Plant Species Loss Affects Life-History Traits of Aphids and Their Parasitoids Authors: Jana S. Petermann, Christine B. Müller, Christiane Roscher, Alexandra Weigelt, Wolfgang W. Weisser, Bernhard Schmid Published: 2010, PLOS ONE DOI: 10.1371/journal.pone.0012053 ABSTRACT The consequences of plant species loss are rarely assessed in a multi-trophic context and especially effects on life-history traits of organisms at higher trophic levels have remained largely unstudied. We used a grassland biodiversity experiment and measured the effects of two components of plant diversity, plant species richness and the presence of nitrogen-fixing legumes, on several life-history traits of naturally colonizing aphids and their primary and secondary parasitoids in the field. We found that, irrespective of aphid species identity, the proportion of winged aphid morphs decreased with increasing plant species richness, which was correlated with decreasing host plant biomass. Similarly, emergence proportions of parasitoids decreased with increasing plant species richness. Both, emergence proportions and proportions of female parasitoids were lower in plots with legumes, where host plants had increased nitrogen concentrations. This effect of legume presence could indicate that aphids were better defended against parasitoids in high-nitrogen environments. Body mass of emerged individuals of the two most abundant primary parasitoid species was, however, higher in plots with legumes, suggesting that once parasitoids could overcome aphid defenses, they could profit from larger or more nutritious hosts. Our study demonstrates that cascading effects of plant species loss on higher trophic levels such as aphids, parasitoids and secondary parasitoids begin with changed life-history traits of these insects. Thus, life-history traits of organisms at higher trophic levels may be useful indicators of bottom-up effects of plant diversity on the biodiversity of consumers. INTRODUCTION The consequences of the ongoing loss of plant species have been studied intensively, but rarely at other trophic levels than that of the plants themselves [1] , [2] . If plant species disappear from a multi-trophic system, it is conceivable that the effects of this loss of primary producers “cascade up” the system and cause changes at higher trophic levels. The first bottom-up cascades [3] were described from aquatic systems, but they have since been suggested to operate in a similar way in terrestrial systems [4] . Indeed, a few field and laboratory experiments have demonstrated terrestrial “community- and population-level” bottom-up cascades [4] , such as effects of plant abundance and richness on the abundance and richness of herbivores or predators [5] , [6] , [7] , [8] , [9] , [10] , [11] . In contrast to these community- and population-level approaches, other studies examined cascading effects at the level of the individual organism (“individual-level cascades” [4] ). These studies typically manipulated plant quality and measured effects on life-history traits of individuals at higher trophic levels [12] , [13] , [14] , [15] , [16] , [17] . In this study, we hypothesize that changes in plant community diversity affect life-history traits of insects at higher trophic levels via changes in plant abundance and quality, effectively linking community- and individual-level cascades. To address this hypothesis, we used aphid–parasitoid communities as a model system. The advantage of this system is that both aphids and their parasitic wasps have a relatively short range of movement, and due to their high degree of specialization are strongly dependent on the presence and quality of their host species [18] , [19] , [20] , [21] . Aphids often prefer single or few host plant species at least during part of their life cycle [18] . Once a winged (alate) aphid colonizer has reached such a host plant, a colony forms via asexual reproduction of non-winged (apterous) individuals. These remain rather immobile on the same plant individual. Wasps parasitize aphids by laying single eggs into their body [22] . The parasitoid larva feeds internally on the aphid until the aphid dies and forms a hard-shelled “mummy” in which the parasitoid larva pupates. The adult primary parasitoid emerges from the mummy shortly after, unless it is itself parasitized by a secondary parasitoid. These secondary parasitoids attack the primary parasitoid either while the aphid is still alive (koinobiont “hyperparasitoids”) or when the aphid has developed into a mummy (idiobiont “mummy parasitoids”). In this study, we recorded four life-history traits of aphids and their primary and secondary parasitoids that we expected to vary along a plant richness gradient: (1) the proportion of winged individuals in aphid colonies, (2) the emergence proportion of parasitoids, (3) the proportion of parasitoids that are female and (4) the body mass of the two most abundant parasitoids. Wings are a costly trait for aphids that has been associated with reduced fecundity and are therefore not constantly expressed [18] , [23] . The production of winged morphs in aphids has been suggested to serve two main purposes besides the obligatory change of host plant species in the life cycles of some species: firstly, the avoidance of deteriorating nutritional conditions, and secondly, the escape from predators [23] , [24] , [25] , [26] . While wing production is subject to intra- and interspecific variation [23] , the proportion of winged aphids in a colony may reflect its nutritional or enemy environment which we anticipate to change with changing plant