- Email: [email protected]
Uncoupling direct and indirect plant defences: Novel opportunities for improving crop security in willow plantations
Agriculture, Ecosystems and Environment 139 (2010) 528–533
Contents lists available at ScienceDirect
Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee
Uncoupling direct and indirect plant defences: Novel opportunities for improving crop security in willow plantations Johan A. Stenberg ∗ , Anna Lehrman, Christer Björkman Swedish University of Agricultural Sciences, Department of Ecology, P.O. Box 7044, SE-75007 Uppsala, Sweden
a r t i c l e
i n f o
Article history: Received 28 May 2010 Received in revised form 19 August 2010 Accepted 22 September 2010 Available online 16 October 2010 Keywords: Bioenergy crop Cropping security Indirect defence Optimal defence Defence syndromes Intraguild predation
a b s t r a c t Increased cropping security, i.e. minimized risk for detrimental stress events and hence improved yield stability, may be achieved by selecting resistant plant genotypes. However, strong direct defences against herbivores have been associated with negative effects on the natural enemies, resulting in weak indirect defences through these plant “bodyguards”. We compared the preference and performance of the most detrimental herbivore, the leaf beetle Phratora vulgatissima, and a biocontrol agent, the predatory bug Anthocoris nemorum, on four willow (Salix) genotypes used in short rotation coppicing. The biocontrol agent is omnivorous and survives on the plant in the absence of prey, without causing any apparent damage to the plant. Two of the genotypes intrinsically attracted and supported the bodyguard, and were at the same time partially capable of deterring and resisting the herbivore. In contrast, the third willow clone exhibited very limited direct and indirect resistance, while the fourth clone displayed intermediate levels of both. Thus, direct and indirect defence against herbivores need not be traded off in this system, but may be intertwined, resulting in “super plants” with higher probability of resisting detrimental herbivores, opening up novel possibilities for increased cropping security, without the use of insecticides. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Herbivores are generally more prone to outbreak in monocultures than in natural systems (Dalin et al., 2009), and may cause severe losses, especially in low-value or organic crops where insecticides are not used. In order to reduce herbivore damage, it is desirable to develop plant genotypes that are intrinsically more resistant to herbivores (direct defence) or that support biological control agents (indirect defence). Intertwining the two defence strategies would probably lead to a more sustainable resistance in the focal plant. Discouraging enough, ecologists have long presumed that there should be an intrinsic trade-off between direct and indirect defences in plants, since both require resources derived from the plant’s metabolic budget (see the Optimal Defence Hypothesis, e.g. Price et al., 1980; Ode et al., 2004). If, however, one disregards examples involving herbivore-mediated effects of plant quality on parasitoids, the literature is rather scarce, and evidence for tradeoffs between direct and indirect defence is mainly limited to plants
∗ Corresponding author. Tel.: +46 18 67 23 67. E-mail addresses: [email protected], [email protected] (J.A. Stenberg), [email protected] (A. Lehrman), [email protected] (C. Björkman). 0167-8809/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.agee.2010.09.013
that produce extra-floral nectar (e.g. Rehr et al., 1973; Rudgers et al., 2004). Trade-offs may also arise when plant traits important for attracting and supporting bodyguards also are favourable for herbivores. For example, many bodyguards are omnivorous and receive important nutrients from the same green plant tissue that the herbivores utilize (Coll, 1996; McMurtry and Croft, 1997; Agrawal et al., 1999; Eubanks and Denno, 2000; Coll and Guershon, 2002; RodriguezSaona and Thaler, 2005; Groenteman et al., 2006). However, the relation between pest insect vs. bodyguard responses to plant genotype has almost entirely been overlooked. Agrawal et al. (1999) and Eubanks and Denno (2000) simultaneously studied the response of herbivores and predators to their shared host-plant and suggested that bodyguards and pest herbivores should respond similarly to shared plant resources (Agrawal et al., 1999; Eubanks and Denno, 2000). We feel, however, that their suggestion may apply more to their specific systems (i.e. cotton and lima bean) rather than having universal significance. Here, we hypothesize that direct and indirect defences may be entangled in more complicated ways. If, for example, the nutritional requirements of pest herbivores and omnivorous bodyguards are different this may open up a possibility for breeders to disassociate the two defence strategies and optimise both of them independently. Rather than looking at direct and indirect defence as two contrasting strategies being subject to trade-off between each other we propose that these two defences are two indispensable aspects
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
of the system, which together underpin an integrated and balanced defence package. Based on empirical evidence and logical reasoning, we argue that in perennial crops it is wise to balance the investment in direct and indirect defences in such a way that they co-vary. In this paper we describe a study of the direct and indirect defences of different Salix (willow) clones against the blue willow beetle, Phratora vulgatissima (Coleoptera: Chrysomelidae), which is a severe pest in European coppicing willow plantations (e.g. Bell et al., 2006). Phratora damage may, however, be reduced by predatory omnivorous bugs such as Anthocoris nemorum (Heteroptera: Anthocoridae), which is considered to be an important bodyguard both for Salix and other plants (e.g. Björkman et al., 2003, 2004). In past breeding programs, the focus has been on direct plant resistance, since existing hypotheses regarding trade-offs have precluded the possibility of simultaneously optimizing direct and indirect defences (Kelly and Curry, 1991; Glynn et al., 2004). Here, we investigated bodyguard attraction to, and fitness on, different clones of Salix, in order to obtain reliable estimates of the clones’ levels of indirect defence. Simultaneously, herbivore attraction to, and reproduction on, the Salix clones were investigated. 2. Materials and methods 2.1. Study species 2.1.1. Plants Willows are grown in monocultures for biomass production in which the use of pesticides is not defendable neither from environmental, nor economic reasons. Due to the frequent hybridization and polyploidization in the genus Salix, variation within and among species may be high and overlapping; this opens up good opportunities to find genotypes with traits suitable for various conditions. It also means that the species concept is problematic – thus clone is often a more relevant identity than species. This taxonomical difficulty is less of a problem in practical bioenergy forestry. Here we chose to work with four clones of Salix, two of which are available from Lantmännen Agroenergi AB: Gudrun and Loden (S. dasyclados), and two breeding clones: 78-0-183 and 78-0-21 (S. viminalis). The specific clones used here were selected as they display phenotypic variation in traits relevant to Phratora and, thus, seem to represent “extremes” in a more or less continuous variation. In addition there is ongoing and existing research on these clones in sister projects. 2.1.2. Herbivore Phratora vulgatissima L. is the most important defoliator of Salix in Europe and has a broad distribution across Europe and Asia. It is restricted to Salix and both adults and larvae skeletonize the leaves (Peacock and Herrick, 2000; Peacock et al., 2001). Overwintering takes place outside Salix plantations in vertical objects that can provide shelter, for example reeds or trees with ageing bark (Björkman and Eklund, 2006). The overwintering adults emerge in April and feed on the young Salix leaves for about two weeks before mating and subsequent egg-laying. Larvae feed on the leaves for about 18–30 days, depending on leaf quality and temperature, and then they pupate in the ground. Adults emerge after about two weeks and feed on the leaves before hibernation. Generally, Phratora has only one generation per year in Sweden. For this study, we collected Phratora individuals from known overwintering sites near Uppsala, Sweden, during late February and kept them in a refrigerator at 5 ◦ C until the beginning of the experiment. The omnivorous predatory bug Anthocoris nemorum L. is widely distributed across Europe and Asia. It can be found in large numbers not only on Salix, but also on, among other plants, apple
