Direct and indirect effects of landscape structure on a ...

Direct and indirect effects of landscape structure on a tri-trophic system within agricultural lands SIMON P. DAOUST,1,4, MARC BE´LISLE,2 JADE SAVAGE,3 AUDREY ROBILLARD,2 RENAUD BAETA,2 1

AND JACQUES

BRODEUR1

Institut de recherche en biologie ve´ge´tale, De´partement de sciences biologiques, Universite´ de Montreál, Montreál, Que´bec, Canada H1X 2B2 2 De´partement de biologie, Universite´ de Sherbrooke, Sherbrooke, Que´bec, Canada J1K 2R1 3 Department of biological sciences, Bishop’s University, Sherbrooke, Que´bec, Canada J1M 1Z7

Citation: Daoust, S., M. Be´lisle, J. Savage, A. Robillard, R. Baeta, and J. Brodeur. 2012. Direct and indirect effects of landscape structure on a tri-trophic system within agricultural lands. Ecosphere 3(11):94. http://dx.doi.org/10.1890/ ES12-00300.1

Abstract. Although several studies have examined the influence of landscape structure and agricultural intensification on species abundance and diversity, few have addressed how these impact populations across multiple trophic levels. We investigated the effects of landscape structure on the tri-trophic interactions between a bird host (the Tree Swallow Tachycineta bicolor (Vieillot)), its blowfly ectoparasites (Protocalliphora Hough), and their parasitoid wasps (Nasonia Ashmead) across 13 spatial scales, along a gradient of agricultural intensification covering 10,200 km2 in southern Que´bec, Canada. We showed that the three taxa responded to landscape structure at distinctive spatial scales that are relative to their size rather than their trophic rank. This response, however, differed according to habitat type. The three organisms responded to the amount of intensive cultures (maize, soybean and other cereals) at smaller spatial scales than to the amount of extensive cultures (hayfields and pastures). Although the number of Tree Swallow fledglings, the number of Protocalliphora sialia Shannon & Dobroscky pupae and the number of P. sialia pupae parasitized by Nasonia sp. per nest were negatively affected by agricultural intensification, our data do not support the prediction that organisms at the higher trophic levels are more susceptible to habitat degradation. Here, ectoparasites at the second trophic level were disproportionally affected by agricultural intensification; the abundance of P. sialia decreased by 80% along the gradient of agricultural intensification compared to a 20% and 35% reduction in the number of Tree Swallow fledglings and in the level of parasitism by Nasonia sp., respectively. Our work highlights the importance of designing protocols that take spatial aspects of trophic interactions into account when studying the impact of habitat loss and fragmentation on populations and communities, as these interactions dictate local biodiversity and community function. Furthermore, our results highlight the importance of considering multiple landscape parameters when identifying the functional spatial scales of an organism, as a failure to do so could lead to an underestimation of the area it uses. Key words: agricultural intensification; functional spatial scale; landscape structure; Nasonia; Protocalliphora; Tree Swallow; tri-trophic interactions. Received 21 September 2012; accepted 1 October 2012; published 2 November 2012. Corresponding Editor: O. Schmitz. Copyright: Ó 2012 Daoust et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits restricted use, distribution, and reproduction in any medium, provided the original author and sources are credited. Current address: Aquatic Biodiversity Section, Watershed Hydrology and Ecology Research Division, St. Lawrence

4

Centre, Environment Canada, Montreál, Que´bec, Canada H2Y 2E7. E-mail: [email protected]

v www.esajournals.org

1

November 2012 v Volume 3(11) v Article 94

DAOUST ET AL.

community whereas predators exploit several prey populations and therefore must switch between patches colonized by prey (Holt 1996, Thies et al. 2003, Tscharntke et al. 2005). While this holds true for invertebrate-vertebrate food chains where species at the highest trophic rank are larger, the predictions by Holt (1996) are not supported in host-parasite/parasitoid interactions where species at the highest trophic ranks are usually much smaller (Thies et al. 2003, 2005). Notwithstanding size, species at higher trophic levels may be more susceptible to extinction after habitat loss, fragmentation and degradation due to agricultural intensification as they are confronted with both the direct effect on their populations and the indirect effects on their host or prey populations (Cagnolo et al. 2009). Within each trophic level mechanisms driving species’ response to landscape structure are species specific (Fisher et al. 2006). This specificity is often attributed to behavioral differences, degrees of habitat specialization and levels of adaptive flexibility (Fisher et al. 2006). Accordingly, in order to understand the relationship between landscape structure and species-specific responses, research protocols must include components and metrics relevant to the target species and landscape metrics should be quantified at the appropriate spatial scales (Thies et al. 2003, Wheatley and Johnson 2009, de Knegt et al. 2010, Wheatley 2010). In this two-year study, we investigated the effects of landscape structure on the tri-trophic interactions between a bird host (the Tree Swallow Tachycineta bicolor (Vieillot)), its blowfly ectoparasites (Protocalliphora Hough), and their parasitoid wasps (Nasonia Ashmead) along a gradient of agricultural intensification covering 10,200 km2 in southern Que´bec, Canada. The gradient is characterized by significant differences in landscape composition, notably in crop diversity and the relative area covered by extensive and intensive cultures as well as forest habitat. Extensive cultures mostly include habitats related to dairy and beef cattle farming, such as hayfields, pastures and fallows, which are typically surrounded by marginal habitats like hedgerows, forest patches, and wetlands. Intensive cultures, on the other hand, are mostly composed of large monocultures of annual row crops, such as maize, soybean and other cereals

INTRODUCTION The significance of spatial context in understanding local ecological processes is increasingly recognized by ecologists (Wiens 1989, Wheatley and Johnson 2009, Fahrig et al. 2011). Species abundance and diversity within a patch of habitat, and thereby trophic interactions and other community function therein, may rely on processes operating at larger spatial scales (Tscharntke and Brandl 2004, Fisher et al. 2005, Thies et al. 2005, Bianchi et al. 2006, Holzschuh et al. 2010). The surrounding matrix of a habitat fragment may increase the amount of available resources or provide additional resources to those within the local habitat fragment. It can also modulate the connectivity of habitat patches and potentially influence meta and local population structure and dynamics (Marshall and Moonen 2002, Steffan-Dewenter 2003, Tscharntke et al. 2005, Fahrig 2007). This notion is especially important in the context of recent and major changes to the agricultural landscape (Robinson and Sutherland 2002). Formerly heterogeneous, extensively cultured lands composed of complex landscapes with well-balanced proportions of arable lands, grasslands, forests, fallows, hedgerows and other semi-natural habitats are being transformed to homogenous intensively cultured lands with simple landscapes containing only fragments of natural or semi-natural land (Matson et al. 1997). Although many studies have quantified the impacts of landscape structure and agricultural intensification on species abundance and diversity (Burel et al. 1998, Benton et al. 2003, Burel et al. 2004, Tscharntke et al. 2005, Donald et al. 2006), few have empirically addressed how agriculture impacts populations across multiple trophic levels. It is relatively well understood that the responses of organisms to changes in landscape structure are contingent on the species’ size and trophic level (Holt 1996, Holt et al.1999, Tscharntke et al. 2005, Thies et al. 2005). Within the framework developed by Holt (1996), organisms at higher trophic levels are predicted to respond to larger spatial-scale patterns and processes. For instance, in a plant-herbivorepredator food chain, abundance of plant species is determined by its microhabitat, herbivores are wider-ranging, but confined within the local v www.esajournals.org