diversity. Whether a parasitoid larva develops and emerges successfully from the aphid mummy as an adult depends amongst other things on aphid quality [27] and resistance [28] , [29] . Therefore, if changes in the plant community affect the aphids' quality in terms of nutritional value, size or defensive abilities, these changes are expected to influence the proportion of successfully emerging parasitoids. Parasitoid body mass is generally correlated with fitness [30] , [31] , which has been shown to be controlled by host nutritional quality and especially by host size [e.g. 21] , [32] . Furthermore, some hymenopteran parasitoids are able to select the sex of their offspring at oviposition: fertilized eggs develop into females, unfertilized ones into males [30] . Because female parasitoids gain more from being large [33] , [34] , female eggs are typically laid into the highest-quality hosts [33] . Thus, heavier parasitoids with a female-biased sex ratio could be expected to emerge in favorable environments. In this study, we measured these life-history variables of aphids and parasitoids in experimental grassland communities varying in plant species richness and legume presence and show that (1) plant species loss has bottom-up cascading effects on life-history traits of organisms at higher trophic levels, (2) these cascades are mediated by diversity-related changes in host plant biomass and host plant nitrogen concentration. METHODS This study was conducted as part of the Jena Experiment, a temperate grassland biodiversity experiment in Jena, Germany [35] . The experimental site is a floodplain area close to the river Saale, which was used for agricultural cropping before the experiment started in 2002. The site has not received fertilizer since then and is mown twice per year, a typical mowing regime of these grasslands. In the 3.5×3.5 m plots that were used for this study, nine dominant plant species from semi-natural, mesophilic grasslands were sown in 2002 as monocultures and mixtures (richness levels of 1, 2, 3, 4, 6, and 9 species [36] ). The original species compositions were maintained by weeding at regular intervals [35] . The plots were arranged in four blocks with increasing distance to the river. Preliminary surveys showed that aphids could be regularly found on four of the nine species at the field site ( Anthriscus sylvestris , Arrhenatherum elatius , Phleum pratense and Trifolium pratense , plant nomenclature follows Rothmaler [37] ). Hence, only plots containing at least one of these host plant species at a minimum abundance (>5% cover in May 2006) were used for this study ( Table S1 ). These selection criteria resulted in an unbalanced design with fewer plots at the high plant-species richness end of the gradient. Therefore, we interpret our results with caution. DATA COLLECTION We identified and counted all aphids (a total of >16,000 individuals) that were part of an established colony on a host plant, including nymphs and winged morphs, in the same 0.2×3 m transect across the middle of the plots four times from May (first appearance of aphids at the field site) to August 2006 (hardly any aphids found after the fourth sampling period). All sampling campaigns were completed within about one week: two campaigns before the first mowing of the field site (around 2 June and 17 June) and two between the first and the second mowing (20 July and 2 August). Sampling was usually done in the mornings from a bench across the plots to avoid disturbing the vegetation and the aphid populations. All parasitized aphids (mummies) encountered in the transects during sampling were collected. Additionally, all mummies in a surrounding plot area (usually 4 m 2 ) were collected at the same time to be able to detect the full parasitoid community of a plot. A total of more than 3,500 mummies were collected, about one third of them from the transects. All mummies were placed individually in gelatin capsules in the field and were taken to the laboratory for rearing of the parasitoids. To induce parasitoid emergence, all mummies remaining after four months were subjected to a 3-month cold-warm-cold cycle with a minimum of 4°C. A total of about 2,000 parasitoids (about 1,700 primary parasitoids and 300 secondary parasitoids) emerged and were identified to species level. We individually weighed all 1,072 emerged individuals of the two most abundant parasitoid species (both primary parasitoid species) on a microbalance (Mettler Toledo MX5) to the nearest microgram after drying at 70°C for 2 days. Those two species were Adialytus arvicola (Stary), predominantly parasitizing the aphid Sipha maydis (aphid nomenclature follows Stresemann [38] ) on the host plant A. elatius ; and Trioxys brevicornis (Haliday), predominantly parasitizing the aphid Cavariella aegopodii on the host plant A. sylvestris . The biomass of aphid host plants was determined by clipping the vegetation at a height of 3 cm in two 20×50 cm areas within each plot in May and August 2006. The harvested biomass was sorted into species, dried at 70°C for 48 h and weighed. The biomass data were averaged over the two sampling areas and summed up over the whole year. Nitrogen concentrations of aphid host plants were measured for each plant species from biomass samples harvested in May 2004 [39] in the same plots that were used for aphid counts or in replicate plots with the same plant composition. DATA ANALYSIS The proportion of winged aphid morphs was calculated separately for each aphid species by dividing the density of winged individuals by the total density of this species. Alatiform nymphs were