529
trees and nettles. Anthocoris overwinters as an adult and becomes active as soon as the temperature rises in the spring. First instar nymphs have been observed sucking plant fluid from leaves and twig bases, which indicates that vegetable matter may be a very important food source for nymphal survival (personal observation). It is, however, very important to note that although Anthocoris may be nourished by Salix, and in extreme situations even completes its development without animal food, previous studies have shown that plant-feeding Anthocoris do not reduce plant fitness (Lauenstein, 1979). Thus, any direct cost of anthocorid feeding on Salix should be incommensurably lower than the cost of Phratora grazing, which may reduce Salix growth with up to 40% (Björkman et al., 2000a). Older anthocorid individuals are predominantly carnivores. Anthocoris is an active and aggressive hunter and has been previously shown to attack more egg batches than other bodyguards in Salix stands (Björkman et al., 2003). Apart from eggs, it may also prey upon small larvae, but it does not typically attack last instar Phratora larvae or adult beetles. Anthocoris eggs are laid inside green plant material such as leaves or stems, including Salix. The species may have one or a few overlapping generations per summer. For this study we collected Anthocoris adults from stinging nettle (Urtica dioica) and first instar nymphs from Salix cinerea and kept them on green beans at room temperature until the beginning of the experiment. It should further be noted that Phratora and Anthocoris have different modes when they feed on Salix leaves. While Phratora defoliates the leaves, causing significant visible damage, Anthocoris pierces and sucks fluid from the leaves, causing no readily apparent damage. The two species may, therefore, be exposed to fundamentally different chemical and physiological plant characters. 2.2. Herbivore preference On June 17, sixteen 50 cm tall potted Salix plants of each of the two clones Gudrun and 78-0-183 were placed at random in a backyard of the university campus, to allow natural colonization by Phratora. The distance between neighboring plants was about 1 m. On August 25, the leaf area that had been consumed by naturally colonizing Phratora (adult and larval grazing) was scored visually for the two clones. A one-way ANOVA, with clone as a fixed factor, was performed to analyze the data. The statistical software package R 2.7.2 was used for all tests. 2.3. Herbivore performance: egg production Pairs (♀♂) of Phratora were placed in plastic jars and fed with detached leaves from one of the four Salix clones. Eggs were counted and leaves were replaced every second day. Because males may stress the females and negatively affect the oviposition rate, each male was removed as soon as egg-laying commenced. Each female was monitored until she died or until no eggs had been laid during two consecutive scoring events. Plant material for this experiment was planted six weeks before the start of the experiment and used for feeding throughout the experiment. Thus, the beetles experienced continuously ageing plants, but leaves were of the same relative age, i.e. leaves from the top 10 cm and bottom 20 cm were not used. Moreover, each individual plant was only used for two consecutive egg counts. Twenty Phratora pairs were used for each clone. An ANOVA with clone as the fixed factor and total egg production as the dependent variable was used to assess direct resistance of the different clones. Individual females that did not produce any eggs at all were removed from the analysis. In addition, an exact binomial test (Crawley, 2002) was used to examine whether the ratio of females failing to produce any eggs at all differed between clones.
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
2.4. Herbivore performance: larval development Eggs from all females in the above experiment were mixed, and 100 eggs were randomly assigned to each Salix clone in groups of five (n = 20), to resemble gregarious feeding during the first larval instar. Each group of five eggs was placed in a plastic jar and the hatched larvae were provided with freshly detached leaves from one of the four clones every second day. As soon as one of the larvae reached the second instar, the other larvae from that group were removed to mimic solitary feeding during the rest of the larval development. The larvae were checked every day and the response variables larval development time (days until prepupal stage), pupal weight, and survival were scored. The effect of clone on development time and pupal weight was assessed using a oneway ANOVA, and differences in survival rate were assessed using an exact binomial test.
a
40
Leaf area consumed (cm2)
530
30
20
10
ND 2.5. Omnivore preference
First instar Anthocoris nymphs were collected from Salix cinerea in the field and placed individually on bagged six-week-old herbivore-free potted saplings of each of the clones (n = 40–60). As it was not possible to make recordings without removing the bags, the bagged plants were left for 30 days, whereafter the bags were removed and survival of the Anthocoris was scored. At this time all Anthocoris had reached adulthood, except dead individuals of which the majority were recovered and found to have died during their second or third instar. 3. Results 3.1. Herbivore preference and performance The mean leaf area consumed was more than 30 times higher on clone 78-0-183 (30.6 cm2 ) than on Gudrun (0.9 cm2 ), demonstrating that Phratora discriminates significantly between these two clones (F1,30 = 11.17, p < 0.002, Fig. 1(a)). In Fig. 1(a) clones Loden and 78-0-21 are labelled “ND” (no data) as they were not included in this experiment. In the no choice experiment, females had the highest mean life time production of eggs when they were fed with leaves from clones 78-0-183 and 78-0-21 (395 and 338 eggs, respectively) and the lowest production on clones Gudrun and Loden (18 and 183 eggs, respectively); the differences between clones were significant (p < 0.001, Fig. 1(b)), coinciding with the feeding preference data. In the no-choice experiment when Phratora females were reared on the four different clones their oviposition rate was almost halved on clone Gudrun compared to the other clones (Fig. 1(c)). When larvae were reared on the four clones, development time (F2,50 = 16.15, p < 0.001), pupal weight (F2,48 = 5.20, p = 0.009) and survival (2 = 49.67, p = 0.001, df = 3) varied significantly between clones (Fig. 2). Larvae performed very poorly on Gudrun, quite well on Loden, and very well on 78-0183 and 78-0-21, in line with the data on adult feeding preference
400
Mean life time egg production
2.6. Omnivore performance
b
300
200
100
0
c
Females not laying any eggs
One detached leaf from each of Gudrun vs. 78-0-183 or Loden vs. 78-0-21 were placed as pairs in moistened oasis in a glass box. The chosen matching of clones in pairs was arbitrary apart from the fact that the two clones in a pair should represent a clone with strong and a clone with weak direct defence. A similar setup was used by Sigsgaard (2004). One Anthocoris was released into each box, and its location was recorded after 24 h. For both clone pairs an exact binomial test was used to assess whether the observed Anthocoris distribution differed significantly from a random 50/50 distribution.
ND
0
10
5
0 Gudrun
Loden
78-0-21
78-0-183
Fig. 1. Preference and performance of adult Phratora vulgatissima (herbivore) on the four Salix clones Gudrun (black bars), Loden (dark grey bars), 78-0-21 (light grey bars), and 78-0-183 (white bars). (a) Leaf area consumed by natural Phratora grazing in an experimental garden (n = 16; potted plants); (b) mean life time egg production by solitary Phratora females fed with detached leaves (n = 20 females); (c) number of Phratora females that died prior to commencing oviposition or that refrained from laying eggs. Error bars indicate standard errors. ND = no data.
