ± 13.1 20.9

3.1.2 Parasitism of Ceutorhynchus obstrictus larvae permanent grasslands reduced parasitism rates. The contribution The parasitism rate of C. obstrictus larvae varied from 8.3% to of herbaceous linear SNHs and permanent grasslands was highest 87% per field, and dropped below the threshold value of effec- close to the focal field and decreased rapidly with distance from tive biological control (32–36%)76 in only two of the 18 focal the focal field. Figure 1 shows the short-scale positive effects of fields (Table 2). The total proportion of SNHs in the surround- herbaceous linear SNHs in the landscape on parasitism in one of ing landscape did not have a significant effect on the parasitism the landscape circles (for a greyscale version see Fig. S2). rate (P = 0.626). The optimal scale of the kernel model was 200 m, indicating that parasitism rates were predominantly affected by 3.2 Effects of adjacent habitats and distance from the edge herbaceous linear SNHs, and permanent grasslands within the of the field first 750 m from the focal field (Table 1, Fig. S1). The model using 3.2.1 Effect of adjacent habitats and distance from the field edge the distance-weighted landscape variables had better explanatory on pest incidence power than the model that assumed equal weight for all distances Infestation rate depended on the type of habitat bordering the (Table 1). Herbaceous linear SNHs increased parasitism rates and focal field (P = 0.032), but there was no effect of sampling distance wileyonlinelibrary.com/journal/ps © 2018 Society of Chemical Industry Pest Manag Sci (2018)

88 Landscape eﬀects on Ceutorhynchus obstrictus infestation and parasitism rates www.soci.org

Figure 1. (a) Map of all habitats in one of the 18 landscape circles with 1 km radius around the sampling point (marked by X) of the focal field (with diagonal stripes from the upper left to the lower right); (b) predicted parasitism rates (from red to blue, highest to lowest) of Ceutorhynchus obstrictus for all agricultural fields based on the spatial configuration of herbaceous linear semi-natural habitats (pink in Fig. 1a) and permanent grasslands (light green in Fig. 1a) in the depicted landscape circle. from the field edge into the crop or the interaction (all P-values from 8.3% to 87% across fields with an average of 55%. The > 0.05). The highest percentage of infested pods was found in level of infestation stayed below the estimated threshold value of focal fields adjacent to herbaceous linear SNHs, where 12% of pods measurable damage (26–40%)12–14 in all studied fields. In 16 of were infested on average, which was significantly higher than in the 18 fields the parasitism rate was above the threshold value focal fields adjacent to linear woody SNHs (P = 0.041) or focal fields of effective biological control (32–36%).76 Thus, hymenopteran bordering a crop field (P = 0.019; Fig. 2). parasitoids controlled cabbage seed weevil populations in most of the studied landscapes. 3.2.2 Effects of adjacent habitats and distance from the field edge At the landscape scale, the infestation rate of OSR pods by on parasitism of Ceutorhynchus obstrictus C. obstrictus was positively affected by herbaceous areal SNHs, The estimated average parasitism rate of C. obstrictus was 52%, and wheat and permanent grassland. When permanent grasslands was not significantly affected by the type of bordering habitat or were within a few hundred metres from the studied fields, they also affected the parasitism rate of C. obstrictus larvae, but nega- distance from the field edge (all P-values > 0.05; Fig. 3). According to the Pearson correlation analysis, there was no relationship tively. Parasitism rates were higher in winter OSR fields with more between infestation rate and parasitism rate (r = 0.39, P = 0.112). herbaceous linear SNHs within close proximity, i.e. a few hundred metres from the studied winter OSR fields. However, at the field scale, herbaceous linear habitats were associated with increased 3.3 Species composition of the parasitoid community infestation rates by C. obstrictus, when they were directly border- In total, 165 parasitoid specimens were reared from cabbage seed ing the studied fields. Thus, the effect of habitats varies across weevil larvae in this study. Trichomalus perfectus was the most scales, and a habitat may enhance both the pest and its enemy, abundant species with 122 specimens, followed by S. gracilis (20 as found here for herbaceous linear habitats. The finding suggests specimens) and M. morys (18 specimens). The remaining reared that SNH habitats play a role in the life cycle of both the pest and parasitoid species were represented by one specimen each: T. its parasitoids at a landscape scale. lucidus (Walker) (Hymenoptera: Pteromalidae), one unidentifiable Trichomalus species, Anaphes fuscipennis (Haliday) (Hymenoptera: Mymaridae), Eurytoma curculionum (Mayr) (Hymenoptera: Eury- 4.1 Landscape-level analyses tomidae), and one unidentifiable parasitoid species. The species In this study, a relationship was found between the incidence of composition of parasitoids was similar across adjacent habitat seed weevil-infested pods and the fraction of herbaceous areal types (Fig. 4). There was a strong negative correlation between the SNHs, permanent grasslands and wheat in the surrounding land- abundance of S. gracilis and T. perfectus (r = −0.79, P < 0.001) and scape. The relationship with wheat has a biological interpretation a moderate negative correlation between the abundance of other because across the study area OSR is mostly followed by wheat in parasitoid species and T. perfectus (r = −0.59, P = 0.009), and M. the crop rotation. Therefore, the area of wheat in this study cor- morys and T. perfectus (r = −0.48, P = 0.046). responds to the area of the previous year’s OSR. Herbaceous areal SNHs and permanent grasslands are perennial habitats, which are ideal overwintering sites and also provide alternative food for C. 4 DISCUSSION obstrictus before and after winter hibernation, explaining their In this study, we determined how local and landscape factors positive effect on infestation rates. Seed weevils may also find affected the incidence of the cabbage weevil C. obstrictus and suitable overwintering sites and alternative food in the vicinity of its parasitoids in OSR. The proportion of C. obstrictus-infested (recently harvested) OSR fields, similarly to pollen beetles, which in pods per plant varied from 2% to 19.5% across fields, with a the study of Rusch et al.,40 were more abundant in overwintering mean of 8%. The parasitism rate of C. obstrictus larvae varied sites close to the previous year’s OSR fields.

89 www.soci.org G Kovács et al.

Figure 2. Estimated proportion of Ceutorhynchus obstrictus-infested oilseed rape pods (± SE) depending on the distance from the ﬁeld edge, and the adjacent habitat type (CT, another crop as control, white; HL, herbaceous linear, light grey; WL, woody linear, dark grey). According to the two-factorial logistic model: Pdistance = 0.911, Padjacent habitat = 0.032, Pdistance*adjacent habitat = 0.671.

Figure 3. Estimated parasitism rate of Ceutorhynchus obstrictus (± SE) in oilseed rape, depending on the distance from the ﬁeld edge and the adjacent habitat type (CT, another crop as control, white; HL, herbaceous linear, light grey; WL, woody linear, dark grey). According to the two-factorial logistic model: Pdistance = 0.188, Padjacent habitat = 0.232, Pdistance*adjacent habitat = 0.406.