rarely encountered and included as non-winged morphs; parasitized aphids were excluded. The proportion of winged aphids was arcsine square-root transformed to improve normality and homoscedasticity. The emergence proportion of parasitoids was calculated as the number of emerged parasitoids divided by the sum of emerged parasitoids and remaining mummies for each aphid species. This analysis was done at the level of the aphid species because in the case of failed emergence we could not confidently distinguish between primary and secondary parasitoids or identify their species. The proportion of female parasitoids was calculated for each parasitoid species by dividing the number of female individuals by the total number of emerged individuals of this species. One parasitoid species, Lysiphlebus fabarum (Marshall) which parasitized the aphid Aphis scaliai on the legume T. pratense , reproduces predominantly asexually in Europe [40] and was excluded from the analysis of proportions of females. All data were aggregated over the four census times. We used multiple regression and ANOVA to analyze the data. Except for the proportion of winged aphid morphs the response variables were not transformed for the final analyses because model-checking procedures indicated that this was not necessary and analyses with transformed variables gave similar results. The biomass of aphid host plants (in g/m 2 ) and their nitrogen concentration (in %) were used as covariables. All analyses were carried out with the statistical software R, version 2.7.2 [41] . RESULTS Overall, host plant biomass declined with increasing plant species richness ( Table 1 ). Host plant nitrogen concentration increased with increasing plant species richness and was higher in legume (2.3±0.8%) than in non-legume plots (1.4±0.3%) and higher for legume ( T. pratense : 2.9±0.5%) than for non-legume host plants ( A. sylvestris : 2.0±0.4%, A. elatius : 1.2±0.3%, P. pratense : 1.3±0.3%). Plant species on which mummies of the two most abundant parasitoid species ( A. arvicola and T. brevicornis ) were found, also showed a decrease in biomass with increasing plant species richness (F 1,24  = 16.6, P<0.001). Their average biomass was higher in plots containing legumes (452±246 g/m 2 ) than in plots from which legumes were absent (198±123 g/m 2 , F 1,24  = 34.7, P<0.001). However, their nitrogen concentration did not significantly increase with plant species richness (F 1,18  = 1.15, P = 0.297). The term representing host plant species identity (“Host plant species”) is split into two contrasts: “non-legume vs. legume” and the term containing the remaining variance from host plant species identity (“rest”). So for example (row 4), non-legume host plant species have a higher biomass than legume host plant species but other host plant species identity effects are not significant. “Plant species richness” and “Legume presence” were tested against “Plotcode”, all other terms against “Residuals”. Directions of significant main effects (with 1 df) are indicated by arrows. P-values<0.05 are printed in bold. MS = mean square. TABLE: ANOVAs for host plant biomass and host plant nitrogen concentration. Host plant biomass Host plant nitrogen concentration df MS F P df MS F P Plant species richness 1 706360 17.3 0.0001 ↓ 1 2.21 4.1 0.0482 ↑ Legume presence 1 3422 0.1 0.7735 1 16.23 30.3 <0.0001 ↑ Plotcode 44 40806 4.8 <0.0001 43 0.54 23.3 <0.0001 Host plant species (non-legume vs. legume) 1 488148 57.0 <0.0001 ↓ 1 32.37 1407.3 <0.0001 ↑ Host plant species (rest) 2 13823 1.6 0.2061 2 0.78 34.0 <0.0001 Plant species richness×Host plant species (non-legume vs. legume) 1 261584 30.6 <0.0001 1 0.01 0.3 0.5832 Plant species richness×Host plant species (rest) 2 50123 5.9 0.0044 2 0.01 0.3 0.7387 Legume presence×Host plant species (rest) 2 19150 2.2 0.1142 1 0.03 1.2 0.2743 Residuals 71 8558 60 0.02 PROPORTION OF WINGED APHIDS Winged individuals were found in colonies of only four out of the ten aphid species ( Table S2 ): Aphis fabae , Cavariella aegopodi (both mainly on the host plant A. sylvestris ), Diuraphis muehlei (on the host plant P. pratense ) and Sipha maydis (on the host plant A. elatius ). The proportion of winged morphs declined with increasing plant species richness, both for these four aphid species ( Fig. 1 ) and when all aphid species were included in the analysis ( Table 2 ). The biomass of the aphid host plants explained part of this effect: higher host plant biomass had an increasing effect on the proportion of winged aphids, but since host plant biomass declined with plant species richness, the proportion of winged individuals declined as well. Aphid load (aphid density per host plant biomass, F 1,28  = 0.083, P = 0.775) did not have a significant effect on the proportions of winged aphids. Parasitoid density in the plots had a positive effect on the proportions of winged aphids (F 1,28  = 12.654, P = 0.001, respectively). Only four aphid species produced winged morphs and are shown here: A. fabae (circles), C. aegopodii (up-facing triangles, both species mainly on the host plant species A. sylvestris ) and D. muehlei (diamonds, on the host plant species P. pratense ) and S. maydis (down-facing triangles, on the host plant species A. elatius ). The size of the plotting symbol is proportional to the biomass (g/m 2 ) of the host plant species of the respective aphid species in the respective plot. Host plant biomass below 200 g/m 2 is represented by a symbol with a fixed minimal size. Proportions of winged aphids are arcsine square-root transformed. 