2


DAOUST ET AL.

which require increased mechanization and high input of chemical fertilizers and pesticides. Their relative abundance is also tightly and negatively correlated with the amount of forest cover in the landscape over a large array of spatial scales (Be´langer and Grenier 2002, Jobin et al. 2005). The nature and management of extensive cultures, as well as their proximity to natural habitats, thus provide more abundant and diversified resources that are easier to reach and exploit as compared to intensive cultures (Tscharntke et al. 2005; Rioux Paquette et al., in press). Previous work within our study system reported that nest box occupancy and breeding success of Tree Swallows, which are aerial insectivores, are negatively correlated with the proportion of intensive culture in the landscape (Ghilain and Be´lisle 2008), likely as a result of lower prey abundance (Rioux Paquette et al., in press). Similarly, Jobin et al. (1996) observed that populations of many bird species are decreasing within the intensively cultivated lands of southern Que´bec; a pattern found across North America (Murphy 2003) and Europe (Donald et al. 2001). Although very little is known of the ecology of Protocalliphora, it is well established that the diversity and abundance of many flying insects, such as non-biting midges, sawflies and moths, are negatively affected by agricultural intensification (Burel et al. 1998, 2004, Wickramasinghe et al. 2004, Schweiger et al. 2005). Parasitoids do not fare any better, as several studies report decreased levels of parasitoid diversity and parasitism within intensively cultivated landscapes (Steffan-Dewenter 2003, Thies et al. 2003, 2004, Tscharntke and Brandl 2004, Tscharntke et al. 2005, Bianchi et al. 2006). These decreases in abundance and diversity of various insect taxa were attributed to lower host diversity and abundance, fewer sources of nectar and shelter sites, reduced functional connectivity that impedes foraging and dispersal movements, and the application of pesticides (Burel et al. 1998, 2004, Tscharntke and Brandl 2004, Wickramasinghe et al. 2004, Tscharntke et al. 2005). The main objectives of the study were to (1) determine the spatial scales at which the organisms maximally respond to landscape composition, (2) measure the influence of landscape composition at the functional spatial scale(s) v www.esajournals.org

identified in objective 1 on the number of Tree Swallow fledglings (proxy for resources available to Protocalliphora larvae), the abundance of Protocalliphora in Tree Swallow nests (proxy for the level of ectoparasitism by Protocalliphora), and the number of Protocalliphora pupae parasitized by Nasonia per nest (level of parasitism by Nasonia), and (3) measure the impact of host availability on the abundance of Protocalliphora and levels of parasitism by Nasonia. Firstly, we predicted that habitat loss associated with an increase in the proportion of intensive agriculture in the landscape leads to reductions in the number of Tree Swallow fledglings, the abundance of Protocalliphora pupae per nest and the level of parasitism by Nasonia. Secondly, as Protocalliphora abundance per nest is positively correlated with bird clutch size (Eeva et al. 1994), and levels of parasitism are generally positively correlated with host density (Hassel 1982), we predicted that both Protocalliphora abundance per nest and the levels of parasitism by Nasonia increase with host abundance regardless of the landscape structure. Lastly, as higher trophic levels may be more negatively affected by the strong ‘‘bottom-up’’ effects associated with landuse change (habitat loss and reduced host abundance) (Thies et al. 2003, Cagnolo et al. 2009), we predicted that the negative effects of agricultural intensification should be greater with increasing trophic rank.

MATERIALS

AND

METHODS

Study organisms The Tree Swallow, an obligate secondary cavity nester, is primarily insectivorous (Winkler et al. 2011). Its breeding home range fluctuates throughout the season: prior to incubation, the range spans .60 km2, but after incubation, it falls to 2–5 km2 (Winkler et al. 2011). Larvae of Holarctic bird blow f1ies are obligate blood-feeding parasites of nestling birds. Adult Protocalliphora are rarely collected in the field, while immatures (larvae and pupae) are commonly found following close examination of nestlings or nest materials. Feeding larvae are typically anchored to the most accessible part of nestlings (feet, legs and belly), although some species can be found in the auditory and nasal cavities (Whitworth and Bennett 1992). In most 3


DAOUST ET AL.

cases, parasitized nestlings have lower hematocrit and haemoglobin levels (Whitworth and Bennett 1992), decreased body temperature and metabolic rates (Simon et al. 2005), reduced growth rates (Whitworth and Bennett 1992), lower fledging survival and reduced dispersal in the first days following fledging (Streby et al. 2009). Nasonia are small insects (approximately 2 mm in length) commonly referred to as ‘‘jewel wasps’’. There are currently four identified species in the genus Nasonia: N. vitripennis (Walker), N. giraulti Darling, N. longcornis Darling and N. oneida Raychoudhury & Desjardins (Darling and Werren 1990, Raychoudhury et al. 2010). Adult females sting and lay their eggs within the puparium of various fly species. Adult wasps typically emerge 2–3 weeks following oviposition, depending on temperature. Whereas N. vitripennis is a generalist that parasitizes blowflies, fleshflies, houseflies and others, the other three species preferentially parasitize Protocalliphora blowflies (Werren and Loehlin 2009).

Nest box monitoring We monitored Tree Swallows during each breeding season (early May to mid-July) of 2008 and 2009. Each year, all nest boxes were visited every two days to determine occupancy, laying date of first egg, clutch size, brood size at hatching, and number of chicks fledged. We considered a nest box occupied when it contained at least one egg. We determined the laying date of the first egg based on the assumption that Tree Swallows lay one egg per day until clutch completion (Winkler et al. 2011).