The weak distance effect of habitats on infestation rate by C. sampling point). To the best of our knowledge, this is the first time obstrictus suggests that habitats outside the 1-km radius affected that a relationship has been described between the parasitism the infestation rate, highlighting the need for increasing the radius rate of seed weevils and landscape elements. Parasitoids of C. of the landscape circles in future studies. This result is consistent obstrictus, just like their hosts, also leave oilseed crops after they with studies on other OSR pests, for example the pollen beetle, emerge from pods to seek easily accessible nectar resources which is found to disperse over 1–2 km.77 Parasitoids, however, and overwintering sites,8,15,24,32,33 but as relatively small specialist disperse at smaller scales than herbivorous insects,47 which is insects they most likely have a shorter dispersal range8,47–50 and supported by the result of this study; parasitoids were affected by therefore stay closer to OSR fields. The proportion of linear grass the distance of different habitats from the study area, at a smaller margins in the surrounding landscape has been previously found landscape scale. According to Tylianakis et al.78 and Lavandero to be positively related to pest control,80 by serving as overwin- et al.,79 the ecosystem service provided by parasitoids is more tering sites for predators,81 therefore, it is possible that parasitoids intensive close to their source/refuge habitats, and the intensity found their food resources on the herbaceous linear SNHs near decreases with increasing distance from these areas. Our results the OSR fields and after feeding stayed there to overwinter. confirmed this. Parasitism rate was increased by herbaceous linear Permanent grasslands in the landscape increased the infestation SNHs and reduced by permanent grasslands when they were rate and reduced parasitism rates. According to earlier studies, it within 750 m of the sampling point (with the largest contribution is possible, that a habitat is either more suitable for pest organ- of herbaceous linear SNHs and permanent grasslands close to the isms than it is for natural enemies, or pests disperse from the given wileyonlinelibrary.com/journal/ps © 2018 Society of Chemical Industry Pest Manag Sci (2018)

90 Landscape eﬀects on Ceutorhynchus obstrictus infestation and parasitism rates www.soci.org

field edge into the field, parasitoids managed to locate their hosts. Furthermore, parasitism was not related to host availability, which supports Ulber and Vidal84 and Ferguson et al.85 who found that T. perfectus (the most abundant parasitoid in our study) did not attack cabbage seed weevil larvae in a density-dependent way. These findings suggest that, even without host density depen- dence, parasitoids can effectively contribute to the biological control of the cabbage seed weevil.

4.3 Species composition of the parasitoid community The parasitoid species composition was not affected by the bordering SNH type. Trichomalus perfectus, one of the most abundant parasitoids of C. obstrictus in Europe,30 was the dominating ectoparasitoid in this study, outcompeting all other parasitoid species present. This supports earlier findings of Murchie33 in the UK, Ulber and Vidal84 in Germany, and Veromann et al.25 in Estonia, Figure 4. Proportion of parasitoid species reared from Ceutorhynchus that in winter OSR, T. perfectus plays the main role in controlling obstrictus larvae collected from oilseed rape depending on the adjacent cabbage seed weevil populations. habitat type. According to the logistic model there were no statistically significant differences between adjacent habitat types (CT, another crop as control; HL, herbaceous linear; WL, woody linear; from bottom to top: 4.4 Limitations of the study design M. morys, white; S. gracilis, dark grey with white numbers; T. perfectus, We acknowledge that besides the proportion and location of light grey; other parasitoid sp., black): PM. morys = 0.750, PS. gracilis = 0.220, semi-natural and crop habitats in a landscape there are other PT. perfectus = 0.200, POther parasitoid sp. = 0.928. important factors that affect pest incidence and control, such as agrochemical inputs, weather conditions and vegetation habitat more successfully than parasitoids.5,39,41,82 Therefore, per- cover/composition of SNHs. Thus, to build a stronger overall manent grasslands in the current study possibly promoted the evidence base, these other factors should also be taken into overwintering and then dispersal of seed weevils more than that consideration in future studies. Furthermore, because most of of the parasitoids. the above-mentioned factors vary between years, long-term The total proportion of SNHs in the landscape did not affect experiments are needed. the plant–herbivore–parasitoid interaction. This can be explained by different SNH types having different effects on infestation and parasitism rates. Another possible reason for this finding is the high 5 CONCLUSIONS proportion of SNH across the landscapes, ≥ 20.9%, and ≤ 75.9%. This study was designed to measure how the incidence of the Thies and Tscharntke38 found that when the area of SNHs dropped cabbage seed weevil and its parasitoids in OSR are affected by below 20%, the parasitism rate of the pollen beetle dropped below semi-natural and crop habitats close to the OSR field and in the the threshold value of effective biological control (∼ 32%). Thus, wider landscape. Based on the results, it was the type of SNHs landscapes in Estonia have (still) sufficient SNH to support natural rather than the total proportion of SNHs in the landscape that influ- enemies of C. obstrictus and keep densities of this pest below enced infestation levels and parasitism rate of this pest. Although damaging levels. herbaceous linear SNHs at a local scale increased the pest infestation rate, parasitism rates in crops next to herbaceous linear SNHs 4.2 Effects of adjacent habitats and distance from the edge were very high (62.5–81.2%). In addition, at the landscape scale, of the field herbaceous linear landscape elements had a significant positive At the focal field scale, a higher percentage of pods was infested in effect on parasitism rates. Herbaceous linear SNHs do not cover fields bordered by an herbaceous linear SNH compared with the a large area in the landscape, but in areas under intensive crop control or fields bordering woody linear SNHs. Hedgerows were production, these might be the only type of SNHs present, which considered to limit pest migration across the landscape, result- underlines their importance in the landscape. We conclude that ing in reduced pest infestation rates.83 However, in this study bet- adding more linear herbaceous SNHs into the agricultural land- ter accessibility of focal fields only explains the higher infestation scape would further increase the potential of natural pest control. rates for those bordered by herbaceous linear SNHs, it does not explain the significantly lower infestation rates in focal fields next to another cropped field, which were similarly easy to access. A ACKNOWLEDGEMENTS more likely explanation is that herbaceous linear SNHs were more We thank all the farmers for allowing access to their fields, all suitable refuges for C. obstrictus, by being less disturbed habitats of those who assisted with the fieldwork, and Ivar Kängsepp for than crops. Although herbaceous linear SNHs did contribute to artwork support. This study was conducted as part of the QuESSA increased infestation rates at the field scale, infestation rates were (Quantification of Ecological Services for Sustainable Agriculture) kept below the threshold by natural enemies, which were sup- project, which received funding from the European Union’s Sev- ported by these very same habitats on a landscape scale. Thus, the enth Framework Programme for research, technological devel- contribution of herbaceous linear SNHs to ecosystem services may opment and demonstration under grant agreement no. 311879. have surpassed its contribution to disservices. The experimental design was developed by the QuESSA project High parasitism rates in this study show that, regardless of the partners.55 The work was also supported by the Institutional adjacent SNH type and distance of the sampled plants from the Research Funding project no. IUT36-2 of the Estonian Research

91 www.soci.org G Kovács et al.