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
0.9
Proportion of individuals selecting
a 16
Larval development time (days)
531
12
8
4
0.6
0.3
0
None
Gudrun
0 Gudrun
Loden
78-0-21
78-0-183
b 9
78-0-183
Loden
78-0-21
Fig. 3. The location of Anthocoris nemorum adults (omnivorous predator) after 2 h in pairwise choice tests with detached herbivore-free leaves of the following Salix clones: Gudrun (black bars) vs. 78-0-183 (white bars), and Loden (dark grey bars) vs. 78-0-21 (light grey bars). n = 24 for Gudrun vs. 78-0-183; n = 22 for Loden vs. 78-0-21.
25
6
Survival rate (%)
Pupal weight (mg)
30
3
20
15
10
None 0 Gudrun
Loden
78-0-21
78-0-183
Fig. 2. Performance of Phratora vulgatissima larvae (herbivore) when fed with detached leaves from the Salix clones Gudrun (no bars), Loden (dark grey bars), 78-0-21 (light grey bars), and 78-0-183 (white bars). (a) Number of days to prepupal stage; (b) pupal weight. n = 17–19 for all treatments, except for Gudrun on which no larvae survived to the prepupal stage. Error bars indicate standard error. “None” indicates that no larvae survived to the prepupal stage.
5
0 Gudrun
Loden
78-0-21
78-0-183
Fig. 4. Survival until adulthood of Anthocoris nemorum when reared on the following herbivore-free Salix clones: Gudrun (black bars), Loden (dark grey bars), 78-0-21 (light grey bars), and 78-0-183 (white bars). n = 40–60.
4. Discussion and egg production (Fig. 2). Because no larva survived to the pupal stage on Gudrun this clone is labelled “None” in Fig. 2(a) and (b).
3.2. Omnivore preference and performance In the preference test, a higher proportion of the Anthocoris individuals were observed on Gudrun and Loden (0.71 and 0.82, respectively) than on 78-0-183 and 78-0-21 (0.29 and 0.18, respectively) (p = 0.064 for Gudrun vs. 78-0-183; p = 0.004 for Loden vs. 78-0-21; Fig. 3). Note, however, that the p-value for Gudrun vs. 78-0-183 was slightly over the 0.05 significance level. In the performance experiment, where Anthocoris were restricted to different Salix clones, a significant difference in survival was detected between clones (2 = 23.72, p < 0.001, df = 3). Survival rates were much higher on Gudrun than on 78-0-21 and 78-0-183, while survival on clone Loden turned out to be intermediate (Fig. 4).
We found that the individual Salix clones that exhibited good direct defence, in terms of herbivore deterrence and resistance, also exhibited good indirect defence, in terms of attracting/retaining and supporting bodyguards. Correspondingly, the clones with limited direct defence also exhibited limited indirect defence. These results show that, in Salix, at least some genotypes do not subscribe to the traditionally presumed dichotomy between direct and indirect defences. Although the present study was not designed to deliver general conclusions about trade-offs in Salix our results hint that the studied defence traits may co-vary, either showing a defensive cocktail that is optimized to counteract herbivores or traded, presumably in favour of some other set of traits (e.g. growth or tolerance). van der Meijden et al. (1988) suggested that plant defence is likely to be traded off against tolerance. In the present case, however, the tolerance of the tested clones does not differ (Martin Weih, unpublished data). Thus, since both the maintenance of indirect (e.g. Gershenzon, 1994; Penuelas and Llusia, 2003) and direct defences (e.g. Stenberg et al., 2006) are generally associated with metabolic costs, these must be traded off against some other
532
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
trait or set of traits, yet to be studied. The extent to which this plant has to trade investments in other traits against defence is currently unknown. Future studies should investigate the basis of genotypic variation in the defences to disentangle the cost issue. In a recent review Agrawal (2007) showed that the scientific support for the widely supposed univariate trade-offs between different direct plant defences, which sometimes inferred from the Optimal Defence Hypothesis, in reality is very scarce. Indeed, univariate trade-offs between different direct defences may be the exception rather than the rule among plants. Ambitious investigations into trade-offs between direct and indirect defences are, however, largely absent from the scientific literature. Our study supports Agrawal’s remark that “there should not be an a priori expectation that two defensive traits should be negatively correlated” even when indirect defences are tested against direct defences. Future studies that include a wider range of plant genotypes should be undertaken in order to elucidate the general pattern of direct and indirect defences in Salix. Although both the bodyguard and the herbivore consumed leaf material from identical plant structures, they showed contrasting responses to that plant material. There are at least two possible reasons for this. As mentioned above, the two species have different feeding modes: the herbivore skeletonizes the leaves, while the bodyguard sucks fluid from them. Thus, although the herbivore and the bodyguard feed on the same plant structure they may be accessing different plant material, which may enable plants to allocate important resources or toxins in such a way that they are mainly exposed to only one of the species. This hypothesis may be too simplistic, however, since in the field psyllids, which have a feeding mode similar to the bodyguard, are mainly found on clones 78-0-183 and 78-0-21 rather than Gudrun and Loden, thus exhibiting similarities to the skeletonizing herbivore Phratora (Stenberg et al., unpublished data). An unexplored possibility may be that the psyllids feed in the symplast and the “bodyguard” in the apoplast. Another alternative hypothesis could be that herbivores and omnivorous bodyguards have vastly different nutritional requirements from the plant; this may enable selection to produce plant material that is favourable to the bodyguards but unfavourable to herbivores – both those that chew and those that suck. Further studies are needed in order to elucidate the exact mechanisms. When developing a strategy for integrated pest management to improve cropping security it is important to relate to how herbivore resistance evolve in natural systems. An important issue is why a plant with relatively strong direct defences would benefit from adding a presumably costly indirect defence. We can suggest three answers to this question. First, if the effect of a second line of defence is at least additive to the first, selection may favour both defences if herbivore pressure is constantly high. Second, given that everything else is equal, we would expect bodyguards to be selected to reside on the plant genotypes on which they are most likely to find high numbers of prey; i.e. on susceptible plants. However, in a co-evolutionary mosaic context (Thompson, 2005), even plants with a relatively strong direct defence system will at least occasionally be subject to high herbivore pressure, and if all bodyguards reside on the susceptible plants then occasional outbreaks may have a devastating impact on plants with no bodyguards. Thus, plants with an unbalanced defence system (high direct defence, no indirect defence) may be completely devoid of bodyguards in the event of an occasional herbivore outbreak. Therefore, a high investment in direct defence may only pay off if it is balanced with a corresponding investment in indirect defences. In other words, an efficient life insurance policy is likely to involve both direct and indirect defences. Third, plants often have more than one important herbivore, and may in some cases target different herbivores with different defences.