Insect collection and identification At the end of their breeding season, and prior to fall migration, Tree Swallows typically assess the quality (e.g., size, presence of dead nestlings, and amount of feces on walls) of the nests that were used during the breeding season to ascertain potential future breeding sites (Winkler et al. 2011), a behavior called ‘‘prospection’’ (Doliguez et al. 2004). In order to avoid interference with the bird’s habitat quality assessment, insect specimens were collected using the following protocol: 2 out of the potential 10 available Tree Swallow nests in each of the 40 farms were examined for Protocalliphora pupae immediately upon fledging, i.e., between June 25th and July 16th in 2008 and 2009. Nests were carefully removed from the box and all pupae present in nest material and in the nest box were collected. Nests were put back into the box upon completion of sampling. Rearing and storing protocols are detailed in Daoust et al. (2012). All Protocalliphora adults and pupae and Nasonia specimens were examined and photographed for measurements necessary to identification using a stereoscope equipped with a digital camera. Measurements were taken using PixeLINK (PixeLink, Ottawa, Que´bec, Canada) imaging software. The second sampling protocol took place during the winters of 2008–2009 and 2009–2010, several months after the birds had fledged. All nests, including those sampled in the first place, were returned to the laboratory for examination. Empty Procalliphora puparia were removed from nest material, air dried and subsequently stored at room temperature in vials. Pupae from which insects failed to emerge were dissected and examined. Specimens were keyed to species

Study region and nest box network The 10,200 km2 study area included a network of 400 nest boxes distributed equally among 40 farms within the Monte´re´gie and Estrie regions in southern Que´bec, Canada (Fig. 1). The area is characterized by an east–west gradient of agricultural intensification, where dairy farming and small-scale, familial farms have been replaced by large-scale, continuous row cropping with full mechanization and high input of pesticides, organic and chemical fertilizers (Be´langer and Grenier 2002; Fig. 1). Nest boxes were built according to North American Bluebird Society’s specifications (i.e., Eastern/Western Bluebird model) and were initially installed in the winter of 2004. Boxes were put up 50 m apart along drainage ditches or fence lines that bordered agricultural fields or pastures. All boxes were mounted on a metal post 1.5 m above the ground and with the opening facing southeast (Ghilain and Be´lisle 2008). Farms were equally distributed between intensive and extensive agricultural zones. See Ghilain and Be´lisle (2008) for the detailed farm selection protocol. Boxes were cleaned from all nest material after each breeding season, when swallows had left for their wintering grounds. v www.esajournals.org

4


DAOUST ET AL.

Fig. 1. Distribution of the 40 farms along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. Land cover types are based on a mosaic of classified LANDSAT-TM satellite images (Canadian Wildlife Service 2004) and include water (black), urban (dark gray), forest (mid-tone gray), extensive cultures (e.g., hayfields and pastures; light gray), and intensive cultures (e.g., maize, cereals, and soybean; white). Open pentagons indicate farm locations. Coordinates are Lambert Conic Conform and refer to the number of meters from a reference point.

within 13 radii: 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 1, 2, 3, 4, 5, 10, and 20 km. We assessed landscape composition up to the 0–5 km radius on a yearly basis using both visual field characterisation and high resolution (1:40,000) orthophotos. Landscape composition beyong the 0.5-km radius was based on a mosaic of georeferenced classified Landsat-7 satellite images taken between August 1999 and May 2003 (pixel resolution 25 m 3 25 m; Canadian Wildlife Service 2004). Although crop rotation modifies landscape composition across years in our study area, such temporal variation was negligible for broad land cover categories such as those we use (i.e., extensive vs. intensive) when measured within

following Sabrosky et al. (1989) and Whitworth (2003) for Protocalliphora, and following Darling and Werren (1990) and Raychoudhury et al. (2010) for Nasonia.

Landscape characterization In our study area, extensive cultures include hayfields, pastures, and fallows, while intensive cultures refer to monocultures of annual row crops, such as maize, soybean, and other cereals (Be´langer and Grenier 2002, Jobin et al. 2005). We characterized the composition of the landscape surrounding each nest box by measuring the relative cover (proportion of the area) of intensive and extensive cultures (Burel et al. 2004) v www.esajournals.org

5


DAOUST ET AL.

radii 1 km based on land cover analyses of the Base de donneés des cultures assureés (Insured Crop Database) of La Financie`re agricole du Que´bec (2011). We determined the range of spatial scales to consider based on Ghilain and Be´lisle (2008) who reported that Tree Swallow fledging success responded to such broad spatial scales. We included smaller radii (0.05–0.5 km) in the study as levels of parasitism by parasitoid wasps respond to landscape structure at smaller spatial scales than birds (Steffan-Dewenter 2003, Holzschuh et al. 2010). Measurements of relative land cover were obtained with ArcView GIS Spatial Analyst 2.0a (ESRI 2005).

presence of exit holes in Protocalliphora pupae obtained during the second sampling event rather than confirmed through rearing of pupae obtained during the first sampling. For this reason, we modelled the level of parasitism by Nasonia sp. on P. sialia. Due to the hierarchical sampling design (i.e., nest boxes nested within farms), we modeled the three response variables using generalized linear mixed models with a log link function Poisson error distribution, and farm identity as a random factor. Models were fitted by maximum likelihood with the lmer function (package lme4 version 0.999375-36) in R (version 2.1.11; R Development Core Team 2010). Statistical analyses We built a series of five competing models for In order to facilitate interpretation of the each response variable discussed above: (1) Nest observed patterns of abundances across the 13 þ Intensive Cultures, (2) Nest, (3) Intensive spatial scales, we first identified how the relative Cultures, (4) Nest þ Extensive Cultures and (5) cover of intensive and extensive cultures changed Extensive Cultures. The ‘‘Nest’’ term signifies over the different spatial scales measured that the model includes non-scalar variables that (Wheatley 2010) and then used correlations were descriptive of the nest (e.g., year, nest (Pearson and Spearman) to assess the relation- weight, etc.), the ‘‘Intensive Cultures’’ and ‘‘Exship between the relative cover of extensive and tensive Cultures’’ terms signify that the model intensive cultures across the 13 spatial scales includes variables that were descriptive of the within the 40 farms. landscape surrounding the nest (i.e., proportion We selected the number of Tree Swallow of intensive and extensive cultures in the fledglings, the abundance of Protocalliphora pu- landscape at different spatial scales, respectivepae and the level of parasitism by Nasonia per ly). More information on the composition and nest as our response variables. We chose to rationale of these groups of explanatory variables model the number of Tree Swallow fledglings are presented in Table 1 and Appendices A, B and not the number a nestling per nest as they and C. Competing models were then compared represent a better proxy of the resources available at predetermined functional spatial scales (see to the developing Protocalliphora larvae in that below) using the second-order Akaike informathey do not decrease through time due to tion criterion (AICc) and corresponding Akaike mortality and because they are strongly correlat- model weights (wi ) following Anderson (2008). ed with the number of nestlings per nest (r ¼ Model selection was performed using the aic0.92). Because Protocalliphora sialia Shannon & tab.mer function (package AICcmodavg version Dobroscky represented .95% of collected spec- 1.11) in R (version 2.1.11; R Development Core imens within our study system (Daoust et al. Team 2010). By contrasting complex models 2012), we chose to model P. sialia abundance containing both ‘‘Nest’’ and ‘‘Landscape’’ variwithin Tree Swallow nests, excluding other ables to ones that included only one of the latter, species from the analyses. Two species of Nasonia we were able to ascertain what single or were identified from parasitized Protocalliphora combination of explanatory parameters were pupae from the first sampling period, N. vitri- most parsimonious at describing the response pennis (Walker) and N. giraulti Darling. Nasonia variable. We were unable to include the two vitripennis represented .98% of reared speci- types of cultures (i.e., intensive and extensive) in mens in our study system (Daoust et al. 2012). the same model, with or without the interaction, That being said, in the majority of cases (2966 vs. as it led to multicollinearity among explanatory 635 and 954 vs. 321, in 2008 and 2009, respec- variables and prevented model convergence. tively) parasitism by Nasonia was inferred by the Hence, we built simpler models including only v www.esajournals.org