Council, the European Social Fund’s Doctoral Studies and the 20 Moreby SJ, Southway S, Barker A and Holland JM, A comparison of Internationalisation Programme DoRa (which was carried out by the effect of new and established insecticides on nontarget inver- Foundation Archimedes). tebrates of winter wheat fields. Environ Toxicol Chem 20:2243–2254 (2001). 21 Gill RJ, Ramos-Rodriguez O and Raine NE, Combined pesticide exposure severely affects individual- and colony-level traits in bees. SUPPORTING INFORMATION Nature 491:105–108 (2012). Supporting information may be found in the online version of this 22 Ceuppens B, Eeraerts M, Vleugels T, Cnops G, Roldan-Ruiz I and Smag- article. ghe G, Effects of dietary lambda-cyhalothrin exposure on bumble- bee survival, reproduction, and foraging behavior in laboratory and greenhouse. J Pest Sci 88:777–783 (2015). 23 Ulber B, Klukowski Z and Williams IH, Impact of insecticides on para- REFERENCES sitoids of oilseed rape pests, in Biocontrol-Based Integrated Manage- 1 Landis DA, Wratten SD and Gurr GM, Habitat management to conserve ment of Oilseed Rape Pests, ed. by Williams IH. Springer, Dordrecht, natural enemies of arthropod pests in agriculture. Annu Rev Entomol pp. 337–355 (2010). 45:175–201 (2000). 24 Williams IH, Parasitoids of cabbage seed weevil, in Biocontrol of Oilseed 2 Duelli P and Obrist MK, Regional biodiversity in an agricultural land- Rape Pests, ed. by Alford DV. Blackwell Science, Oxford, pp. 97–112 scape: the contribution of seminatural habitat islands. Basic Appl Ecol (2003). 4:129–138 (2003). 25 Veromann E, Williams IH, Kaasik R and Luik A, Potential of para- 3 Holland JM, Bianchi FJ, Entling MH, Moonen A-C, Smith BM and Jean- sitoids to control populations of the weevil Ceutorhynchus obstric- neret P, Structure, function and management of semi-natural habi- tus (Marsham) on winter oilseed rape. Int J Pest Manag 57:85–92 tats for conservation biological control: a review of European stud- (2011). ies. Pest Manag Sci 72:1638–1651 (2016). 26 Kovács G, Kaasik R, Metspalu L, Williams IH, Luik A and Veromann E, 4 Holland JM, Douma JC, Crowley L, James L, Kor L, Stevenson DRW et al., Could Brassica rapa, Brassica juncea and Sinapis alba facilitate the Semi-natural habitats support biological control, pollination and soil control of the cabbage seed weevil in oilseed rape crops? Biol Control conservation in Europe. A review. Agron Sustain Dev 37:31 (2017). 65:124–129 (2013). 5 Tscharntke T, Karp DS, Chaplin-Kramer R, Batáry P, DeClerck F, Grat- 27 Kovács G, Kaasik R, Treier K, Luik A and Veromann E, Do different field ton C et al., When natural habitat fails to enhance biological pest bordering elements affect cabbage seed weevil damage and its control – five hypotheses. Biol Conserv 204:449–458 (2016). parasitism rate differently in winter oilseed rape? IOBC-WPRS Bull 6 Landis DA, Designing agricultural landscapes for biodiversity-based 116:75–80 (2016). ecosystem services. Basic Appl Ecol 18:1–12 (2017). 28 Kovács G, Kaasik R, Kaart T, Metspalu L, Luik A and Veromann E, In 7 Alford DV, Nilsson C and Ulber B, Insect pests of oilseed rape crops, in search of secondary plants to enhance the efficiency of cabbage Biocontrol of Oilseed Rape Pests, ed. by Alford DV. Blackwell Science, seed weevil management. BioControl 62:29–38 (2017). Oxford, pp. 9–42 (2003). 29 Ulber B, Williams IH, Klukowski Z, Luik A and Nilsson C, Parasitoids of 8 Williams IH, The major insect pests of oilseed rape in Europe and their oilseed rape pests in Europe: key species for conservation biocontrol, management: an overview, in Biocontrol-Based Integrated Manage- in Biocontrol-Based Integrated Management of Oilseed Rape Pests, ed. ment of Oilseed Rape Pests, ed. by Williams IH. Springer, Dordrecht, by Williams IH. Springer, Dordrecht, pp. 45–76 (2010). pp. 1–43 (2010). 30 Murchie AK and Williams IH, A bibliography of the parasitoids of the 9 Kalischuk AR and Dosdall LM, Susceptibilities of seven Brassicaceae cabbage seed weevil (Ceutorhynchus assimilis Payk.). IOBC-WPRS Bull species to infestation by the cabbage seedpod weevil (Coleoptera: 21:163–169 (1998). Curculionidae). Can Entomol 136:265–276 (2004). 31 Williams IH and Ferguson AW, Spatio-temporal distributions of pests 10 Ulmer BJ and Dosdall LM, Glucosinolate profile and oviposition behav- and their parasitoids on the oilseed rape crop, in Biocontrol-Based ior in relation to the susceptibilities of Brassicaceae to the cabbage Integrated Management of Oilseed Rape Pests, ed. by Williams IH. seedpod weevil. Entomol Exp Appl 121:203–213 (2006). Springer, Dordrecht, pp. 245–271 (2010). 11 Ulmer BJ and Dosdall LM, Spring emergence biology of the cabbage 32 Dmoch J, Study on the parasites of the cabbage seed weevil (Ceuth- seedpod weevil (Coleoptera: Curculionidae). Ann Entomol Soc Am orrhynchus assimilis Payk.). I. Species composition and economic 99:64–69 (2006). importance of the larval ectoparasites. Rocz Nauk Rol E 5:99–112 12 Free JB and Williams IH, A survey of the damage caused to crops of oil-seed rape (Brassica napus L.) by insect pests in South–Central (1975). England and their effect on seed yield. J Agric Sci 90:417–424 (1978). 33 Murchie AK, Parasitoids of cabbage seed weevil and brassica pod 13 Lerin J, Assessment of yield loss in winter rape due to seedpod weevil midge in oilseed rape. PhD thesis. University of Keele (1996). (Ceutorhynchus assimilis Payk.) II. Yield loss in a cage experiment. 34 Graham MWR d V, The Pteromalidae of North-Western Europe Agronomie 4:147–154 (1984). (Hymenoptera: Chalcidoidea). Bull Br Mus Nat Hist Entomol 16:1–908 14 Buntin GD, Damage loss assessment and control of the cabbage (1969). seedpod weevil (Coleoptera: Curculionidae) in winter canola using 35 Gurr GM, van Emden HF and Wratten SD, Habitat manipulation and insecticides. J Econ Entomol 92:220–227 (1999). natural enemy efficiency: implications for the control of pests, in 15 Dmoch J, The dynamics of a population of the cabbage seedpod weevil Conservation Biological Control, ed. by Barbosa P. Academic Press, (Ceutorhynchus assimilis Payk.) and the development of winter rape. San Diego, CA, pp. 155–183 (1998). Part I. Ekol Pol Ser A 13:249–287 (1965). 36 Wäckers FL, Assessing the suitability of flowering herbs as parasitoid 16 Dosdall LM and Mason PG, Key pests and parasitoids of oilseed rape food sources: flower attractiveness and nectar accessibility. Biol or canola in North America and the importance of parasitoids in Control 29:307–314 (2004). integrated management, in Biocontrol-Based Inegrated Management 37 van Rijn PCJ and Wäckers FL, Nectar accessibility determines fitness, of Oilseed Rape Pests, ed. by Williams IH. Springer, Dordrecht, pp. flower choice and abundance of hoverflies that provide natural pest 167–213 (2010). control. J Appl Ecol 53:925–933 (2016). 17 Dosdall LM, Moisey D, Cárcamo H and Dunn R, Cabbage Seedpod 38 Thies C and Tscharntke T, Biological rape pest control in Weevil Fact Sheet, Agri-facts Practical Information for Alberta’s Agri- spatio-temporally changing landscapes, in Biocontrol-Based Inte- culture Industry. Revised January 2014 Agdex 622–21. [Online]. grated Management of Oilseed Rape Pests, ed. by Williams IH. (2001). Available: http://www1.agric.gov.ab.ca/$department/ Springer, Dordrecht, pp. 273–284 (2010). deptdocs.nsf/all/agdex2538/$file/622-21.pdf [12 December 2017]. 39 Rusch A, Valantin-Morison M, Sarthou J-P and Roger-Estrade J, 18 Heimbach U and Müller A, Incidence of pyrethroid-resistant oilseed Multi-scale effects of landscape complexity and crop manage- rape pests in Germany. Pest Manag Sci 69:209–216 (2013). ment on pollen beetle parasitism rate. Landsc Ecol 26:473–486 19 Zimmer CT, Köhler H and Nauen R, Baseline susceptibility and insec- (2011). ticide resistance monitoring in European populations of Meligethes 40 Rusch A, Valantin-Morison M, Roger-Estrade J and Sarthou J-P, Local aeneus and Ceutorhynchus assimilis collected in winter oilseed rape. and landscape determinants of pollen beetle abundance in overwin- Entomol Exp Appl 150:279–288 (2014). tering habitats. Agric For Entomol 14:37–47 (2012). wileyonlinelibrary.com/journal/ps © 2018 Society of Chemical Industry Pest Manag Sci (2018)