Our results open up the possibility for breeders to simultaneously select both for high direct and indirect defences. The positive effects of these traits, however, have to be contrasted to possible costs, which potentially may comprise a reduced intrinsic growth rate. In the case of Salix short rotation coppices, however, where up to 40% of the biomass may be lost to herbivores (Björkman et al., 2000a), potential costs of resistance should quickly be outweighed by the reduced leaf-beetle herbivory. One weakness with our study is that we did not examine bodyguard performance in the presence of the herbivore, but choose to study the direct effect of plant genotype on the bodyguard vs. on the herbivore separately. As showed by Agrawal et al. (1999) herbivore feeding may induce chemical defences in the plant which ultimately may also affect omnivorous bodyguards. Further, the propensity of the bodyguard to feed on herbivores rather than on the host plant may depend on host-plant quality. These aspects, which may further clarify the relationship between the presumably entangled direct and indirect defences will be subject to future studies. 5. Conclusions Previous hypotheses, most importantly the univariate tradeoff model often inferred from the Optimal Defence Hypothesis, have suggested that univariate trade-offs between defence traits may be common. Subsequently, the Defence Syndrome Hypothesis suggested that linkages between defences arise because of physiological constraints, such that trade-offs would rather be found between groups of defence traits (e.g. direct vs. indirect defence, or direct defence vs. herbivore escape). This paper does not discount previous hypotheses, but adds the hypothesis that plants may not always be obliged to trade-off between different defences, but that the cost may be off-set from traits not associated with herbivory at all. Thus, our findings open up novel opportunities for integrated pest management and improved cropping security in coppicing willow plantations and possibly in other crops. However, possible trade-offs with other traits, such as intrinsic growth, need to be further investigated. Acknowledgements We would like to thank two anonymous reviewers for helpful comments on a previous version of this paper. We are further grateful to Karin Eklund and Anna-Sara Liman for technical assistance during summer 2008. This study was funded by the Swedish Energy Agency (Energimyndigheten), the Swedish research council Formas, and Lantmännen Agroenergi AB. References Agrawal, A.A., Kobayashi, C., Thaler, J.S., 1999. Influence of prey availability and induced host-plant resistance on omnivory by western flower thrips. Ecology 80, 518–523. Agrawal, A.A., 2007. Macroevolution of plant defense strategies. Trends in Ecology and Evolution 22, 103–109. Bell, A.C., Clawson, S., Watson, S., 2006. The long-term effect of partial defoliation on the yield of short-rotation coppice willow. Annals of Applied Biology 48, 97–103. Björkman, C., Höglund, S., Eklund, K., Larsson, S., 2000a. Effects of leaf beetle damage on stem wood production in coppicing willow. Agricultural and Forest Entomology 2, 131–139. Björkman, C., Dalin, P., Eklund, K., 2003. Generalist natural enemies of a willow leaf beetle (Phratora vulgatissima): abundance and feeding habits. Journal of Insect Behavior 16, 747–764. Björkman, C., Bommarco, R., Eklund, K., Höglund, S., 2004. Harvesting disrupts biological control of herbivores in a short-rotation coppice system. Ecological Applications 14, 1624–1633. Björkman, C., Eklund, K., 2006. Factors affecting willow leaf beetles (Phratora vulgatissima) when selecting overwintering sites. Agricultural and Forest Entomology 8, 97–101.
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533 Coll, M., 1996. Feeding and ovipositing on plants by an omnivorous insect predator. Oecologia 105, 214–220. Coll, M., Guershon, M., 2002. Omnivory in terrestrial arthropods: mixing plant and prey diets. Annual Review of Entomology 47, 267–297. Crawley, M.J., 2002. Statistical Computing: an Introduction to Data Analysis using S-Plus. John Wiley and Sons Ltd., Chichester. Dalin, P., Kindvall, O., Björkman, C., 2009. Reduced population control of an insect pest in managed willow monocultures. PLoS One 4 (1–6), e5487. Eubanks, M.D., Denno, R.F., 2000. Host plants mediate omnivore–herbivore interactions and influence prey suppression. Ecology 81, 936–947. Gershenzon, J., 1994. The cost of plant chemical defense against herbivory: a biochemical perspective. In: Bernays, E.A. (Ed.), Insect–Plant Interactions, vol. 5. CRC Press, Boca Raton, pp. 105–173. Glynn, C., Rönnberg-Wästljung, A.C., Julkunen-Tiitto, R., Weih, M., 2004. Willow genotype, but not drought treatment, affects foliar phenolic concentrations and leaf-beetle resistance. Entomologia Experimentalis et Applicata 113, 1–14. Groenteman, R., Guershon, M., Coll, M., 2006. Effects of leaf nitrogen content on oviposition site selection, offspring performance, and intraspecific interactions in an omnivorous bug. Ecological Entomology 31, 155–161. Kelly, M.T., Curry, J.P., 1991. The influence of phenolic compounds on the suitability of 3 Salix species as hosts for the willow beetle Phratora vulgatissima. Entomologia Experimentalis et Applicata 61, 25–32. Lauenstein, G., 1979. Zur Aufnahme von Pflanzen durch die Räuberische Blumenwanze Anthocoris nemorum [Hem.: Heteroptera]. Entomophaga 24, 431–441. McMurtry, J.A., Croft, B.A., 1997. Life-styles of phytoseiid mites and their roles in biological control. Annual Review of Entomology 42, 291–321. Ode, P.J., Berenbaum, M.R., Zangerl, A.R., Hardy, I.C.W., 2004. Host plant, host plant chemistry and the polyembryonic parasitoid Copidosoma sosares: indirect effects in a tritrophic interaction. Oikos 104, 388–400.
533
Peacock, L., Herrick, S., 2000. Responses of the willow beetle Phratora vulgatissima to genetically and spatially diverse Salix spp. plantations. Journal of Applied Ecology 37, 821–831. Peacock, L., Hunter, T., Turner, H., Brain, P., 2001. Does host genotype diversity affect the distribution of insect and disease damage in willow cropping systems? Journal of Applied Ecology 38, 1070–1081. Penuelas, J., Llusia, J., 2003. BVOCs: plant defense against climate warming? Trends in Plant Science 8, 105–109. Price, P.W., Bouton, C.E., Gross, P., McPheron, B.A., Thompson, J.N., Weis, A.E., 1980. Interactions among three trophic levels: influence of plants on interactions between insect herbivores and natural enemies. Annual Review of Ecology and Systematics 11, 41–65. Rehr, S.S., Feeny, P.P., Janzen, D.H., 1973. Chemical defence in in Central American non-ant-acacias. Journal of Animal Ecology 42, 405–416. Rodriguez-Saona, C., Thaler, J.S., 2005. Herbivore-induced responses and patch heterogeneity affect abundance of arthropods on plants. Ecological Entomology 30, 156–163. Rudgers, J.A., Strauss, S.Y., Wendel, J.F., 2004. Trade-offs among anti-herbivore resistance traits: insights from Gossypieae (Malvaceae). American Journal of Botany 91, 878–880. Sigsgaard, L., 2004. Oviposition preference of Anthocoris nemorum and A. nemoralis for apple and pear. Entomologia Experimentalis et Applicata 111, 215–223. Stenberg, J.A., Witzell, J., Ericson, L., 2006. Tall herb herbivory resistance reflects historic exposure to leaf beetles in a boreal archipelago age-gradient. Oecologia 148, 414–425. Thompson, J.N., 2005. The Geographic Mosaic of Coevolution. University of Chicago Press, Chicago. van der Meijden, E., Wijn, M., Verkaar, H.J., 1988. Defense and regrowth, alternative plant strategies in the struggle against herbivores. Oikos 51, 355–363.