95% confidence intervals are shown around the regression line. Proportions were arcsine square-root transformed prior to analyses. The first analysis (a) was done without covariables, the second analysis (b) with biomass and nitrogen concentration of aphid host plants as covariables. “Plant species richness” and “Legume presence” were tested against “Plotcode”, all other terms against “Residuals”. Directions of significant effects of the covariables are indicated by arrows. P-values<0.05 are printed in bold. % SS = percent sum of squares explained. TABLE: ANOVAs for the proportion of winged aphids. a) without covariables b) with covariables df %SS F P df %SS F P Host plant biomass 1 13.2 29.6 <0.0001 ↑ Host plant nitrogen concentration 1 0.1 0.2 0.6902 Plant species richness 1 7.0 7.1 0.0111 1 4.1 3.9 0.0546 Legume presence 1 0.0 0.0 0.9595 1 0.1 0.1 0.7865 Plotcode 42 41.9 2.1 0.0097 41 42.5 2.3 0.0082 Aphid identity 9 27.4 6.3 <0.0001 9 23.2 5.8 0.0001 Plant species richness×Aphid identity 6 2.1 0.7 0.6386 6 2.4 0.9 0.4996 Legume presence×Aphid identity 4 0.9 0.5 0.7436 3 0.7 0.5 0.6687 Residuals 43 20.7 31 13.8 PARASITOID EMERGENCE The proportion of emerged parasitoids declined with increasing plant species richness. Furthermore, the proportion was lower in plots containing legumes. However, the latter was largely due to the fact that parasitoids of aphid species with legume host plants showed very low emergence proportions ( Fig. 2 , Table 3 ). Host plant nitrogen concentration had a significant negative effect on emergence ( Table 3 ). When considering only aphid species that were hosts for the two most abundant parasitoids, A. arvicola and T. brevicornis (mummies only on non-legume host plants), emergence proportions were similar in legume and non-legume plots (F 1,25  = 0.029, P = 0.867). Closed black symbols depict parasitoids emerged from mummies on non-legume host plants ( A. sylvestris , A. elatius , P. pratense ) growing in plant communities without legumes, grey symbols depict parasitoids emerged from mummies on those non-legume host plant species in plant communities with legumes and open symbols depict parasitoids emerged from mummies on a legume host plant ( T. pratense ). The regression line (shown with 95% confidence intervals) was fitted to all data points. The first analysis (a) was done without covariables, the second analysis (b) with biomass and nitrogen concentration of aphid host plants as covariables. “Plant species richness” and “Legume presence” were tested against “Plotcode”, all other terms against “Residuals”. Directions of significant effects of the covariables are indicated by arrows. P-values<0.05 are printed in bold. % SS = percent sum of squares explained. TABLE: ANOVAs for the parasitoid emergence proportion. a) without covariables b) with covariables df %SS F P df %SS F P Host plant biomass 1 1.4 1.4 0.2591 Host plant nitrogen concentration 1 18.6 17.6 0.0007 ↓ Plant species richness 1 8.9 10.4 0.0024 1 5.9 7.4 0.0092 Legume presence 1 5.5 6.4 0.0149 1 0.0 0.1 0.8075 Plotcode 44 37.7 1.2 0.3621 43 33.9 0.7 0.7818 Aphid identity 6 23.6 5.3 0.0014 6 16.2 2.6 0.0629 Plant species richness×Aphid identity 4 4.9 1.7 0.1941 4 5.4 1.3 0.3198 Legume presence×Aphid identity 2 1.6 1.1 0.3523 1 1.6 1.5 0.2433 Residuals 24 17.8 16 16.9 PROPORTION OF FEMALE PARASITOIDS The proportion of female parasitoids was lower in the presence of legumes, irrespective of the trophic level or species identity of the parasitoid ( Fig. 3 , Table 4 ). Host plant biomass had a significant negative effect on the proportion of female parasitoids and explained part of the negative legume effect ( Table 4 ). There was no significant legume effect for the two most abundant parasitoids, A. arvicola and T. brevicornis (F 1,24  = 0.528, P = 0.475). The dashed line shows the 1∶1 sex ratio. The first analysis (a) was done without covariables, the second analysis (b) with biomass and nitrogen concentration of aphid host plants as covariables. “Plant species richness” and “Legume presence” were tested against “Plotcode”, all other terms against “Residuals”. “Trophic level” signifies the contrast between primary and secondary parasitoids. Directions of significant effects of the covariables are indicated by arrows. P-values<0.05 are printed in bold. % SS = percent sum of squares explained. TABLE: ANOVAs for the proportion of female parasitoids. a) without covariables b) with covariables df %SS F P df %SS F P Host plant biomass 1 3.6 6.5 0.0148 ↓ Host plant nitrogen concentration 1 1.6 2.9 0.0988 Plant richness 1 0.7 0.9 0.3461 1 0.8 0.9 0.3432 Legume presence 1 3.3 4.3 0.0452 1 0.3 0.3 0.5732 Plotcode 39 30.1 1.6 0.0555 37 30.9 1.5 0.1081 Trophic level 1 0.0 0.0 0.9736 1 0.4 0.8 0.3735 Parasitoid identity 18 20.5 2.3 0.0077 18 22.3 2.2 0.0191 Plant species richness×Trophic level 1 0.0 0.0 0.9086 1 0.1 0.1 0.7182 Plant species richness×Parasitoid identity 15 7.1 1.0 0.5051 14 10.1 1.3 0.2496 Legume presence×Trophic level 1 0.1 0.1 0.7156 1 0.2 0.4 0.5296 Legume presence×Parasitoid identity 12 7.9 1.3 0.2216 12 9.2 1.4 0.2157 Residuals 62 30.4 37 20.5 PARASITOID BODY MASS The body mass of the two most abundant parasitoids, A. arvicola and T. brevicornis was similar across all plant species richness levels but was higher in the presence of legumes for males and females of both parasitoid species ( Figs. 4 and S1 , Table 5 ). Because both species emerged only from mummies on non-legume host plants, these effects were indirect: they were largely due to increased biomass and nitrogen concentration of these non-legume host plants in plots containing legumes. Average dry mass of the two most abundant