92 Landscape eﬀects on Ceutorhynchus obstrictus infestation and parasitism rates www.soci.org

41 Rusch A, Valantin-Morison M, Roger-Estrade J and Sarthou JP, Using 63 Lancashire PD, Bleiholder H, Boom TVD, Langelüddeke P, Stauss R, landscape indicators to predict high pest infestations and success- Weber E et al., A uniform decimal code for growth stages of crops ful natural pest control at the regional scale. Landsc Urban Plan and weeds. Ann Appl Biol 119:561–601 (1991). 105:62–73 (2012). 64 Dosdall LM, Ulmer BJ, Gibson GAP and Cárcamo HA, The 42 Skellern MP, Welham SJ, Watts NP and Cook SM, Meteorological and spatio-temporal distribution dynamics of the cabbage seed- landscape influences on pollen beetle immigration into oilseed rape pod weevil, Ceutorhynchus obstrictus (Coleoptera: Curculionidae), crops. Agric Ecosyst Environ 241:150–159 (2017). and its larval parasitoids in canola in western Canada. Biocontrol Sci 43 Skellern MP and Cook SM, Prospects for improved off-crop habi- Technol 16:987–1006 (2006). tat management for pollen beetle control in oilseed rape. 65 Trjapitzin VA, Hymenoptera III, 2. Chalcidoidea 18. Mymaridae, in Keys Arthropod–Plant Interact (in press). https://doi.org/10.1007/s11829- to the Insects of the European Part of the USSR, ed. by Medvedev 018-9598-9. GS. Academy of Sciences of the USSR, Institute of Zoology, Nauka, 44 Rusch A, Valantin-Morison M, Sarthou JP and Roger-Estrade J, Inte- Leningrad, pp. 516–537 (1978). grating crop and landscape management into new crop protection 66 Gibson GA, Gates MW and Buntin GD, Parasitoids (Hymenoptera: Chal- strategies to enhance biological control of oilseed rape pests, in cidoidea) of the cabbage seedpod weevil (Coleoptera: Curculion- Biocontrol-Based Integrated Management of Oilseed Rape Pests, ed. by idae) in Georgia, USA. J Hymenopt Res 15:187–207 (2006). Williams IH. Springer, Dordrecht, pp. 415–448 (2010). 67 Baur H, Muller FJ, Gibson GAP, Mason PG and Kuhlmann U, A review 45 Zaller JG, Moser D, Drapela T, Schmöger C and Frank T, Insect pests in of the species of Mesopolobus (Chalcidoidea: Pteromalidae) associ- winter oilseed rape affected by field and landscape characteristics. ated with Ceutorhynchus (Coleoptera: Curculionidae) host-species of Basic Appl Ecol 9:682–690 (2008). European origin. Bull Entomol Res 97:387 (2007). 46 Valantin-Morison M, Meynard J-M and Doré T, Effects of crop manage- 68 Muller FJ, Baur H, Gibson GAP, Mason PG and Kuhlmann U, Review ment and surrounding field environment on insect incidence of the species of Trichomalus (Chalcidoidea: Pteromalidae) associ- in organic winter oilseed rape (Brassica napus L.). Crop Prot ated with Ceutorhynchus (Coleoptera: Curculionidae) host species of 26:1108–1120 (2007). European origin. Can Entomol 139:643–657 (2007). 47 Roland J and Taylor PD, Insect parasitoid species respond to forest 69 Clark JS, Silman M, Kern R, Macklin E and HilleRisLambers J, Seed dis- structure at different spatial scales. Nature 386:710–713 (1997). persal near and far: patterns across temperate and tropical forests. 48 Tscharntke T, Klein AM, Kruess A, Steffan-Dewenter I and Thies C, Ecology 80:1475–1494 (1999). Landscape perspectives on agricultural intensification and biodiver- 70 Robinet C, Kehlenbeck H, Kriticos DJ, Baker RHA, Battisti A, Brunel S sity – ecosystem service management. Ecol Lett 8:857–874 (2005). et al., A suite of models to support the quantitative assessment of 49 Kruess A, Effects of landscape structure and habitat type on a spread in pest risk analysis. PLOS ONE 7:e43366 (2012). plant–herbivore–parasitoid community. Ecography 26:283–290 71 Miguet P, Fahrig L and Lavigne C, How to quantify a (2003). distance-dependent landscape effect on a biological response. 50 Kruess A and Tscharntke T, Habitat fragmentation, species loss, and Methods Ecol Evol 8:1717–1724 (2017). biological control. Sci New Ser 264:1581–1584 (1994). 72 Goedhart PW, Lof ME, Bianchi FJJA, Baveco HJM and van der 51 Taimr L, Sedivy J, Bergmannova E and Hanker I, Further experi- Werf W, Modelling mobile agent-based ecosystem services using ence obtained in studies on dispersal flights of Meligethes aeneus kernel-weighted predictors. Methods Ecol Evol 9:1241–1249 (2018). F., marked with P23 (Coleoptera). Acta Entomol Bohemoslovic https://doi.org/10.1111/2041-210X.12972. 64:325–332 (1967). 73 Zuur AF, Ieno EN, Walker NJ, Saveliev AA and Smith GM, Mixed Effects 52 Williams IH and Cook SM, Crop location by oilseed rape pests and Models and Extensions in Ecology with R. Springer, New York, NY host location by their parasitoids, in Biocontrol-Based Integrated (2009). Management of Oilseed Rape Pests, ed. by Williams IH. Springer, 74 Barton´ K, Multi-Model Inference. R package version 1.15.6. (2016). Dordrecht, pp. 215–244 (2010). 75 R Core Team, A Language and Environment for Statistical Computing.R 53 Estonian Environment Agency, Estonian Weather Service of the Estonian Foundation for Statistical Computing, Vienna (2016). Environment Agency. [Online]. Available: http://www.ilmateenistus 76 Hawkins BA and Cornell HV, Maximum parasitism rates and successful .ee/kliima/kliimanormid/ohutemperatuur/?lang=en#keskm_min_ biological control. Science 266:1886 (1994). %C3%B5hut [24 May 2017]. 77 Juhel AS, Barbu CM, Franck P, Roger-Estrade J, Butier A, Bazot M 54 Estonian Environment Agency, Meteorological Yearbook of Estonia et al., Characterization of the pollen beetle, Brassicogethes aeneus, 2014. [Online]. Estonian Environment Agency, Tallinn (2015) Avail- dispersal from woodlands to winter oilseed rape fields. PLoS One able: http://www.ilmateenistus.ee/wp-content/uploads/2016/02/ 12:e0183878 (2017). aastaraamat_2014.pdf [27 June 2018]. 78 Tylianakis JM, Didham RK and Wratten SD, Improved fitness of aphid 55 QuESSA Europe, Quantification of Ecological Services for Sustainable parasitoids receiving resource subsidies. Ecology 85:658–666 Agriculture, 2013–2017. [Online]. Available: http://www.quessa.eu/ (2004). [13 November 2017]. 79 Lavandero B, Wratten SD, Didham RK and Gurr G, Increasing floral 56 Holland JM, Jeanneret P, Herzog F, Moonen AC, Rossing W, van der diversity for selective enhancement of biological control agents: a Werf W et al., Project: quantification of ecological services for sus- double-edged sward? Basic Appl Ecol 7:236–243 (2006). tainable agriculture. IOBC-WPRS Bull 100:55–58 (2014). 80 Holland JM, Oaten H, Moreby S, Birkett T, Simper J, Southway S 57 Picchi MS, Bocci G, Petacchi R and Entling MH, Effects of local and et al., Agri-environment scheme enhancing ecosystem services: a landscape factors on spiders and olive fruit flies. Agric Ecosyst Environ demonstration of improved biological control in cereal crops. Agric 222:138–147 (2016). Ecosyst Environ 155:147–152 (2012). 58 Sutter L, Albrecht M and Jeanneret P, Landscape greening and local 81 Griffiths GJK, Holland JM, Bailey A and Thomas MB, Efficacy and creation of wildflower strips and hedgerows promote multiple economics of shelter habitats for conservation biological control. ecosystem services. J Appl Ecol 55:612–620 (2018). Biol Control 45:200–209 (2008). 59 Pfister SC, Sutter L, Albrecht M, Marini S, Schirmel J and Entling 82 Midega CAO, Jonsson M, Khan ZR and Ekbom B, Effects of landscape MH, Positive effects of local and landscape features on preda- complexity and habitat management on stemborer colonization, tory flies in European agricultural landscapes. Agric Ecosyst Environ parasitism and damage to maize. Agric Ecosyst Environ 188:289–293 239:283–292 (2017). (2014). 60 Fusser MS, Pfister SC, Entling MH and Schirmel J, Effects of field margin 83 Bhar R and Fahrig L, Local vs. landscape effects of woody field borders type and landscape composition on predatory carabids and slugs in as barriers to crop pest movement. Conserv Ecol 2:3 (1998). wheat fields. Agric Ecosyst Environ 247:182–188 (2017). 84 Ulber B and Vidal S, Influence of host density and host distribution 61 PRIA, Interactive map of the Estonian Agricultural Registers and Infor- on parasitism of Ceutorhynchus assimilis by Trichomalus perfectus. mation Board. [Online]. Available: https://kls.pria.ee/kaart/ [24 May IOBC-WPRS Bull 21:185–195 (1998). 2017]. 85 Ferguson AW, Klukowski Z, Walczak B, Perry JN, Mugglestone MA, Clark 62 Estonian Land Board, Web Map of the Estonian Land Board. SJ et al., The spatio-temporal distribution of adult Ceutorhynchus [Online]. Available: http://xgis.maaamet.ee/maps/XGis?app_ assimilis in a crop of winter oilseed rape in relation to the distribution id=UU82A&user_id=at&LANG=2&WIDTH=1304& of their larvae and that of the parasitoid Trichomalus perfectus. amp;HEIGHT=797&zlevel=0,552500,6505000 [24 May 2017]. Entomol Exp Appl 95:161–171 (2000).