Contents lists available at ScienceDirect
Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee
Uncoupling direct and indirect plant defences: Novel opportunities for improving crop security in willow plantations Johan A. Stenberg ∗ , Anna Lehrman, Christer Björkman Swedish University of Agricultural Sciences, Department of Ecology, P.O. Box 7044, SE-75007 Uppsala, Sweden
a r t i c l e
i n f o
Article history: Received 28 May 2010 Received in revised form 19 August 2010 Accepted 22 September 2010 Available online 16 October 2010 Keywords: Bioenergy crop Cropping security Indirect defence Optimal defence Defence syndromes Intraguild predation
a b s t r a c t Increased cropping security, i.e. minimized risk for detrimental stress events and hence improved yield stability, may be achieved by selecting resistant plant genotypes. However, strong direct defences against herbivores have been associated with negative effects on the natural enemies, resulting in weak indirect defences through these plant “bodyguards”. We compared the preference and performance of the most detrimental herbivore, the leaf beetle Phratora vulgatissima, and a biocontrol agent, the predatory bug Anthocoris nemorum, on four willow (Salix) genotypes used in short rotation coppicing. The biocontrol agent is omnivorous and survives on the plant in the absence of prey, without causing any apparent damage to the plant. Two of the genotypes intrinsically attracted and supported the bodyguard, and were at the same time partially capable of deterring and resisting the herbivore. In contrast, the third willow clone exhibited very limited direct and indirect resistance, while the fourth clone displayed intermediate levels of both. Thus, direct and indirect defence against herbivores need not be traded off in this system, but may be intertwined, resulting in “super plants” with higher probability of resisting detrimental herbivores, opening up novel possibilities for increased cropping security, without the use of insecticides. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Herbivores are generally more prone to outbreak in monocultures than in natural systems (Dalin et al., 2009), and may cause severe losses, especially in low-value or organic crops where insecticides are not used. In order to reduce herbivore damage, it is desirable to develop plant genotypes that are intrinsically more resistant to herbivores (direct defence) or that support biological control agents (indirect defence). Intertwining the two defence strategies would probably lead to a more sustainable resistance in the focal plant. Discouraging enough, ecologists have long presumed that there should be an intrinsic trade-off between direct and indirect defences in plants, since both require resources derived from the plant’s metabolic budget (see the Optimal Defence Hypothesis, e.g. Price et al., 1980; Ode et al., 2004). If, however, one disregards examples involving herbivore-mediated effects of plant quality on parasitoids, the literature is rather scarce, and evidence for tradeoffs between direct and indirect defence is mainly limited to plants
∗ Corresponding author. Tel.: +46 18 67 23 67. E-mail addresses: [email protected], [email protected] (J.A. Stenberg), [email protected] (A. Lehrman), [email protected] (C. Björkman). 0167-8809/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.agee.2010.09.013
that produce extra-floral nectar (e.g. Rehr et al., 1973; Rudgers et al., 2004). Trade-offs may also arise when plant traits important for attracting and supporting bodyguards also are favourable for herbivores. For example, many bodyguards are omnivorous and receive important nutrients from the same green plant tissue that the herbivores utilize (Coll, 1996; McMurtry and Croft, 1997; Agrawal et al., 1999; Eubanks and Denno, 2000; Coll and Guershon, 2002; RodriguezSaona and Thaler, 2005; Groenteman et al., 2006). However, the relation between pest insect vs. bodyguard responses to plant genotype has almost entirely been overlooked. Agrawal et al. (1999) and Eubanks and Denno (2000) simultaneously studied the response of herbivores and predators to their shared host-plant and suggested that bodyguards and pest herbivores should respond similarly to shared plant resources (Agrawal et al., 1999; Eubanks and Denno, 2000). We feel, however, that their suggestion may apply more to their specific systems (i.e. cotton and lima bean) rather than having universal significance. Here, we hypothesize that direct and indirect defences may be entangled in more complicated ways. If, for example, the nutritional requirements of pest herbivores and omnivorous bodyguards are different this may open up a possibility for breeders to disassociate the two defence strategies and optimise both of them independently. Rather than looking at direct and indirect defence as two contrasting strategies being subject to trade-off between each other we propose that these two defences are two indispensable aspects
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
of the system, which together underpin an integrated and balanced defence package. Based on empirical evidence and logical reasoning, we argue that in perennial crops it is wise to balance the investment in direct and indirect defences in such a way that they co-vary. In this paper we describe a study of the direct and indirect defences of different Salix (willow) clones against the blue willow beetle, Phratora vulgatissima (Coleoptera: Chrysomelidae), which is a severe pest in European coppicing willow plantations (e.g. Bell et al., 2006). Phratora damage may, however, be reduced by predatory omnivorous bugs such as Anthocoris nemorum (Heteroptera: Anthocoridae), which is considered to be an important bodyguard both for Salix and other plants (e.g. Björkman et al., 2003, 2004). In past breeding programs, the focus has been on direct plant resistance, since existing hypotheses regarding trade-offs have precluded the possibility of simultaneously optimizing direct and indirect defences (Kelly and Curry, 1991; Glynn et al., 2004). Here, we investigated bodyguard attraction to, and fitness on, different clones of Salix, in order to obtain reliable estimates of the clones’ levels of indirect defence. Simultaneously, herbivore attraction to, and reproduction on, the Salix clones were investigated. 2. Materials and methods 2.1. Study species 2.1.1. Plants Willows are grown in monocultures for biomass production in which the use of pesticides is not defendable neither from environmental, nor economic reasons. Due to the frequent hybridization and polyploidization in the genus Salix, variation within and among species may be high and overlapping; this opens up good opportunities to find genotypes with traits suitable for various conditions. It also means that the species concept is problematic – thus clone is often a more relevant identity than species. This taxonomical difficulty is less of a problem in practical bioenergy forestry. Here we chose to work with four clones of Salix, two of which are available from Lantmännen Agroenergi AB: Gudrun and Loden (S. dasyclados), and two breeding clones: 78-0-183 and 78-0-21 (S. viminalis). The specific clones used here were selected as they display phenotypic variation in traits relevant to Phratora and, thus, seem to represent “extremes” in a more or less continuous variation. In addition there is ongoing and existing research on these clones in sister projects. 2.1.2. Herbivore Phratora vulgatissima L. is the most important defoliator of Salix in Europe and has a broad distribution across Europe and Asia. It is restricted to Salix and both adults and larvae skeletonize the leaves (Peacock and Herrick, 2000; Peacock et al., 2001). Overwintering takes place outside Salix plantations in vertical objects that can provide shelter, for example reeds or trees with ageing bark (Björkman and Eklund, 2006). The overwintering adults emerge in April and feed on the young Salix leaves for about two weeks before mating and subsequent egg-laying. Larvae feed on the leaves for about 18–30 days, depending on leaf quality and temperature, and then they pupate in the ground. Adults emerge after about two weeks and feed on the leaves before hibernation. Generally, Phratora has only one generation per year in Sweden. For this study, we collected Phratora individuals from known overwintering sites near Uppsala, Sweden, during late February and kept them in a refrigerator at 5 ◦ C until the beginning of the experiment. The omnivorous predatory bug Anthocoris nemorum L. is widely distributed across Europe and Asia. It can be found in large numbers not only on Salix, but also on, among other plants, apple