6


DAOUST ET AL. Table 1. Definition and justification of explanatory variables used to quantify the influence of landscape structure on number of Tree Swallow fledglings, the number of Protocalliphora sialia pupae per nest and the level of parasitism by Nasonia sp. on P. sialia pupae per nest. Justification Explanatory variable

Definition (units)

Year Hatching date

Julian date

Nest weight

grams (g)

No. fledglings

no. Tree Swallow fledglings/nest

No. P. sialia pupae

no. P. sialia pupae/nest

Intensive culture

relative amount of cover within a given extent around a nest box (%) relative amount of cover within a given extent around a nest box (%)

Extensive culture

No. Tree Swallow fledgings interannual variation clutch size decreases through the season (1)

nestling survival decreases with no. Protocalliphora (4) breeding success decreases with culture type (5) breeding success increases with culture type (5)

Protocalliphora sialia abundance

Level of parasitism by Nasonia sp.

idem Protocalliphora abundance increase through the season (2) no. Protocalliphora may increase with nest size (2) no. Protocalliphora may increase with no. fledglings (2)

idem higher levels of parasitism through the season (3)

no. Protocalliphora may decreases with with culture type (6) no. Protocalliphora may increase with culture type (6)

levels of parasitism may increase with nest size levels of parasitism may increase with no. fledglings levels of parasitism may increase with no. Protocalliphora pupae levels of parasitism may decrease with culture type (6) levels of parasitism may increase with culture type (6)

Notes: (1) Winkler and Allen 1996, (2) Bennett and Whitworth 1992, (3) Bennett and Whitworth 1991, (4) Streby et al. 2008, (5) Ghilain and Be´lisle 2008, (6) Tscharntke and Brandl 2004.

one type of culture. Explanatory variables included in a given model were never strongly correlated (0.01 , r , 0.46). To investigate how resource availability and landscape structure influenced our tri-trophic model, the spatial scales at which the organisms maximally respond to landscape composition had to be examined first. Functional spatial scales were determined for each response variable by fitting two general models (i.e., Nest þ Intensive Cultures and Nest þ Extensive Cultures) at each of the 13 spatial scales and contrasting them among spatial scales using Akaike model weights (wi ) based on DAICc. Spatial scales for which Akaike model weights showed clear peaks, thus corresponding to the scales at which organisms were most likely to respond to landscape composition, where identified as the functional spatial scales. The effect of each explanatory variable was accordingly estimated at each of those functional spatial scales. For model sets where no single candidate model stood out significantly above others (wi 0.95), effect sizes were computed by model averaging based on AICc using the modavg function (package AICcmodavg version 1.11), whereby parameter estimates for each v www.esajournals.org

predictor are averaged only over models including this predictor and standard errors take into account model selection uncertainty (Anderson 2008). To assess the fit of models, we provide a coefficient of determination (R2) computed as the Pearson’s r2 of the fitted values vs. observed values. In order to test our final prediction, that organisms at higher trophic levels are more negatively affected by agricultural intensification, we compared the fitted values obtained from the most parsimonious models, between the highest and lowest proportions of intensive cultures in the landscape.

RESULTS Landscape structure Two landscape metrics (i.e., relative cover of intensive and extensive cultures) were characterized at 13 spatial scales (Fig. 2a, b). The intensive zone of the study was characterized by high proportions of intensive culture, ranging from 60–90% (Fig. 2a). Similarly, the extensive zone of the study was characterized by high proportions of extensive cultures, ranging from 45–90% (Fig. 2b). The relative cover of intensive cultures was 7


DAOUST ET AL.

Tree Swallow fledging Within nest boxes where at least one egg hatched, the number of fledglings averaged 4.29 (SD ¼ 1.71) in 2008 and 4.04 (SD ¼ 1.88) in 2009. Contrasts between spatial scales revealed that the Nest þ Extensive Cultures model performed best at the scales of 0.3 and 5 km (wi ¼ 0.26 and wi ¼ 0.26, respectively; Fig. 3a) whereas the Nest þ Intensive Cultures model performed best a scale of 0.3 km (wi ¼ 0.35; Fig. 3a). The relative amount of intensive cultures within 0.3 km was strongly correlated with the relative amount of extensive cultures at 5 km (r ¼ 0.69; Appendix D: Table D1). Contrasting the five candidate models within the functional spatial scales revealed that both nest and landscape components influenced the number of Tree Swallow hatchlings that fledged from the nest at the 0.3 and 5-km functional scales (Appendix B: Table B1). At the scale of 0.3 km, model averaging indicated, however, that the number of Tree Swallow fledglings was only influenced by hatching date, whereby fewer birds fledged later in the season (Table 2). On the other hand, the number of Tree Swallow fledglings increased with the relative amount of extensive cultures within 5 km of the nest (Fig. 4) and decreased as the season progressed (Table 2). Fig. 2. Average values and associated variation (SD) for two landscape parameters: proportion of intensive (a) and extensive (b) culture in the landscape surrounding Tree Swallow nests across 13 spatial scales (while holding grain constant) from 40 farms along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009.