93 SUPPORTING INFORMATION

Eﬀects of land use on infestation and parasitism rates of cabbage seed weevil in oilseed rape Gabriella Kovács a*, Riina Kaasik a, Marjolein E Lof b, Wopke van der Werf b, Tanel Kaart c, John M Holland d, Anne Luik a, Eve Veromann a a Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Fr. R. Kreutzwaldi 5, 51006 Tartu, Estonia b Centre for Crop Systems Analysis, Department of Plant Sciences, Wageningen University and Research, P.O. Box 430, 6700 AK, Wageningen, The Netherlands c Institute of Veterinary Medicine and Animal Sciences, Estonian University of Life Sciences, Fr. R. Kreutzwaldi 62, 51006 Tartu, Estonia d Game & Wildlife Conservation Trust, Fordingbridge, Hampshire, SP6 1EF, United Kingdom * Corresponding author: Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Fr. R. Kreutzwaldi 5, 51006 Tartu, Estonia e-mail: [email protected]

CONTENTS: Table S1 Mapped habitat categories Table S2 Crop management Figure S1 Cross-section of 2-dimensional t-distribution for various values of the scale parameter u Figure S2 Greyscale version of Figure 1

Table S1 Mapped habitat categories Semi-natural habitat (SNH) Crops (continued) WA: natural or semi-natural woody areal Perennial herbaceous crops elements Permanent grassland (NOT SNH) WL: woody linear elements Perennial woody crops HA: herbaceous areal elements HL: herbaceous linear elements Apples (Malus spp.) FA: Temporary in-field SNH Sea-buckthorn (Hippophae rhamnoides) Crops Currant – red/black / goosberry (Ribes rubrum, R. nigrum, R. uva-crispa) Annual herbaceous crops Urban areas Wheat (Triticum aestivum & associated sp) Barley (Hordeum sativum) Urban area, % of green area in element <25% Oats (Avena sativa) Urban area, % of green area in element 26 to Rye (Secale cereale) 50% Potato (Solanum tuberosum) Urban area, % of green area in element 51 to Field beans (Vicia faba) 75% Peas (all types) (Pisum spp) Urban area, % of green area in element >75% Maize (Zea mays) Water courses Oilseed rape (Brassica hybrid) (e.g. rivers, streams, canals, ditches >1.5m wide) Commercial horticulture Non-Habitat Medicago sativa (e.g. roads, lakes) Rotational grassland < 5 years old Buckwheat (Fagorpyrum esculentum) Purple tansy (Phacelia tanacetifolia) Jerusalem artichoke (Helianthus tuberosus) Turfgrass sod (Bluegrass Poa pratensis ca 40%, Red Fescue Festuca rubra spp. ca 60% mixture)