529
trees and nettles. Anthocoris overwinters as an adult and becomes active as soon as the temperature rises in the spring. First instar nymphs have been observed sucking plant fluid from leaves and twig bases, which indicates that vegetable matter may be a very important food source for nymphal survival (personal observation). It is, however, very important to note that although Anthocoris may be nourished by Salix, and in extreme situations even completes its development without animal food, previous studies have shown that plant-feeding Anthocoris do not reduce plant fitness (Lauenstein, 1979). Thus, any direct cost of anthocorid feeding on Salix should be incommensurably lower than the cost of Phratora grazing, which may reduce Salix growth with up to 40% (Björkman et al., 2000a). Older anthocorid individuals are predominantly carnivores. Anthocoris is an active and aggressive hunter and has been previously shown to attack more egg batches than other bodyguards in Salix stands (Björkman et al., 2003). Apart from eggs, it may also prey upon small larvae, but it does not typically attack last instar Phratora larvae or adult beetles. Anthocoris eggs are laid inside green plant material such as leaves or stems, including Salix. The species may have one or a few overlapping generations per summer. For this study we collected Anthocoris adults from stinging nettle (Urtica dioica) and first instar nymphs from Salix cinerea and kept them on green beans at room temperature until the beginning of the experiment. It should further be noted that Phratora and Anthocoris have different modes when they feed on Salix leaves. While Phratora defoliates the leaves, causing significant visible damage, Anthocoris pierces and sucks fluid from the leaves, causing no readily apparent damage. The two species may, therefore, be exposed to fundamentally different chemical and physiological plant characters. 2.2. Herbivore preference On June 17, sixteen 50 cm tall potted Salix plants of each of the two clones Gudrun and 78-0-183 were placed at random in a backyard of the university campus, to allow natural colonization by Phratora. The distance between neighboring plants was about 1 m. On August 25, the leaf area that had been consumed by naturally colonizing Phratora (adult and larval grazing) was scored visually for the two clones. A one-way ANOVA, with clone as a fixed factor, was performed to analyze the data. The statistical software package R 2.7.2 was used for all tests. 2.3. Herbivore performance: egg production Pairs (♀♂) of Phratora were placed in plastic jars and fed with detached leaves from one of the four Salix clones. Eggs were counted and leaves were replaced every second day. Because males may stress the females and negatively affect the oviposition rate, each male was removed as soon as egg-laying commenced. Each female was monitored until she died or until no eggs had been laid during two consecutive scoring events. Plant material for this experiment was planted six weeks before the start of the experiment and used for feeding throughout the experiment. Thus, the beetles experienced continuously ageing plants, but leaves were of the same relative age, i.e. leaves from the top 10 cm and bottom 20 cm were not used. Moreover, each individual plant was only used for two consecutive egg counts. Twenty Phratora pairs were used for each clone. An ANOVA with clone as the fixed factor and total egg production as the dependent variable was used to assess direct resistance of the different clones. Individual females that did not produce any eggs at all were removed from the analysis. In addition, an exact binomial test (Crawley, 2002) was used to examine whether the ratio of females failing to produce any eggs at all differed between clones.
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
2.4. Herbivore performance: larval development Eggs from all females in the above experiment were mixed, and 100 eggs were randomly assigned to each Salix clone in groups of five (n = 20), to resemble gregarious feeding during the first larval instar. Each group of five eggs was placed in a plastic jar and the hatched larvae were provided with freshly detached leaves from one of the four clones every second day. As soon as one of the larvae reached the second instar, the other larvae from that group were removed to mimic solitary feeding during the rest of the larval development. The larvae were checked every day and the response variables larval development time (days until prepupal stage), pupal weight, and survival were scored. The effect of clone on development time and pupal weight was assessed using a oneway ANOVA, and differences in survival rate were assessed using an exact binomial test.
a
40
Leaf area consumed (cm2)
530
30
20
10
ND 2.5. Omnivore preference
First instar Anthocoris nymphs were collected from Salix cinerea in the field and placed individually on bagged six-week-old herbivore-free potted saplings of each of the clones (n = 40–60). As it was not possible to make recordings without removing the bags, the bagged plants were left for 30 days, whereafter the bags were removed and survival of the Anthocoris was scored. At this time all Anthocoris had reached adulthood, except dead individuals of which the majority were recovered and found to have died during their second or third instar. 3. Results 3.1. Herbivore preference and performance The mean leaf area consumed was more than 30 times higher on clone 78-0-183 (30.6 cm2 ) than on Gudrun (0.9 cm2 ), demonstrating that Phratora discriminates significantly between these two clones (F1,30 = 11.17, p < 0.002, Fig. 1(a)). In Fig. 1(a) clones Loden and 78-0-21 are labelled “ND” (no data) as they were not included in this experiment. In the no choice experiment, females had the highest mean life time production of eggs when they were fed with leaves from clones 78-0-183 and 78-0-21 (395 and 338 eggs, respectively) and the lowest production on clones Gudrun and Loden (18 and 183 eggs, respectively); the differences between clones were significant (p < 0.001, Fig. 1(b)), coinciding with the feeding preference data. In the no-choice experiment when Phratora females were reared on the four different clones their oviposition rate was almost halved on clone Gudrun compared to the other clones (Fig. 1(c)). When larvae were reared on the four clones, development time (F2,50 = 16.15, p < 0.001), pupal weight (F2,48 = 5.20, p = 0.009) and survival (2 = 49.67, p = 0.001, df = 3) varied significantly between clones (Fig. 2). Larvae performed very poorly on Gudrun, quite well on Loden, and very well on 78-0183 and 78-0-21, in line with the data on adult feeding preference
400
Mean life time egg production
2.6. Omnivore performance
b
300
200
100
0
c
Females not laying any eggs
One detached leaf from each of Gudrun vs. 78-0-183 or Loden vs. 78-0-21 were placed as pairs in moistened oasis in a glass box. The chosen matching of clones in pairs was arbitrary apart from the fact that the two clones in a pair should represent a clone with strong and a clone with weak direct defence. A similar setup was used by Sigsgaard (2004). One Anthocoris was released into each box, and its location was recorded after 24 h. For both clone pairs an exact binomial test was used to assess whether the observed Anthocoris distribution differed significantly from a random 50/50 distribution.
ND
0
10
5
0 Gudrun
Loden
78-0-21
78-0-183
Fig. 1. Preference and performance of adult Phratora vulgatissima (herbivore) on the four Salix clones Gudrun (black bars), Loden (dark grey bars), 78-0-21 (light grey bars), and 78-0-183 (white bars). (a) Leaf area consumed by natural Phratora grazing in an experimental garden (n = 16; potted plants); (b) mean life time egg production by solitary Phratora females fed with detached leaves (n = 20 females); (c) number of Phratora females that died prior to commencing oviposition or that refrained from laying eggs. Error bars indicate standard errors. ND = no data.