Protocalliphora nest abundance Within nest boxes where at least one swallow egg hatched, the number of P. sialia pupae averaged 24.20 (SD ¼ 23.38) in 2008 and 19.63 (SD ¼ 39.20) in 2009. Contrasts between spatial scales revealed that the Nest þ Extensive Cultures model clearly performed best at a spatial scale of 2 km (wi ¼ 1; Fig. 3b), whereas the Nest þ Intensive Cultures model performed best at a scale of 0.2 km (wi ¼ 1; Fig. 3b). The relative amount of intensive cultures within 0.2 km was only weakly correlated with the relative amount of extensive cultures within 2 km (r ¼ 0.35; Appendix A: Table A1). Contrasting the five candidate models within the functional spatial scales revealed strong evidence that the number of P. sialia pupae per nest was also influenced by both nest and landscape components at the 0.2 and 2-km scales (Appendix C: Table C1). The number of P. sialia pupae per nest increased with the number of fledglings per nest (Fig. 5a) and with hatching date at both scales (Fig. 5b, Table 3). However,

relatively stable across spatial scales until the 2km scale, where it rose from ;30% to 65% and stabilized from the 3-km scale onwards (Fig. 2a). On the other hand, the relative cover of extensive cultures gradually decreased across the spatial scales (Fig. 2b). Pearson and Spearman correlations were relatively similar, indicating that the relationship between the relative amount of extensive and intensive cultures was linear (Appendix D: Table D1). Furthermore, there were strong correlations (r . 0.46) between the relative amount of intensive and extensive cultures within landscapes among spatial scales, indicating that patterns at smaller scales were correlated with those at larger scales. v www.esajournals.org

8


DAOUST ET AL.

while the number of P. sialia pupae per nest decreased with the relative amount of intensive cultures within 0.2 km of the nest (Fig. 5c), it increased with the relative amount of extensive cultures within 2 km of the nest (Fig. 5d, Table 3).

Level of parasitism by Nasonia sp. Within nest boxes where at least one P. sialia pupa was found, the proportion of P. sialia pupae parasitized by Nasonia sp. was 53.68% in 2008 and 34.71% in 2009. Contrasts between spatial scales indicated that the Nest þ Extensive Cultures model performed best at spatial scales of 0.05 and 0.3 km (wi ¼ 0.26 and wi ¼ 0.61, respectively; Fig. 3c), whereas the Nest þ Intensive Cultures model performed best at 0.2 km (wi ¼ 0.91; Fig. 3c). The relative amount of intensive cultures within 0.2 km was strongly correlated with the relative amount of extensive cultures at 0.05 and 0.3 km (r ¼ 0.80 and r ¼ 0.84, respectively; Appendix A: Table A1). Contrasting the five candidate models for the number of P. sialia pupae parasitized by Nasonia sp. per nest within the functional spatial scales again provided evidence that both nest and landscape components have to be taken into account, and this, at the scales of 0.05, 0.2, and 0.3 km (Appendix D: Table D1). The number of P. sialia pupae parasitized by Nasonia sp. increased with the number of Tree Swallow fledglings per nest and the number of P. sialia pupae per nest at all three scales (Fig. 6b, Table 4). The number of P. sialia pupae parasitized by Nasonia sp. also increased with nestling hatching date at the 0.2 and 0.3-km scales (Fig. 6a, Table 4). Regarding the effect of landscape composition, the number of P. sialia pupae parasitized by Nasonia sp. decreased with the relative amount of intensive cultures within 0.2 and 0.3 km of the nest and increased with the relative amount of extensive cultures within 0.3 km of the nest (Fig. 6c, Table 4).

Fig. 3. Akaike weights of the models explaining (a) the number of Tree Swallow hatchlings that fledged (n ¼ 387; nests were included in the analysis when 1 nestling developed within it), (b) abundance of Protocalliphora sialia pupae (n ¼ 387; nests were included in the analysis when 1 nestling developed within it) and (c) level of parasitism by Nasonia sp. on Protocalliphora sialia (n ¼ 207; nests were included in the analysis when 1 P. sialia developed within it) from nests within the experimental nest boxes at 13 scales scales along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. Models are generalized mixed models with a log link for poisson error distribution with farm ID as random factor. See Table 1 for variable definition and justification.


Trophic rank and susceptibility to agricultural intensification We obtained the fitted values from the Nest þ Intensive Cultures models at the spatial scales of 0.3, 0.2 and 0.2 for the number of Tree Swallow fledglings, the abundance of P. sialia and the level of parasitism by Nasonia sp. per nest, respectively. The number of Tree Swallow fledglings per 9


DAOUST ET AL. Table 2. Model averaged parameters explaining the number of Tree Swallow hatchlings that fledged from the nest within nest boxes along a gradient of agricultural intensification at pre-identified functional spatial scales in southern Que´bec, Canada (n ¼ 387; nests were included in the analysis when 1 nestling developed within it). Parameter

Estimate

Unconditional SE

Lower CI

Upper CI

0.002 0.159 0.021 0.001 0.095 0.015

0.002 0.682 0.056 0.001 0.176 0.066

0.001 1.496 0.032 0.001 0.249 0.115

0.005 1.178 0.098 0.004 0.439 0.145

0.002 0.159 0.021 0.001 0.209 0.180

0.002 0.682 0.056 0.001 0.258 0.082

0.001 1.496 0.032 0.001 0.296 0.020

0.005 1.178 0.098 0.004 0.715 0.341

0.3 km No. Protocalliphora sialia pupae Year Hatching date Nest weight Intensive culture Extensive culture 5 km No. Protocalliphora sialia pupae Year Hatching date Nest weight Intensive culture Extensive culture

nest decreased by 20.4% (from 4.9 to 3.9) between the least and most intensive farmlands. A reduction of 80% (from 20 to 4) in the abundance of P. sialia pupae per nest was also observed between the least and most intensive farmlands. Lastly, there was a reduction of 35.4% (from 20 to 13) in the level of parasitism by Nasonia sp. per nest between the least and most intensive farmlands.

that not only do interacting organisms at different trophic levels respond to landscape structure at distinctive spatial scales, but this response can also differ based on which landscape parameter is considered. Furthermore, the number of Tree Swallow fledglings, the abundance of P. sialia and the level of parasitism by Nasonia sp. were negatively affected by agricultural intensification. Major results are summarized in Table 5. We reported that the functional spatial scales to which the studied organisms responded to landscape structure differed according to the landscape parameters modeled. For instance, Tree Swallows responded to intensive cultures

DISCUSSION To our knowledge, this is the first study to investigate the effects of agricultural intensification using a tri-trophic animal system. We show

Fig. 4. Influence of landscape composition (proportion of extensive culture within a 5 km radius of the nest) on number of Tree Swallow fledglings per nest along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009 (n ¼ 387; nests were included in the analysis when 1 nestling developed within it).


10


DAOUST ET AL.