94 , ‡ ) § date of application, growth stage date of application, growth code of the crop no insecticides no insecticides no insecticides Proteus OD, 30.04.2014, 53-56 Proteus OD, 28.04.2014, 53-55 Proteus OD, 23.04.2014, 30-32 Proteus OD, 23.04.2014, 30-32 Proteus OD, 24.04.2014, 30-32 Proteus OD, 29.04.2014, 52-55 Proteus OD, 23.04.2014, 30-32 Proteus OD, 23.04.2014, 30-32 Proteus OD, 23.04.2014, 30-32 Proteus OD, 23.04.2014, 30-32 Proteus OD, 23.04.2014, 30-32 Proteus OD, 29.04.2014, 55-59 Proteus OD, 29.04.2014, 55-57 Proteus OD, 23.04.2014, 30-32 Proteus OD, 23.04.2014, 30-32 Proteus Insecticide application (product Insecticide synthetic synthetic/organic (pig slurry) synthetic synthetic synthetic synthetic synthetic/organic (pig slurry) synthetic synthetic synthetic synthetic synthetic synthetic synthetic synthetic synthetic/organic (slurry) synthetic synthetic Type of fertilizer Type 229 354 229 156 331 229 242 242 229 156 233.3 233.3 233.3 233.3 233.3 233.3 233.3 233.3 (kg/ha) Total amount Total of N fertilizer (fungicide or insecticide) both both both both both fungicide fungicide fungicide fungicide fungicide both both fungicide fungicide fungicide no seed dressing no seed dressing no seed dressing Seed dressings dressings Seed inversion inversion minimum non-inversion tillage: disc harrow minimum non-inversion tillage: disc harrow minimum non-inversion tillage: disc harrow inversion minimum non-inversion tillage: stubble cultivation minimum non-inversion tillage: cultivation minimum non-inversion tillage: cultivation inversion minimum non-inversion inversion minimum non-inversion tillage: cultivation minimum non-inversion tillage: cultivation inversion inversion inversion inversion tillage Type of tillage Type Visby Sherpa Thorin Thorin Thorin Visby Rohan Rohan Rohan Rohan Thorin Visby Rohan Rohan Abakus Rohan Rohan Rohan Crop Crop variety 6 9.7 6.4 12. 4.5 9.5 26.8 41.3 20.6 12.4 11.5 73.8 43.7 16.6 37.8 52.3 52.7 10.5 Size Size (ha) 1 5 7 6 9 2 3 4 8 11 12 13 10 14 16 17 15 18 Field Field ID † According to the BBCH-scale of oilseed rape (Lancashire et al., 1991) - principal growth stages 3 (stem elongation) and 5 (inflorescence emergence) stages 3 (stem elongation) and 5 (inflorescence et al., 1991) - principal growth to the BBCH-scale of oilseed rape (Lancashire According Abbreviations: HL – herbaceous linear semi-natural habitat, WL – woody linear semi-natural habitat HL – herbaceous linear semi-natural habitat, Abbreviations: (pyrethroid) Thiacloprid (neonicotinoid) and Deltamethrin OD – Proteus ingredients: Active Adjacent Adjacent element HL HL HL HL HL HL WL WL WL WL WL WL another crop another crop another crop another crop another crop another crop Crop management S2 Crop Table † ‡ § Reference: Lancashire, P.D., Bleiholder, H., Boom, T.V.D., Langelüddeke, P., Stauss, R., Weber, E., Witzenberger, A., 1991. A uniform decimal code for growth growth for code decimal uniform A 1991. A., Witzenberger, E., Weber, R., Stauss, P., Langelüddeke, T.V.D., Boom, H., Bleiholder, P.D., Lancashire, Reference: 119, 561–601. doi:10.1111/j.1744-7348.1991.tb04895.x Biol. Ann. Appl. and weeds. stages of crops

FigureFigure S1 Cross-sectionS1 Cross-section of 2-dimensional of 2-dimensional t-distribution, t-distribution, with with shape shape parameter parameter = = 25, 25, and and scale scale parametersparameters ranging ranging from from 200 200 m inm blackin black (the (the optimal optimal scale scale of ofthe the parasitism parasitism model) model) to to 600 600 m min red in(the red optimal (the optimal scale ofscale the of infestation the infestation model). model). It shows It shows that for that larger for larger values values of the of scale the factorscale (>300factor m) habitats(>300 m )at habitats larger distancesat larger distances than 1 km than from 1 kmthe fromobservation the observation point also point have also an have eﬀect an on the effectecosystem on the service ecosystem (or disservice) service (or disservice)

FigureFFigure igureS2 Greyscale S2 S2 Greyscale Greyscale version version of Figureversion of Figure 1 a) of Map 1 Figurea) ofMap all of1.habitats aa)ll habitatsMap in one of in ofall one the habitats of 18 the landscape 18 inlandscape one circles of circles thewith 181with km landscape 1radius km radius around circlesaround the sampling thewith sampling 1 point point (marked(markedkm with radius X)with of around X)the of focal the thefocalfield sampling (fieldwith (crosshatch);with crosshatch);point b) (marked predicted b) predicted withparasitism parasitismX) of rates the (fromrates focal (from white field whiteto (withblack, to black, crosshatch);highest highest to lowest) to b)lowest) of predicted Ceutorhynchus5 of Ceutorhynchus obstrictusobstrictus for all for agricultural all agricultural fields fieldsbased based on the on spatial the spatial configuration configuration of herbaceous of herbaceous linear linear semi- naturalsemi-natural habitats habitats (‘Herb. (‘Herb. linear’, linear’, black blackin Fig. in 1a) Fig. 1a) and permanentandparasitism permanent grasslands ratesgrasslands (‘Grassland’, (from (‘Grassland’, white grey to with grey black, diagonal with diagonalhighest stripes stripes fromto lowest) thefrom upper the of upperleft Ceutorhynchus to left the tolower the lower right inrightobstrictus Fig. in 1a) Fig. in 1a) thefor ind allepicted the agricultural depicted landscape landscape circle circle fields based on the spatial configuration of herbaceous linear semi-natural habitats ('Herb. linear' black in Fig. 1a) and permanent grasslands ('Grassland', grey with diagonal stripes from the upper left to the lower right in Fig. 1a) in the depicted landscape circle

6 6 96 CURRICULUM VITAE

First name: Gabriella Last name: Kovács Date of birth: August 15th, 1986 Address: Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences (EULS), Fr. R. Kreutzwaldi 5, Tartu E-mail: [email protected]

Professional employment: Since 2014 Chair of Plant Health, Institute of Agricultural and Environmental Sciences, EULS, Specialist (1.00) 2010–2014 Department of Plant Protection, Institute of Agricultural and Environmental Sciences, EULS, Specialist (0.50)

Education: 2010–2018 PhD studies in agricultural sciences, EULS 2005–2010 Five-year university studies in agricultural engineer- ing in environmental management (organic farming, plant protection), Szent István University, Hungary 2009–2010 Foreign visiting student (fall semester), EULS

Academic Degree: 2010 Diploma (MSc), Agricultural Engineer in Envi- ronmental Management, Szent István University, Hungary

Research interests: Agricultural entomology, natural enemies, ecosystem services, sustainable plant protection

Language skills: Hungarian – native speaker; English – proficient user; Estonian – independent user

97 Honours & awards: 2013 Young researcher’s award, Research Centre of Organic Farming, EULS 2011–2013 SA Archimedes DoRa activity 8 scholarship -to attend a conference in 2013(2x), 2012, 2011 -to visit a laboratory abroad in 2012 -to attend a course in 2011