primary parasitoids A. arvicola and T. brevicornis ± standard error (mg) in plots without legumes (“no leg”) and with legumes (“leg”). Both parasitoid species developed only in aphids on non-legume host plant species ( A. elatius and A. sylvestris , respectively). The first analysis (a) was done without covariables, the second analysis (b) with biomass and nitrogen concentration of aphid host plants as covariables. Both parasitoid species developed only in aphids on non-legume host plant species. The covariables and “Sex” (incl. all interactions with “Sex”) were tested against “Residuals”, all other terms against “Plotcode”. “Parasitoid identity” was tested at the “Plotcode” level in this analysis because in almost all cases only one of the two species was present in a plot. Directions of significant effects of the covariables are indicated by arrows. P-values<0.05 are printed in bold. % SS = percent sum of squares explained. TABLE: ANOVAs for the body mass of the two most abundant primary parasitoid species ( A. arvicola and T. brevicornis ). a) without covariables b) with covariables df %SS F P df %SS F P Host plant biomass 1 6.1 67.1 <0.0001 ↑ Host plant nitrogen concentration 1 2.5 27.4 <0.0001 ↑ Plant species richness 1 0.3 1.4 0.2452 1 0.8 2.7 0.1228 Legume presence 1 7.7 34.6 <0.0001 1 0.6 2.2 0.1584 Parasitoid identity 1 1.7 7.7 0.0103 1 1.1 3.9 0.0664 Plant species richness×Parasitoid identity 1 0.2 0.7 0.4190 1 0.1 0.5 0.5017 Legume presence×Parasitoid identity 1 0.0 0.1 0.7088 1 0.2 0.8 0.3812 Plotcode 24 5.3 2.6 0.0001 16 4.7 3.2 <0.0001 Sex 1 2.8 32.8 <0.0001 1 2.4 26.3 <0.0001 Plant species richness×Sex 1 0.1 1.7 0.1917 1 0.2 1.7 0.1894 Legume presence×Sex 1 0.2 2.0 0.1551 1 0.2 2.3 0.1297 Parasitoid identity×Sex 1 0.7 8.5 0.0036 1 0.5 5.9 0.0158 Residuals 936 80.9 877 80.5 DISCUSSION Our results demonstrate that plant species richness and the presence of legumes in plant communities affect the life history of aphids and their parasitoids both directly and indirectly via host plant biomass and nitrogen concentration. These effects are paralleled by community-level cascading effects of plant species richness on species richness and densities of aphids and their parasitoids that we reported previously from the same field site [11] . The proportion of winged aphids decreased with increasing plant species richness, partly as a result of a simultaneous decrease in aphid host plant biomass. Whereas low host plant abundance typically has a negative effect on herbivore densities [5] , [6] , [11] , [42] , the proportion of winged individuals was expected to rise with resource limitation since wing production after the initial colonization period in early spring is a strategy to escape deteriorating conditions by dispersing [23] , [25] . In addition to the unexpected direction of the effect, we found no influence of crowding (aphid load) or of plant quality (in terms of plant nitrogen concentration) on winged morph production, suggesting that limited resource availability was not the cause for decreasing proportions of winged morphs with increasing plant species richness. However, Züst et al. [43] suggested that resources might sometimes be too low to induce wing production at all. So the resource-limitation argument could be reversed to explain the extremely low proportions of winged aphids in communities with high plant species richness and low host plant biomass. Parasitoid density had an increasing effect on aphid wing production indicating that increased pressure from natural enemies might have played a role in this context. However, we found relevant proportions of winged individuals in only four aphid species. It has been shown that intra- and interspecific variation in wing induction can be very large [23] . Furthermore, other abiotic effects that we did not measure (e.g. temperature) may be important triggers [23] , cautioning against drawing strong conclusions from the results. PARASITOID EMERGENCE AND PROPORTION OF FEMALE PARASITOIDS Emergence proportions of parasitoids similarly decreased with increasing plant species richness. However, our design was unbalanced, because of a disproportionately low number of plant communities at high species richness levels. Therefore, we have to interpret our results with caution. Both emergence proportions and proportions of females were lower in plant communities containing legumes compared with legume-free communities. These relationships were partially explained by changes in host-plant nitrogen concentration (parasitoid emergence) and in host plant biomass (proportion of females). Plant nitrogen concentrations, which are often increased even in non-legumes by the presence of nitrogen-fixing legumes [44] , [45] , can have positive or negative effects on higher trophic levels. On the one hand, plants with higher nitrogen concentrations may be more nutritious for herbivores, which are generally nitrogen-limited [46] , [47] , [48] , [49] . On the other hand, nutrient availability is known to affect plant allocation to secondary metabolites [50] , [51] and plants may use surplus nitrogen to produce higher levels of nitrogen-based defense compounds [51] , [52] . One trophic level above herbivores, the same dichotomy may become apparent. Natural enemies of herbivores are even more strongly limited by nitrogen than the herbivores themselves [53] , [54] and might do better on nitrogen-rich hosts. However, some herbivores can use nitrogen-based plant compounds or simply their own high nutritional status on nitrogen-rich host plants to mount defenses against predators and parasitoids [55] , [56] . While