Fig. 5. Influence of landscape composition on number of Protocalliphora sialia pupae per Tree Swallow nest along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. (a) Effect of the number of Tree Swallow nestlings per nest on the number of P. sialia per nest based on the Nest þ Intensive model in Appendix C. (b) Effect of the Tree Swallow nestling hatching date on the number of P. sialia per nest based on the Nest þ Intensive model in Appendix C. (c) Effect of the proportion of intensive culture within a 0.2 km radius of the nest on the number of P. sialia per nest based on the Nest þ Intensive model in Appendix C. (d) Effect of the proportion of extensive culture within a 2 km radius of the nest on the number of P. sialia per nest based on the Nest þ Extensive model in Appendix C (n ¼ 387; nests were included in the analysis when 1 nestling developed within it).

Table 3. Parameters of the most parsimonious models explaining Protocalliphora sialia abundance within the nest boxes along a gradient of agricultural intensification at pre-identified functional spatial scales in southern Que´bec, Canada. (n ¼ 387; nests were included in the analysis when 1 nestling developed within it). Parameter Model 1: Nest þ Intensive ¼ 0.2 km Year No. fledglings Hatching date Nest weight Intensive culture Model 4: Nest þ Extensive ¼ 2 km Year No. fledglings Hatching date Nest weight Extensive culture


Estimate

Unconditional SE

Lower CI

Upper CI

1.162 0.254 0.052 0.006 2.688

0.045 0.017 0.004 0.007 0.191

1.250 0.221 0.044 0.007 3.062

1.074 0.287 0.06 0.0018 2.314

1.147 0.277 0.052 0.005 1.496

0.045 0.017 0.004 0.001 0.180

1.235 0.244 0.044 0.004 1.144

1.059 0.31 0.06 0.006 1.848

11


DAOUST ET AL.

at a relatively small scale (0.3 km) while they responded to extensive cultures at a much larger scale (5 km). Similarly, P. sialia responded to intensive cultures at a smaller scale (0.2 km) than to extensive cultures (2 km). In contrast, the functional spatial scales experienced by Nasonia sp. were relatively similar between landscape parameters: 0.2 and 0.3 km for intensive and extensive cultures, respectively. We do not have a comprehensive explanation for these patterns. It might be that this differential response to the two landscape parameters is in fact a correlation artefact. Analyses revealed substantial correlations between relative amounts of intensive cultures at smaller scales and those of extensive cultures at larger scales (Appendix D). For example, we present strong correlations between the small spatial scales perceived by both Tree Swallows and Nasonia sp. in response to intensive cultures and the larger scales perceived in response to extensive cultures. However this is not true for P. sialia, as there is little correlation between the spatial scales it responded to. In this case, we attribute this differential perception to the underlying variations in the landscape structure between the intensive and extensive portions of our study system. As previously stated, the landscape matrix surrounding a patch significantly influences the structural connectivity of the habitat (Steffan-Dewenter 2003). For instance, the spatial distribution of crop and noncrop habitats in the surrounding landscape may affect extinction and recolonization patterns within patches via dispersal (Jonsen and Fahrig 1997). The intensive portion of our study system is characterized by high percentage of intensive cultures (70–90%) within the first 2 km surrounding bird nests. Such high levels of intensive cultures could then be construed as a hostile environment (as per Bianchi et al. 2006) that may act as a barrier to P. sialia dispersal. We postulate that the functional spatial scale of 0.2 km identified when modelling P. sialia abundance as a function of intensive cultures represents the maximal level for the negative effects associated with intensive cultures. Conversely, the proportion of extensive cultures within the extensive zone of the study decreased gradually across spatial scales, reaching its lowest level (;40%) between radii of 2–20 km. Non-crop habitats such as fallows, pastures, hedgerows and wood-

Fig. 6. Influence of landscape composition within 0.3 km of Tree Swallow nests on number of Protocalliphora sialia pupae parasitized by Nasonia sp. per Tree Swallow nest along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. (a) Effect of the Tree Swallow nestling hatching date on the number of P. sialia pupae parasitized by Nasonia sp. per nest based on model averaged estimates. (b) Effect of the number of number of P. sialia pupae per nest on the number of P. sialia pupae parasitized by Nasonia sp. per nest based on model averaged estimates. (c) Effect of the proportion of extensive culture on the number of P. sialia pupae parasitized by Nasonia sp. per Tree Swallow nest based on the based on model averaged estimates (n ¼ 207; nests were included in the analysis when 1 nestling developed within it).


12


DAOUST ET AL. Table 4. Parameters of the most parsimonious and averaged models explaining the level of parasitism by Nasonia sp. on Protocalliphora sialia within nest boxes along a gradient of agricultural intensification at pre-identified functional spatial scales in southern Que´bec, Canada (n ¼ 207; nests were included in the analysis when 1 P. sialia developed within it). Parameter

Estimate

Unconditional SE

Lower CI

Upper CI

0.128 0.036 0.019 0.027 0.002 0.063 0.073

0.027 0.001 0.077 0.059 0.001 0.127 0.117

0.076 0.034 0.170 0.089 0.004 0.311 0.157

0.180 0.038 0.132 0.142 0.000 0.186 0.303

0.135 0.036 0.010 0.028 0.002 0.589 0.278

0.027 0.001 0.077 0.006 0.001 0.217 0.178

0.081 0.033 0.161 0.016 0.004 1.013 0.071

0.188 0.038 0.014 0.040 0.000 0.164 0.628

0.138 0.035 0.011 0.028 0.002 0.802 0.367

0.027 0.001 0.077 0.006 0.001 0.252 0.128

0.084 0.033 0.162 0.016 0.004 1.295 0.116

0.192 0.038 0.140 0.041 0.000 0.308 0.618

Nest þ Extensive ¼ 0.05 km No. fledglings No. Protocalliphora sialia pupae Year Hatching date Nest weight Intensive culture Extensive culture 0.2 km No. fledglings No. Protocalliphora sialia pupae Year Hatching date Nest weight Intensive culture Extensive culture 0.3 km No. fledglings No. Protocalliphora sialia pupae Year Hatching date Nest weight Intensive culture Extensive culture

lots, which are characteristic of extensive farming, act as biodiversity reservoirs providing alternate food sources, shelters, overwintering sites to many species and facilitate dispersal (Matson et al. 1997). The 2-km scale identified when modelling P. sialia abundance, as a function of extensive cultures in the landscape, is likely to represent the true functional spatial scale at which P. sialia perceives and utilizes habitat. The functional spatial scales identified for Tree Swallows, P. sialia and Nasonia sp. also appear to

be determined, at least partly, by their trophic rank; the higher up the trophic ladder, the smaller the spatial scales at which the organisms responded to the proportion of extensive cultures (Table 5). These findings are consistent with studies on other host-parasitoid associations (Thies et al. 2003, Holzschuh et al. 2010). However, this pattern does not concur with the trophic rank/species-area relationships theory proposed by Holt (1996) and Holt et al. (1999), that organisms at higher trophic levels should

Table 5. Summary of the effect of landscape structure on the functional spatial scales and trophic interactions of the Tree Swallows, Protocalliphora sialia and Nasonia sp. along a gradient of agricultural intensification covering 10,200 km2 in southern Que´bec, Canada. The impact of agricultural intensification is a combination of direct (loss of nesting sites, overwintering sites, shelters and sources of food) and indirect effects (host availability and level of parasitism by Nasonia sp. on P. sialia) of agricultural intensification. Parameters Functional spatial scale (km) Extensive cultures Intensive cultures Influence of culture type Extensive cultures Intensive cultures Impact of agricultural intensification

Tree Swallow

P. sialia

Nasonia sp.