Participation in research projects: 2016–2019 8T160091PKTK ”Novel biosafe IPM strategies to manage pesticide resistance in pollen beetles”, senior research staff 2013–2017 European Union 7th Framework Programme, project no. 8-2/T12156PKTK ”Quantification of ecological services for sustainable agriculture”, senior research staff 2015−2016 8-2/T15129PKTK ”Promotion of recent advances in research towards improved integrated crop protection in oilseed production through the Interna- tional Organisation for Biological and Integrated Control (IOBC)”, senior research staff 2011−2014 Estonian Science Foundation Grant no. 8895 ”Impact of host plants on the major pests of cruciferous plants and their parasitoids in different cropping systems”, senior research staff

Professional training: 2018 PhD course “Multivariate Data Analyses Using PC- ORD“, EULS 2012 MSc/PhD course “Chemical ecology 2”, University of Tartu (UT) 2012 BOVA MSc course “Causes and consequences of pollinator decline”, Open University (EULS) 2011 MSc/PhD course “Chemical ecology”, UT 2011 BeeNOVA PhD course “Communication in insects”, University of Helsinki, Finland 2011 Erasmus intensive programme “EPSEN – Ecological production systems for environmental and human health”, Slovak University of Agriculture in Nitra, Slovakia

98 2011 PhD course “English for specific purposes. Language structures in oral and written expression”, EULS 2011−2012 PhD curriculum including: Copyright and legal protection of intellectual property, Methodology of research, Academic writing and presentation, Mathematical statistics and odelling, Philosophy of science, Learning and teaching in higher eduation, Practice learning in univerity teaching, Enomology; EULS/UT

99 ELULOOKIRJELDUS

Eesnimi: Gabriella Perenimi: Kovács Sünniaeg: 15. august 1986 Aadress: Põllumajandus- ja keskkonnainstituut, Eesti Maaüli- kool (EMÜ), Fr. R. Kreutzwaldi 5, Tartu E-post [email protected]

Teenistuskäik: Alates 2014 Taimetervise õppetool, põllumajandus- ja keskkonnainstituut, EMÜ, spetsialist (1.00) 2010–2014 Taimekaitse osakond, põllumajandus- ja keskkonnainstituut, EMÜ, spetsialist (0.50)

Haridustee: 2010–2018 Doktoriõpe põllumajanduse erialal, EMÜ 2005–2010 Magistriõpe põllumajandusinsener keskkonnakor- ralduse erialal (mahepõllumajandus, taimekaitse), Szent István Ülikool, Ungari 2009–2010 Väliskülalisüliõpilane (sügissemester), EMÜ

Teaduskraad: 2010 Diplom (MSc), põllumajandusinsener keskkonna- korralduse erialal, Szent István Ülikool, Ungari

Teadustöö põhisuunad: entomoloogia, kahjurite looduslikud vaenlased, ökosüsteemiteenused, jätkusuutlik taimekaitse

Keeleoskus: Ungari – emakeel, inglise – vilunud keelekasutaja, eesti – iseseisev keelekasutaja

Teaduspreemiad ja stipendiumid: 2013 Eesti Maaülikooli Mahekeskuse noorteadlase pree- mia

100 2011–2013 SA Archimedes DoRa T8 stipendium - konverentsil osalemiseks, 2013(2x), 2012, 2011 - enesetäiendamiseks välislaboris, 2012 - kursusel osalemiseks, 2011

Uurimisprojektides osalemine: 2016–2019 8T160091PKTK „Integreeritud taimekaitse hiilamardika tõrjeks“, põhitäitja 2013–2017 8-2/T12156PKTK „Ökosüsteemi teenuste roll jät- kusuutlikus põllumajanduses“, põhitäitja 2015−2016 8-2/T15129PKTK „Õlikultuuride integreeritud taimekaitsestrateegiate väljatöötamise ja uurimise edendamine rahvusvahelise võrgustiku IOBC (Inter- national Organisation for Biological and Integrated Control) kaudu“, põhitäitja 2011−2014 ETF8895 „Peremeestaimede mõju ristõieliste õli- kultuuride võtmekahjuritele ja nende parasitoididele erinevates viljelustingimustes“, põhitäitja

Enesetäiendus ja koolitused: 2018 Mitme muutujaga andmeanalüüs kasutades PC-ORDi, EMÜ 2012 Keemiline ökoloogia 2, Tartu Ülikool (TÜ) 2012 Tolmeldajate arvukuse vähenemise põhjused ja taga- järjed, BOVA kraadiõppurite kursus, Avatud ülikool (EMÜ) 2011 Keemiline ökoloogia, TÜ 2011 Putukate kommunikeerumine, BeeNOVA, Helsingi Ülikool, Soome 2011 Ökoloogilised tootmissüsteemid keskkonna- ja inim- tervise jaoks (EPSEN), Erasmuse intensiivne kursus, Slovakkia Põllumajandusülikool Nitras, Slovakkia 2011 Inglise erialakeel. Keelestruktuurid erialase teksti kirjutamises ja suulises väljenduses, EMÜ 2011−2012 Doktorantide üld- ja erialaõpe: intellektuaalomandi kaitse ja autoriõigus, teadustöö metodoloogia, akadeemiline kirjutamine ja esitlus, matemaatiline statistika ja modelleerimine, teadusfilosoofia, õppi- mine ja õpetamine kõrgkoolis, ülikooli pedagoogi- line praktika, entomoloogia; EMÜ/TÜ

101 LIST OF PUBLICATIONS

Scholarly articles indexed by Web of Science (1.1.)

Kovács, G.; Kaasik, R.; Lof, M.; van der Werf, W.; Kaart, T.; Holland, J.H.; Luik, A.; Veromann, E. 2018. Effects of land use on infestation and parasitism rates of cabbage seed weevil in oilseed rape. Pest Management Science DOI: https://doi.org/10.1002/ps.5161 Kovács, G.; Kaasik, R.; Kaart, T.; Metspalu, L.; Luik, A.; Veromann, E. (2017). In search of secondary plants to enhance the efficiency of cabbage seed weevil management. Biocontrol, 62, 29−38. Metspalu, L.; Veromann, E.; Kaasik, R.; Kovács, G.; Williams, I. H.; Mänd, M. (2015). Comparison of sampling methods for estimating the abundance of Meligethes aeneus on oilseed crops. International Journal of Pest Management, 61, 312−319. Veromann, E.; Kaasik, R.; Kovács, G.; Metspalu, L.; Williams, I.H.; Mänd, M. (2014). Fatal attraction: search for a dead-end trap crop for the pollen beetle (Meligethes aeneus). Arthropod-Plant Interac- tions, 8, 373−381. Kaasik, R.; Kovács, G.; Kaart, T.; Metspalu, L.; Williams, I.H.; Vero- mann, E. (2014). 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. Biological control, 69, 65−71. Kaasik, R.; Kovács, G.; Toome, M.; Metspalu, L.; Veromann, E. (2014). The relative attractiveness of Brassica napus, B. rapa, B. juncea and Sinapis alba to pollen beetles. Biocontrol, 59, 19−28. Kovács, G.; Kaasik, R.; Metspalu, L.; Williams, I. H.; Luik, A.; Vero- mann, E. (2013). Could Brassica rapa, Brassica juncea and Sinapis alba facilitate the control of the cabbage seed weevil in oilseed rape crops? Biological control, 65, 124−129. Veromann, E.; Toome, M.; Kännaste, A.; Kaasik, R.; Copolovici, L.; Flink, J.; Kovács, G.; Narits, L.; Luik, A.; Niinemets, Ü. (2013). Effects of nitrogen fertilization on insect pests, their parasitoids, plant diseases and volatile organic compounds in Brassica napus. Crop Protection, 43, 79−88.