some of the early defense reactions (encapsulation and abortion of parasitoid eggs) probably went unnoticed in our study, the lower parasitoid emergence success on nitrogen-rich host plants could still be due to increased aphid defenses. Since mated parasitoid females can determine the sex of their offspring at oviposition [30] , we speculate that they might have sensed a potentially different host quality and avoided “wasting” females on risky hosts. On the other hand, the parasitoid sexes might have differential mortality [30] , with female larvae more prone to abortion as a result of lower host quality, potentially explaining the lower emergence success in legume plots. Interestingly, even secondary parasitoids showed the effect of legumes on the proportions of females, implying a cascade to the third trophic level above the plants. PARASITOID BODY MASS In contrast to general parasitoid emergence and proportions of females, the body mass of the two most abundant primary parasitoids, A. arvicola and T. brevicornis , was higher in the presence of legumes in the plant community. Host plant nitrogen concentration and host plant biomass had a positive effect in this case. Since we did not find a negative effect of legume presence on emergence or the proportion of females in these two species, we conclude that they might be resistant to potentially increased aphid defenses in legume plots. This resistance may even explain their high abundance at our field site. The two species seemed to instead profit from the presence of legumes. We did not measure aphid body mass or nitrogen content and hence have no conclusive evidence for either higher nitrogen content of hosts [53] , [54] , [57] or larger host size indeed mediating higher parasitoid body mass in communities with legumes. Another field experiment found a higher body-mass gain of herbivores (grasshoppers) in legume-containing plant communities [58] . Since host size is typically strongly correlated with parasitoid size [59] , a size-mediated effect seems plausible. CONCLUSIONS We found considerable effects of plant species richness and community composition cascading up the food web to affect aphid and parasitoid life-history traits. Legumes exerted a particularly strong effect in some cases. This plant functional group had an impact up to the highest trophic level considered in this study, sometimes mediated via an additional interaction at the plant level, thus affecting even insect communities on non-legume host plants. In contrast, we found only plant-species richness effects but no effects of legume presence on aphid and parasitoid richness and densities in a community-level study at the same field site [11] , underlining the importance of life-history studies to detect different pathways of effects of plant species loss on associated biodiversity. This difference in the effect of legume presence between the two studies may indicate that cascading bottom-up effects on insect life-history traits are more strongly resource-quality mediated, whereas cascading effects on insect species richness and density seem to be rather mediated by abundance and richness of lower trophic levels. On the other hand, it is possible that insect life-history studies detect effects of plant species loss at an earlier stage than richness or abundance studies. Similar observations have been made in analyses of vegetation responses to habitat change, where life-history changes could be used as early indicators [60] . The altered life-history traits of aphids and parasitoids are expected to eventually result in altered species abundances and richness at these consumer trophic levels. One example for this link is the apparently low suitability of species-rich plant communities for parasitoids. In the present study, we found for example lower emergence success, potentially due to increased aphid defenses. These are paralleled by low parasitoid densities and species richness [11] , possibly a result of reduced parasitoid reproduction. We therefore advocate the additional examination of individual-level cascades to predict and explain population- and community-level cascades of plant species loss on higher trophic levels. SUPPORTING INFORMATION Body mass of two primary parasitoids in communities with and without legumes. Average dry mass of a) females and b) males of the two most abundant primary parasitoids A. arvicola and T. brevicornis ± standard error (mg) in plots without legumes (“no leg”) and with legumes (“leg”). Both parasitoid species developed only in aphids on non-legume host plant species ( A. elatius and A. sylvestris , respectively). (0.06 MB TIF) List of all 47 sampled plots. For each plot, block number, plot number, plant species richness, the presence of legumes, the presence of the nine plant species in the plant assemblage and the total host plant biomass (sum across all aphid host plant species and both harvests in 2006) are given. In some cases, two replicate plots with the same plant composition were used. The presence of a plant species refers to its presence in the seed mixture at plot establishment in 2002, independent of its actual abundance through 2006. Only the first four plant species hosted aphids in our study. Ant: Anthriscus sylvestris, Arr: Arrhenatherum elatius, Phl: Phleum pratense, Tri p: Trifolium pratense, Alo: Alopecurus pratensis, Dac: Dactylis glomerata, Ger: Geranium pratense, Poa: Poa trivialis, Tri r: Trifolium repens. (0.07 MB DOC) List of plant and insect species. For each species, average densities (biomass in g per m2 for plants, individuals per m2 for aphids and