5.0 0.3

2.0 0.2

0.3 0.2

" no. fledglings/nest ... 20% # fledglings/nest

" no. pupae/nest # no. pupae/nest 80% # no. pupae/ nest

" level of parasitism # level of parasitism 35% # level of parasitism

Direct and indirect effects combined.


13


DAOUST ET AL.

experience the landscape at larger spatial scales. This theory assumes that communities are closed to immigration (isolated habitats), composed of ‘‘stacked specialist’’ food chains (where each plant species supports a specialist herbivore, which in turn sustains a specialist carnivore, and so forth), and although not directly stated, it is implied that the size of organisms increased with trophic rank. Because host-parasite-parasitoid systems typically violate at least one of these assumptions (i.e., increasing size with trophic rank), there is a need to develop alternative conceptions for interactions along chains of hosts and parasites. The body size of the organisms within host-parasitoid associations usually does not increase with trophic level as is often the case in vertebrate predator-prey associations. The ability of an organism to disperse determines its functional spatial scale and, in turn, the extent to which landscape elements contribute to its population dynamics and trophic interactions. In many instances, functional spatial scales are strongly related to body size (Roland and Taylor 1997, Holland et al. 2005). Indeed, in both tachinid parasitoids (Roland and Taylor 1997) and longhorn beetles (Holland et al. 2005), the functional spatial scales increase with average species size. Thus, in host-parasitoid systems, body size rather than trophic rank is likely to act as a key determinant of functional spatial scales. In agreement with our first prediction, we showed that the number of Tree Swallow fledglings was positively associated with the proportion of extensive cultures in the landscape and decreased as the breeding season progressed. These results parallel those of Ghilain and Be´lisle (2008) in which the higher number of fledglings in more extensive farmlands was attributed to an increase availability of insect prey, especially when provisioning nestlings as food availability is a major determinant of Tree Swallow nestling growth and survival (McCarty and Winkler 1999, Ghilain and Be´lisle 2008). Moreover, clutch size did not diminish sufficiently (0.8 egg) along the agricultural intensification gradient to account for the observed difference in fledgling number (2.6 fledglings). Protocalliphora sialia abundance was shown to decrease as the proportion of intensive cultures in the landscape increased. The current literature is replete with examples reporting lower popuv www.esajournals.org

lation abundances of various taxa within intensively cultured farmlands (Burel et al. 1998, Benton et al. 2003, Burel et al. 2004, Tscharntke et al. 2005, Donald et al. 2006). Due to its life history, P. sialia is potentially more sensitive to agricultural intensification than other animals at lower trophic levels as it is both directly and indirectly affected by agricultural intensification. First, agricultural intensification can act directly on P. sialia populations by reducing sources of nectar as well as the number of shelter and overwintering sites necessary for adult survival (Bennett and Whitworth 1991). Second, it can act indirectly by diminishing bird host populations (Sabrosky et al. 1989, Donald et al. 2001, Murphy 2003, Bianchi et al. 2006). Interestingly, the number of P. sialia pupae per nest increased as the bird-breeding season progressed, as reported in several other studies (Bortolotti 1984, Roby et al. 1992); indicating that birds nesting earlier in the season benefit from lower levels of ectoparasitism. Protocalliphora sialia also benefitted from increased bird host availability within nests, thus supporting our second prediction. Density-dependent processes are indeed likely to govern the abundance of P. sialia per nest. As in many other parasitic insect models (Hassel 1982, Walde and Murdoch 1988), host patch density could influence the ovipositing behavior of P. sialia, with individual females investing less time and laying fewer eggs in lower quality patches. In addition, as P. sialia females cue primarily on host odours (Bennett and Whitworth 1991), high host densities emitting stronger olfactory cues should attract more P. sialia females, a behaviour which has also been reported in parasitoids (Turlings et al. 1993, Mackauer and Michaud 1996, Sullivan et al. 2000). These hypotheses are not mutually exclusive and additional studies will be required to elucidate the ovipositing behavior of Protocalliphora and the influence of host density and habitat structure on it. In agreement with our first two predictions, parasitism by Nasonia sp. on P. sialia decreased with the amount of intensive culture in the landscape and increased with the number of P. sialia pupae per nest. Like P. sialia, Nasonia sp. were directly affected by agricultural intensification as adults depend on sources of nectar and shelter sites, which are less common in inten14


DAOUST ET AL.

sively cultured lands (Steffan-Dewenter 2003, Tscharntke et al. 2005). As with P. sialia, levels of Nasonia sp. parasitism are likely to be governed by density-dependent processes. Lastly, our data do not support the prediction that organisms at higher trophic levels should be more susceptible to habitat loss, fragmentation and degradation than those at lower trophic levels. We showed that blowfly ectoparasites (second trophic level) are disproportionally affected by agricultural intensification; the effect of the relative cover of intensive cultures in the landscape on the abundance of P. sialia pupae per nest decreased by 80% from low to high intensive farms as compared to a 20% and 35% reduction in the number of Tree Swallow fledglings and in the level of parasitism by Nasonia sp., respectively. We propose two, non-mutually exclusive hypotheses to explain these observations. First, due to their unique life history, P. sialia are highly dependent on a handful of altricial farmland bird species whose populations are greatly reduced by agricultural intensification (Jobin et al. 1996, Murphy 2003, Ghilain and Be´lisle 2008). This significant decrease in host availability within intensive farmlands leads to a concomitant decrease in P. sialia. Second, because parasitoids are relatively free of natural enemies within our system (having found no evidence of hyperparasitism in our samples) and because Nasonia sp. (the most abundant species collected; Daoust et al. 2012) is a generalist parasitoid of calyptrate flies, whose abundances only significantly differ between intensive and extensively managed habitats later in the season (Rioux-Paquette et al. in press), Nasonia sp. parasitoids may also contribute an important ‘‘top-down’’ effect on P. sialia populations. Therefore, it is the combined effects of both ‘‘bottom-up’’ and ‘‘top-down’’ forces that could potentially lead to the disproportionate drop in P. sialia abundance across the gradient of agricultural intensification. To conclude, habitat loss, fragmentation and degradation associated with the intensification of agricultural practices impact trophic processes by contributing to the reduction of Tree Swallow fledglings per nest, ectoparasitism by P. sialia and parasitism by Nasonia sp.. The spatial scales at which the organisms respond increase with their size and change depending on habitat type. With croplands and pastures occupying .40% of the v www.esajournals.org