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
0.9
Proportion of individuals selecting
a 16
Larval development time (days)
531
12
8
4
0.6
0.3
0
None
Gudrun
0 Gudrun
Loden
78-0-21
78-0-183
b 9
78-0-183
Loden
78-0-21
Fig. 3. The location of Anthocoris nemorum adults (omnivorous predator) after 2 h in pairwise choice tests with detached herbivore-free leaves of the following Salix clones: Gudrun (black bars) vs. 78-0-183 (white bars), and Loden (dark grey bars) vs. 78-0-21 (light grey bars). n = 24 for Gudrun vs. 78-0-183; n = 22 for Loden vs. 78-0-21.
25
6
Survival rate (%)
Pupal weight (mg)
30
3
20
15
10
None 0 Gudrun
Loden
78-0-21
78-0-183
Fig. 2. Performance of Phratora vulgatissima larvae (herbivore) when fed with detached leaves from the Salix clones Gudrun (no bars), Loden (dark grey bars), 78-0-21 (light grey bars), and 78-0-183 (white bars). (a) Number of days to prepupal stage; (b) pupal weight. n = 17–19 for all treatments, except for Gudrun on which no larvae survived to the prepupal stage. Error bars indicate standard error. “None” indicates that no larvae survived to the prepupal stage.
5
0 Gudrun
Loden
78-0-21
78-0-183
Fig. 4. Survival until adulthood of Anthocoris nemorum when reared on the following herbivore-free Salix clones: Gudrun (black bars), Loden (dark grey bars), 78-0-21 (light grey bars), and 78-0-183 (white bars). n = 40–60.
4. Discussion and egg production (Fig. 2). Because no larva survived to the pupal stage on Gudrun this clone is labelled “None” in Fig. 2(a) and (b).
3.2. Omnivore preference and performance In the preference test, a higher proportion of the Anthocoris individuals were observed on Gudrun and Loden (0.71 and 0.82, respectively) than on 78-0-183 and 78-0-21 (0.29 and 0.18, respectively) (p = 0.064 for Gudrun vs. 78-0-183; p = 0.004 for Loden vs. 78-0-21; Fig. 3). Note, however, that the p-value for Gudrun vs. 78-0-183 was slightly over the 0.05 significance level. In the performance experiment, where Anthocoris were restricted to different Salix clones, a significant difference in survival was detected between clones (2 = 23.72, p < 0.001, df = 3). Survival rates were much higher on Gudrun than on 78-0-21 and 78-0-183, while survival on clone Loden turned out to be intermediate (Fig. 4).
We found that the individual Salix clones that exhibited good direct defence, in terms of herbivore deterrence and resistance, also exhibited good indirect defence, in terms of attracting/retaining and supporting bodyguards. Correspondingly, the clones with limited direct defence also exhibited limited indirect defence. These results show that, in Salix, at least some genotypes do not subscribe to the traditionally presumed dichotomy between direct and indirect defences. Although the present study was not designed to deliver general conclusions about trade-offs in Salix our results hint that the studied defence traits may co-vary, either showing a defensive cocktail that is optimized to counteract herbivores or traded, presumably in favour of some other set of traits (e.g. growth or tolerance). van der Meijden et al. (1988) suggested that plant defence is likely to be traded off against tolerance. In the present case, however, the tolerance of the tested clones does not differ (Martin Weih, unpublished data). Thus, since both the maintenance of indirect (e.g. Gershenzon, 1994; Penuelas and Llusia, 2003) and direct defences (e.g. Stenberg et al., 2006) are generally associated with metabolic costs, these must be traded off against some other
532
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533
trait or set of traits, yet to be studied. The extent to which this plant has to trade investments in other traits against defence is currently unknown. Future studies should investigate the basis of genotypic variation in the defences to disentangle the cost issue. In a recent review Agrawal (2007) showed that the scientific support for the widely supposed univariate trade-offs between different direct plant defences, which sometimes inferred from the Optimal Defence Hypothesis, in reality is very scarce. Indeed, univariate trade-offs between different direct defences may be the exception rather than the rule among plants. Ambitious investigations into trade-offs between direct and indirect defences are, however, largely absent from the scientific literature. Our study supports Agrawal’s remark that “there should not be an a priori expectation that two defensive traits should be negatively correlated” even when indirect defences are tested against direct defences. Future studies that include a wider range of plant genotypes should be undertaken in order to elucidate the general pattern of direct and indirect defences in Salix. Although both the bodyguard and the herbivore consumed leaf material from identical plant structures, they showed contrasting responses to that plant material. There are at least two possible reasons for this. As mentioned above, the two species have different feeding modes: the herbivore skeletonizes the leaves, while the bodyguard sucks fluid from them. Thus, although the herbivore and the bodyguard feed on the same plant structure they may be accessing different plant material, which may enable plants to allocate important resources or toxins in such a way that they are mainly exposed to only one of the species. This hypothesis may be too simplistic, however, since in the field psyllids, which have a feeding mode similar to the bodyguard, are mainly found on clones 78-0-183 and 78-0-21 rather than Gudrun and Loden, thus exhibiting similarities to the skeletonizing herbivore Phratora (Stenberg et al., unpublished data). An unexplored possibility may be that the psyllids feed in the symplast and the “bodyguard” in the apoplast. Another alternative hypothesis could be that herbivores and omnivorous bodyguards have vastly different nutritional requirements from the plant; this may enable selection to produce plant material that is favourable to the bodyguards but unfavourable to herbivores – both those that chew and those that suck. Further studies are needed in order to elucidate the exact mechanisms. When developing a strategy for integrated pest management to improve cropping security it is important to relate to how herbivore resistance evolve in natural systems. An important issue is why a plant with relatively strong direct defences would benefit from adding a presumably costly indirect defence. We can suggest three answers to this question. First, if the effect of a second line of defence is at least additive to the first, selection may favour both defences if herbivore pressure is constantly high. Second, given that everything else is equal, we would expect bodyguards to be selected to reside on the plant genotypes on which they are most likely to find high numbers of prey; i.e. on susceptible plants. However, in a co-evolutionary mosaic context (Thompson, 2005), even plants with a relatively strong direct defence system will at least occasionally be subject to high herbivore pressure, and if all bodyguards reside on the susceptible plants then occasional outbreaks may have a devastating impact on plants with no bodyguards. Thus, plants with an unbalanced defence system (high direct defence, no indirect defence) may be completely devoid of bodyguards in the event of an occasional herbivore outbreak. Therefore, a high investment in direct defence may only pay off if it is balanced with a corresponding investment in indirect defences. In other words, an efficient life insurance policy is likely to involve both direct and indirect defences. Third, plants often have more than one important herbivore, and may in some cases target different herbivores with different defences.