parasitoids, sums over all sampling dates) are given. Furthermore, proportions of winged aphids, proportions of emerged parasitoids, proportions of female parasitoids and parasitoid body mass (mg) across all respective plots are given. All means are shown ±1 standard error. Plant nomenclature follows Rothmaler [1] , aphid nomenclature follows Stresemann [2] . Authorities for parasitoids are given in parentheses behind species names. Two rare aphid species and one rare parasitoid species could not be identified due to the lack of material and were assigned to morphospecies. Three species of Alloxysta have not been described and were given provisional names (Frank van Veen, personal communication). excl.: Lysiphlebus fabarum reproduces asexually in Europe (i.e. the proportion of females is aways 1) and was excluded from the analysis of proportions of females; n.a.: not available ( Aphis fabae and the unidentified aphid species were never parasitized in our study; the sex of the unidentified parasitoids could not be determined). References: 1. Rothmaler R (2002) Exkursionsflora von Deutschland; Jäger EJ, Werner K, editors. Heidelberg-Berlin: Spektrum. 2. Stresemann E (1994) Exkursionsfauna von Deutschland, Wirbellose: Insekten- 2.Teil. Jena: Gustav Fischer Verlag. (0.06 MB DOC) REFERENCES [1] Duffy JE Cardinale BJ France KE McIntyre PB Thebault E 2007 The functional role of biodiversity in ecosystems: incorporating trophic complexity. Ecol Lett 10 522 538 [2] Schmid B Balvanera P Cardinale BJ Godbold J Pfisterer AB 2009 Consequences of species loss for ecosystem functioning: meta-analyses of data from biodiversity experiments. Naeem S Bunker DE Hector A Loreau M Perrings C Biodiversity, Ecosystem Functioning, and Human Wellbeing: An Ecological and Economic Perspective Oxford Oxford University Press 14 29 [3] Hunter MD Price PW 1992 Playing chutes and ladders: Heterogeneity and the relative roles of bottom-up and top-down forces in natural communities. Ecology 73 724 732 [4] Kagata H Ohgushi T 2006 Bottom-up trophic cascades and material transfer in terrestrial food webs. Ecol Res 21 26 34 [5] Haddad NM Tilman D Haarstad J Ritchie M Knops JMH 2001 Contrasting effects of plant richness and composition on insect communities: A field experiment. Am Nat 158 17 35 [6] Knops JMH Tilman D Haddad NM Naeem S Mitchell CE 1999 Effects of plant species richness on invasion dynamics, disease outbreaks, insect abundances and diversity. Ecol Lett 2 286 293 [7] Siemann E Tilman D Haarstad J Ritchie M 1998 Experimental tests of the dependence of arthropod diversity on plant diversity. Am Nat 152 738 750 [8] Dyer LA Stireman JO 2003 Community-wide trophic cascades and other indirect interactions in an agricultural community. Basic Appl Ecol 4 423 432 [9] Gange AC Brown VK Aplin DM 2003 Multitrophic links between arbuscular mycorrhizal fungi and insect parasitoids. Ecol Lett 6 1051 1055 [10] Aquilino KM Cardinale BJ Ives AR 2005 Reciprocal effects of host plant and natural enemy diversity on herbivore suppression: an empirical study of a model tritrophic system. Oikos 108 275 282 [11] Petermann JS Müller CB Weigelt A Weisser WW Schmid B 2010 Effects of plant species loss on aphid-parasitoid communities. J Anim Ecol 79 709 720 [12] Harvey JA van Dam NM Gols R 2003 Interactions over four trophic levels: foodplant quality affects development of a hyperparasitoid as mediated through a herbivore and its primary parasitoid. J Anim Ecol 72 520 531 [13] Soler R Bezemer TM van der Putten WH Vet LEM Harvey JA 2005 Root herbivore effects on above-ground herbivore, parasitoid and hyperparasitoid performance via changes in plant quality. J Anim Ecol 74 1121 1130 [14] Wurst S Jones TH 2003 Indirect effects of earthworms ( Aporrectodea caliginosa ) on an above-ground tritrophic interaction. Pedobiologia 47 91 97 [15] de Sassi C Müller CB Krauss J 2006 Fungal plant endosymbionts alter life history and reproductive success of aphid predators. Proceedings of the Royal Society B 273 1301 1306 [16] Härri SA Krauss J Müller CB 2008 Trophic cascades initiated by fungal plant endosymbionts impair reproductive performance of parasitoids in the second generation. Oecologia 157 399 407 [17] Härri SA Krauss J Müller CB 2008 Fungal endosymbionts of plants reduce lifespan of an aphid secondary parasitoid and influence host selection. Proceedings of the Royal Society B 275 2627 2632 [18] Dixon AFG 1998 Aphid Ecology London Chapman and Hall [19] Dixon AFG Kindlmann P Leps J Holman J 1987 Why there are so few species of aphids, especially in the tropics. Am Nat 129 580 592 [20] van Veen FJF Müller CB Pell JK Godfray HCJ 2008 Food web structure of three guilds of natural enemies: predators, parasitoids and pathogens of aphids. J Anim Ecol 77 191 200 [21] Bukovinszky T van Veen FJF Jongema Y Dicke M 2008 Direct and indirect effects of resource quality on food web structure. Science 319 804 807 [22] Völkl W Mackauer M Pell JK Brodeur J 2007 Predators, parasitoids and pathogens. Van Emden HF Harrington RH Aphids as Crop Pests Oxford CAB International 187 233 [23] Müller CB Williams IS Hardie J 2001 The role of nutrition, crowding and interspecific interactions in the development of winged aphids. Ecol Entomol 26 330 340 [24] Sloggett JJ Weisser WW 2002 Parasitoids induce production of the dispersal morph of the pea aphid, Acyrthosiphon pisum . Oikos 98 323 333 [25] Sutherland ORW 1967 Role of host plant in production of winged forms by a green strain of pea aphid Acyrthosiphon pisum Harris. Nature 216 387 388 [26] Weisser WW Braendle C Minoretti N 1999 Predator-induced morphological shift in the pea