102 Veromann, E.; Metspalu, L.; Williams, I.H.; Hiiesaar, K.; Mand, M.; Kaasik, R.; Kovács, G.; Jogar, K.; Svilponis, E.; Kivimagi, I.; Ploomi, A.; Luik, A. (2012). Relative attractiveness of Brassica napus, Brassica nigra, Eruca sativa and Raphanus sativus for pollen beetle (Meligethes aeneus) and their potential for use in trap cropping. Arthropod-Plant Interactions, 6 (3), 385−394.

Articles published in conference proceedings (3.4.)

Treier, K.; Kovács, G.; Kaasik, R.; Männiste, M.; Veromann, E. (2017). The abundance of overwintered predatory arthropods in agricultural landscape elements. IOBC-WPRS Bulletin, 122, 68−73. Kovács, G.; Kaasik, R.; Treier, K.; Luik, A.; Veromann, E. (2016). Do different field bordering elements affect cabbage seed weevil damage and its parasitism rate differently in winter oilseed rape? IOBC- WPRS Bulletin, 116, 75−80. Lof, M. E.; Veromann, E.; Kovács, G.; Kaasik, R.; van der Werf, W. (2016). Modelling ecosystem services in landscapes. 1−1. Kovács, G.; Kaasik, R.; Luik, A.; Veromann, E. (2014). The expanding oilseed rape insect pest community in Estonia. IOBC-WPRS Bul- letin, 104, 25−28. Kaasik, R.; Kovács, G.; Mölder, J.; Treier, K.; Vaino, L.; Veromann, E. (2014). The impact of semi-natural habitats on the abundance of pollen beetle adults on winter oilseed rape fields. IOBC-WPRS Bul- letin, 104, 85−89. Kovács, G.; Kaasik, R.; Metspalu, L.; Veromann, E. (2013). The attractiveness of wild cruciferous plants on the key parasitoids of Meligethes aeneus. IOBC-WPRS Bulletin, 96, 81−92. Kovács, G.; Kaasik, R.; Veromann, E. (2013). The impact of companion planting on the abundance of Lepidopteran pests on white cabbage. Future IPM in Europe (on USB stick): Future IPM in Europe, 19- 21.03.2013, Riva del Garda, Italy.

103 Articles published in local conference proceedings (3.5.)

Veromann, E.; Sosare, K.; Kovács, G.; Kaasik, K. (2017). Põllumajan- dusmaastiku elementide tähtsus parasitoidide talvitumispaikadena. Mahepõllumajanduse arengusuunad – teaduselt mahepõllumajan- dusele, 154−159. Veromann, E.; Rannamäe, K.; Kaasik, R.; Kovács, G.; Metspalu, L. (2016). Seltsilistaimede mõju kapsakoi (Plutella xylostella) arvuku- sele ja parasiteerituse tasemele valgel peakapsal. Eesti Taimekaitse 95, 25−30. Kaasik, R.; Kovács, G.; Veromann, E. (2013). Kahjurite ja parasitoidide talvitumine rapsipõllul ja selle serva-aladel. Agronoomia 2013: Agro- noomia 2013, 144−149. Kaasik, R.; Kovács, G.; Veromann, E. (2012). Kahjurite arvukuse hinda- mise tähtsus suurkapsaliblika näitel. Agronoomia 2012, 129−134. Kovács, G.; Kaasik, R.; Luik, A.; Veromann, E. (2012). Kivikilbik mee- litab kapsakoid. Mahepõllumajanduse arengusuunad – teaduselt mahepõllumajandusele, 50−52. Kovács, G.; Kaasik, R.; Veromann, E. (2012). Kõdrasääsk (Dasineura brassicae) - uus kahjustaja rapsil. Agronoomia 2012, 141−144. Kaasik, R.; Kovács, G.; Luik, A.; Veromann, E. (2012). Seltsilistaimed kapsaliblikate mõjutajatena. Mahepõllumajanduse arengusuunad – teaduselt mahepõllumajandusele, 38−40. Veromann, E.; Kaasik, R.; Kovács, G.; Luik, A. (2012). Till aitab kap- sast öölase eest peita. Mahepõllumajanduse arengusuunad – teaduselt mahepõllumajandusele, 92−94. Kaasik, R., Kovács, G., Luik, A., Veromann, E. (2011). Seltsilistaimede mõju entomofauna mitmekesisusele valgel peakapsal. Agronoomia 2010/2011, 165−172.

104

VIIS VIIMAST KAITSMIST

KAAREL SOOTS POST-HARVEST PROCESSING TECHNOLOGY FOR CULTIVATED BERRIES KULTUURMARJADE KORISTUSJÄRGSE TÖÖTLEMISE TEHNOLOOGIA Professor Jüri Olt 6. aprill 2018

ANDRES JÄÄRATS THE EFFECT OF PLANTING STOCK AND SOIL SCARIFICATION ON FOREST REGENERATION ISTUTUSMATERJALI JA MAAPINNA ETTEVALMISTAMISE MÕJU METSA UUNEDAMISELE Emeriitdotsent Heino Seemen, dotsent Ivar Sibul, vanemteadur Arvo Tullus (Tartu Ülikool) 1. juuni 2018

KATRIN KALDRE INVASIVE NON-INDIGENOUS CRAYFISH SPECIES AS A THREAT TO THE NOBLE CRAYFISH (ASTACUS ASTACUS L.) POPULATIONS IN ESTONIA INVASIIVSED VÄHI VÕÕRLIIGID JA NENDE OHUSTAV MÕJU JÕEVÄHI (ASTACUS ASTACUS L.) ASURKONDADELE EESTIS Emeriitprofessor Tiit Paaver, professor Riho Gross 15. juuni 2018

PILLE TOMSON ROLE OF HISTORICAL SLASH AND BURN CULTIVATION IN THE DEVELOPMENT OF CULTURAL LANDSCAPES AND FOREST VEGETATION IN SOUTH ESTONIA AJALOOLISE ALEPÕLLUNDUSE ROLL LÕUNA-EESTI MAASTIKE JA METSATAIMESTIKU KUJUNEMISEL Professor Robert Gerald Henry Bunce, professor Kalev Sepp 24. oktoober 2018

MARI TILK GROUND VEGETATION DIVERSITY AND GEOBOTANICAL ANALYSIS IN THE SOUTH-WEST ESTONIAN DUNE PINE FORESTS ALUSTAIMESTIKU MITMEKESISUS JA GEOBOTAATILINE ANALÜÜS EDELA-EESTI LUITEMÄNNIKUTES Vanemteadur Katri Ots, juhtivteadur Malle Mandre, teadur Tea Tullus 12. detsember 2018

ISSN 2382-7076 ISBN 978-9949-629-52-7 (trükis) ISBN 978-9949-629-53-4 (pdf)