Our results open up the possibility for breeders to simultaneously select both for high direct and indirect defences. The positive effects of these traits, however, have to be contrasted to possible costs, which potentially may comprise a reduced intrinsic growth rate. In the case of Salix short rotation coppices, however, where up to 40% of the biomass may be lost to herbivores (Björkman et al., 2000a), potential costs of resistance should quickly be outweighed by the reduced leaf-beetle herbivory. One weakness with our study is that we did not examine bodyguard performance in the presence of the herbivore, but choose to study the direct effect of plant genotype on the bodyguard vs. on the herbivore separately. As showed by Agrawal et al. (1999) herbivore feeding may induce chemical defences in the plant which ultimately may also affect omnivorous bodyguards. Further, the propensity of the bodyguard to feed on herbivores rather than on the host plant may depend on host-plant quality. These aspects, which may further clarify the relationship between the presumably entangled direct and indirect defences will be subject to future studies. 5. Conclusions Previous hypotheses, most importantly the univariate tradeoff model often inferred from the Optimal Defence Hypothesis, have suggested that univariate trade-offs between defence traits may be common. Subsequently, the Defence Syndrome Hypothesis suggested that linkages between defences arise because of physiological constraints, such that trade-offs would rather be found between groups of defence traits (e.g. direct vs. indirect defence, or direct defence vs. herbivore escape). This paper does not discount previous hypotheses, but adds the hypothesis that plants may not always be obliged to trade-off between different defences, but that the cost may be off-set from traits not associated with herbivory at all. Thus, our findings open up novel opportunities for integrated pest management and improved cropping security in coppicing willow plantations and possibly in other crops. However, possible trade-offs with other traits, such as intrinsic growth, need to be further investigated. Acknowledgements We would like to thank two anonymous reviewers for helpful comments on a previous version of this paper. We are further grateful to Karin Eklund and Anna-Sara Liman for technical assistance during summer 2008. This study was funded by the Swedish Energy Agency (Energimyndigheten), the Swedish research council Formas, and Lantmännen Agroenergi AB. References Agrawal, A.A., Kobayashi, C., Thaler, J.S., 1999. Influence of prey availability and induced host-plant resistance on omnivory by western flower thrips. Ecology 80, 518–523. Agrawal, A.A., 2007. Macroevolution of plant defense strategies. Trends in Ecology and Evolution 22, 103–109. Bell, A.C., Clawson, S., Watson, S., 2006. The long-term effect of partial defoliation on the yield of short-rotation coppice willow. Annals of Applied Biology 48, 97–103. Björkman, C., Höglund, S., Eklund, K., Larsson, S., 2000a. Effects of leaf beetle damage on stem wood production in coppicing willow. Agricultural and Forest Entomology 2, 131–139. Björkman, C., Dalin, P., Eklund, K., 2003. Generalist natural enemies of a willow leaf beetle (Phratora vulgatissima): abundance and feeding habits. Journal of Insect Behavior 16, 747–764. Björkman, C., Bommarco, R., Eklund, K., Höglund, S., 2004. Harvesting disrupts biological control of herbivores in a short-rotation coppice system. Ecological Applications 14, 1624–1633. Björkman, C., Eklund, K., 2006. Factors affecting willow leaf beetles (Phratora vulgatissima) when selecting overwintering sites. Agricultural and Forest Entomology 8, 97–101.
J.A. Stenberg et al. / Agriculture, Ecosystems and Environment 139 (2010) 528–533 Coll, M., 1996. Feeding and ovipositing on plants by an omnivorous insect predator. Oecologia 105, 214–220. Coll, M., Guershon, M., 2002. Omnivory in terrestrial arthropods: mixing plant and prey diets. Annual Review of Entomology 47, 267–297. Crawley, M.J., 2002. Statistical Computing: an Introduction to Data Analysis using S-Plus. John Wiley and Sons Ltd., Chichester. Dalin, P., Kindvall, O., Björkman, C., 2009. Reduced population control of an insect pest in managed willow monocultures. PLoS One 4 (1–6), e5487. Eubanks, M.D., Denno, R.F., 2000. Host plants mediate omnivore–herbivore interactions and influence prey suppression. Ecology 81, 936–947. Gershenzon, J., 1994. The cost of plant chemical defense against herbivory: a biochemical perspective. In: Bernays, E.A. (Ed.), Insect–Plant Interactions, vol. 5. CRC Press, Boca Raton, pp. 105–173. Glynn, C., Rönnberg-Wästljung, A.C., Julkunen-Tiitto, R., Weih, M., 2004. Willow genotype, but not drought treatment, affects foliar phenolic concentrations and leaf-beetle resistance. Entomologia Experimentalis et Applicata 113, 1–14. Groenteman, R., Guershon, M., Coll, M., 2006. Effects of leaf nitrogen content on oviposition site selection, offspring performance, and intraspecific interactions in an omnivorous bug. Ecological Entomology 31, 155–161. Kelly, M.T., Curry, J.P., 1991. The influence of phenolic compounds on the suitability of 3 Salix species as hosts for the willow beetle Phratora vulgatissima. Entomologia Experimentalis et Applicata 61, 25–32. Lauenstein, G., 1979. Zur Aufnahme von Pflanzen durch die Räuberische Blumenwanze Anthocoris nemorum [Hem.: Heteroptera]. Entomophaga 24, 431–441. McMurtry, J.A., Croft, B.A., 1997. Life-styles of phytoseiid mites and their roles in biological control. Annual Review of Entomology 42, 291–321. Ode, P.J., Berenbaum, M.R., Zangerl, A.R., Hardy, I.C.W., 2004. Host plant, host plant chemistry and the polyembryonic parasitoid Copidosoma sosares: indirect effects in a tritrophic interaction. Oikos 104, 388–400.
533
Peacock, L., Herrick, S., 2000. Responses of the willow beetle Phratora vulgatissima to genetically and spatially diverse Salix spp. plantations. Journal of Applied Ecology 37, 821–831. Peacock, L., Hunter, T., Turner, H., Brain, P., 2001. Does host genotype diversity affect the distribution of insect and disease damage in willow cropping systems? Journal of Applied Ecology 38, 1070–1081. Penuelas, J., Llusia, J., 2003. BVOCs: plant defense against climate warming? Trends in Plant Science 8, 105–109. Price, P.W., Bouton, C.E., Gross, P., McPheron, B.A., Thompson, J.N., Weis, A.E., 1980. Interactions among three trophic levels: influence of plants on interactions between insect herbivores and natural enemies. Annual Review of Ecology and Systematics 11, 41–65. Rehr, S.S., Feeny, P.P., Janzen, D.H., 1973. Chemical defence in in Central American non-ant-acacias. Journal of Animal Ecology 42, 405–416. Rodriguez-Saona, C., Thaler, J.S., 2005. Herbivore-induced responses and patch heterogeneity affect abundance of arthropods on plants. Ecological Entomology 30, 156–163. Rudgers, J.A., Strauss, S.Y., Wendel, J.F., 2004. Trade-offs among anti-herbivore resistance traits: insights from Gossypieae (Malvaceae). American Journal of Botany 91, 878–880. Sigsgaard, L., 2004. Oviposition preference of Anthocoris nemorum and A. nemoralis for apple and pear. Entomologia Experimentalis et Applicata 111, 215–223. Stenberg, J.A., Witzell, J., Ericson, L., 2006. Tall herb herbivory resistance reflects historic exposure to leaf beetles in a boreal archipelago age-gradient. Oecologia 148, 414–425. Thompson, J.N., 2005. The Geographic Mosaic of Coevolution. University of Chicago Press, Chicago. van der Meijden, E., Wijn, M., Verkaar, H.J., 1988. Defense and regrowth, alternative plant strategies in the struggle against herbivores. Oikos 51, 355–363.