aphid. Proceedings of the Royal Society B 266 1175 1182 [27] Ashouri A Michaud D Cloutier C 2001 Recombinant and classically selected factors of potato plant resistance to the Colorado potato beetle, Leptinotarsa decemlineata , variously affect the potato aphid parasitoid Aphidius nigripes . Biocontrol 46 401 418 [28] Ferrari J Müller CB Kraaijeveld AR Godfray HCJ 2001 Clonal variation and covariation in aphid resistance to parasitoids and a pathogen. Evolution 55 1805 1814 [29] Li S Falabella P Giannantonio S Fanti P Battaglia D 2002 Pea aphid clonal resistance to the endophagous parasitoid Aphidius ervi . J Insect Physiol 48 971 980 [30] Godfray HCJ 1994 Parasitoids: Behavioral and Evolutionary Ecology; Krebs J Clutton-Brock TH Princeton, New Jersey Princeton University Press [31] Visser ME 1994 The importance of being large: The relationship between size and fitness in females of the parasitoid Aphaereta minuta (Hymenoptera: Braconidae). J Anim Ecol 63 963 978 [32] Sequeira R Mackauer M 1992 Nutritional ecology of an insect host-parasitoid association: The pea aphid- Aphidius ervi s ystem. Ecology 73 183 189 [33] Charnov EL Los-den Hartogh RL Jones WT van den Assem J 1981 Sex ratio evolution in a variable environment. Nature 289 27 33 [34] Ueno T 1999 Host-size-dependent sex ratio in a parasitoid wasp. Researches on Population Ecology 41 47 57 [35] Roscher C Schumacher J Baade J Wilcke W Gleixner G 2004 The role of biodiversity for element cycling and trophic interactions: an experimental approach in a grassland community. Basic Appl Ecol 5 107 121 [36] Roscher C Temperton VM Scherer-Lorenzen M Schmitz M Schumacher J 2005 Overyielding in experimental grassland communities - irrespective of species pool or spatial scale. Ecol Lett 8 419 429 [37] Rothmaler R 2002 Exkursionsflora von Deutschland; Jäger EJ Werner K Heidelberg-Berlin Spektrum [38] Stresemann E 1994 Exkursionsfauna von Deutschland, Wirbellose: Insekten- 2.Teil Jena Gustav Fischer Verlag [39] Roscher C Thein S Schmid B Scherer-Lorenzen M 2008 Complementary nitrogen use among potentially dominant species in a biodiversity experiment varies between two years. J Ecol 96 477 488 [40] Belshaw R Quicke DLJ Völkl W Godfray HCJ 1999 Molecular markers indicate rare sex in a predominantly asexual parasitoid wasp. Evolution 53 1189 1199 [41] R Development Core Team 2008 R: A Language and Environment for Statistical Computing. 2.7.2 ed Vienna, Austria Foundation for Statistical Computing [42] Otway SJ Hector A Lawton JH 2005 Resource dilution effects on specialist insect herbivores in a grassland biodiversity experiment. J Anim Ecol 74 234 240 [43] Züst T Härri SA Müller CB 2008 Endophytic fungi decrease available resources for the aphid Rhopalosiphum padi and impair their ability to induce defences against predators. Ecol Entomol 33 80 85 [44] Temperton VM Mwangi PN Scherer-Lorenzen M Schmid B Buchmann N 2007 Positive interactions between nitrogen-fixing legumes and four different neighbouring species in a biodiversity experiment. Oecologia 151 190 205 [45] Wacker L Baudois O Eichenberger-Glinz S Schmid B 2009 Effects of plant species richness on stand structure and productivity. Journal of Plant Ecology 2 95 106 [46] Lawton JH McNeill S 1979 Between the devil and the deep blue sea: on the problems of being an herbivore. Anderson RM Turner BD Taylor LR Population dynamics Oxford Blackwell Scientific 223 244 [47] White TCR 1984 The abundance of invertebrate herbivores in relation to the availability of nitrogen in stressed food plants. Oecologia 63 90 105 [48] Davidson AW Potter DA 1995 Response of plant-feeding, predatory, and soil-inhabiting invertebrates to Acremonium endophyte and nitrogen-fertilization in tall fescue turf. J Econ Entomol 88 367 379 [49] Kempel A Brandl R Schadler M 2009 Symbiotic soil microorganisms as players in aboveground plant-herbivore interactions - the role of rhizobia. Oikos 118 634 640 [50] Bryant JP Chapin FS Klein DR 1983 Carbon/nutrient balance of boreal plants in relation to vertebrate herbivory. Oikos 40 357 368 [51] Mattson WJ 1980 Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics 11 119 161 [52] Lyons PC Plattner RD Bacon CW 1986 Occurrence of peptide and clavine ergot alkaloids in tall fescue grass. Science 232 487 489 [53] Mayntz D Toft S 2001 Nutrient composition of the prey's diet affects growth and survivorship of a generalist predator. Oecologia 127 207 213 [54] Fagan WF Siemann E Mitter C Denno RF Huberty AF 2002 Nitrogen in insects: Implications for trophic complexity and species diversification. Am Nat 160 784 802 [55] Francis F Lognay G Wathelet J-P Haubruge E 2001 Effects of allelochemicals from first (Brassicaceae) and second ( Myzus persicae and Brevicoryne brassicae ) trophic levels on Adalia bipunctata . J Chem Ecol 27 243 256 [56] Barbosa P Gross P Kemper J 1991 Influence of plant allelochemicals on the tobacco hornworm and its parasitoid, Cotesia congregata . Ecology 72 1567 1575 [57] Hunter MD 2003 Effects of plant quality on the population ecology of parasitoids. Ecol Lett 5 1 8 [58] Pfisterer AB Diemer M Schmid B 2003 Dietary shift and lowered biomass gain of a generalist herbivore in species-poor experimental plant communities. Oecologia 135 234 241 [59] Cohen JE Jonsson T Muller CB Godfray HCJ Savage VM 2005 Body sizes of hosts and parasitoids in individual feeding relationships. Proceedings of the National Academy of Sciences 102 684 689 [60] Bühler C Schmid B 2001 The influence of management regime and altitude on the population structure of Succisa pratensis : implications for vegetation monitoring. J Appl Ecol 38 689 698