total terrestrial surface (Foley et al. 2005), it is critical for ecologists to design research protocols that include spatial aspects of trophic interactions as they dictate local biodiversity and community functioning. Furthermore, our results highlight the importance of considering multiple landscape parameters when quantifying the functional spatial scales of an organism. This is especially important for conservation biologists who use functional spatial scales as a method of delineating conservation areas. Within our system, using only one landscape parameter would lead to a false estimation of the area used by an animal.

ACKNOWLEDGMENTS We thank Arnaud Ghilain and all members the TRSW group at Universite´ de Sherbrooke for their support and hard work. This work would not have been possible without the help of Joseé Doyon, Brian Mader, Krista Gigue`re, Simon Legault, Mathieu Be´langer-Morin, Fanny Maure, Simon Laurin-Lemay, Louis Laplante, Jeremy Daoust, Christian Daoust, ¸ Michelle Payette Daoust, Sylvain Daoust, Jean-Francois Limoges and Erfan Vafaie in the field. We also thank Lenore Fahrig and Patrick James for their useful comments. Financial support was provided to J.S., M.B. and J.B. from the National Sciences and Engineering Research Council of Canada (NSERC). SPD was supported by scholarships from the Fonds que´bećois de la recherche sur la nature et les technologies (FQRNT) and the Que´bec center for biodiversity studies (QCBS).

LITERATURE CITED Anderson, D. R. 2008. Model based inference in the life sciences: a primer on evidence. Springer, New York, New York, USA. Be´langer, L., and M. Grenier. 2002. Agriculture intensification and forest fragmentation in the St. Lawrence valley, Quebec, Canada. Landscape Ecology 17:94–112. Bennett, G. F., and T. L. Whitworth. 1991. Studies on the life history of some species of Protocalliphora (Diptera: Calliphoridae). Canadian Journal of Zoology 69:2048–2058. Bennett, G. F., and T. L. Whitworth. 1992. Host, nest, and ecological relationships of species of Protocalliphora (Diptera: Calliphoridae). Canadian Journal of Zoology 70:51–61. Benton, T. G., J. A. Vickery, and J. D. Wilson. 2003. Farmland biodiversity: is habitat heterogeneity the key? Trends in Ecology and Evolution 18:182–188. Bianchi, F. J. J. A., C. J. H. Booij, and T. Tscharntke.

15


DAOUST ET AL. 2006. Sustainable pest regulation in agricultural landscapes: a review on landscape composition, biodiversity and natural control. Proceedings of the Royal Society of London B 273:1715–1727. Bortolotti, G. 1984. Frequency of Protocalliphora avium (Diptera: Calliphoridae) infestations on bald eagles (Haliaeetus leucocephalus). Canadian Journal of Zoology 63:165–168. Burel, F., J. Baudry, A. Butet, P. Clergeau, Y. Delette, D. Le Coeur, F. Dubs, N. Morvan, G. Paillat, S. Petit, C. Thenail, E. Brunel, and J.-C. Lefeuvre. 1998. Comparative biodiversity along a gradient of agricultural landscape. Acta Oecologica 19:47–60. Burel, F., A. Butet, Y. R. Delette, and N. Milla`n de la Pena. 2004. Differential responses of selected taxa to landscape context and agricultural intensification. Landscape and Urban Planning 67:195–204. Cagnolo, L., G. Valladares, A. Salvo, M. Cabido, and M. Zak. 2009. Habitat fragmentation and species loss across three interacting trophic levels: effects of life-history and food-web traits. Conservation Biology 23:1167–1175. Canadian Wildlife Service. 2004. Occupation du sol a partir des images classifieés Landsat-7, Sud du Que´bec, 1999–2003. Environnement Canada, region du Que´bec, Que´bec, Canada. Daoust, S. P., J. Savage, T. L. Whitworth, M. Be´lisle, and J. Brodeur. 2012. Diversity and abundance of ectoparasitic blowflies Protocalliphora (Diptera: Calliphoridae) and their Nasonia parasitoids (Hymenoptera: Pteromalidae) in Tree Swallow nests within agricultural lands of southern Que´bec, Canada. Annals of the Entomological Society of America 105:471–478. Darling, C. D., and J. H. Werren. 1990. Biosystematics of Nasonia (Hymenoptera: Pteromalidae): two new species reared from bird’s nests in North America. Annals of the Entomological Society of America 83:352–370. de Knegt, H. J., F. van Langevelde, M. B. Coughenour, A. K. Skidmore, W. F. de Boer, I. M. A. Heitkonig, et al. 2010. Spatial autocorrelation and the scaling of species-environment relationships. Ecology 91:2445–2465. Doliguez, B., T. Paert, E. Danchin, J. Clobert, and L. Gustafsson. 2004. Availability and use of public information and conspecific density for settlement decisions in the collared flycatcher. Journal of Animal Ecology 73:75–87. Donald, P. F., R. E. Green, and M. F. Heath. 2001. Agricultural intensification and the collapse of europe’s farmland bird populations. Proceedings of the Royal Society London B 268:25–29. Donald, P. F., F. J. Sanderson, I. J. Burfield, and P. J. van Bommel. 2006. Further evidence of continent-wide impacts of agricultural intensification on European farmland birds, 1990-2000. Agriculture, Ecosys-


tems & Environment 116:189–196. Eeva, T., E. Lehikoinen, and J. Nurmi. 1994. Effects of ectoparasites on breeding success of great tits (Parus major) and pied flycatchers (Ficedula hypoleuca) in an air pollution gradient. Canadian Journal of Zoology 72:624–635. ESRI. 2005. Patch analyst 2.0a patch for French language. ESRI, Redlands, California, USA. Fahrig, L. 2007. Non-optimal animal movement in human-altered landscapes. Functional Ecology 21:1003–1015. Fahrig, L., J. Baudry, L. Brotons, F. Burel, T. O. Crist, R. J. Fuller, C. Sirami, G. M. Siriwardena, and J.-L. Martin. 2011. Functional landscape heterogeneity and animal biodiversity in agricultural landscapes. Ecology Letters 14:101–112. Financie`re agricole du Que´bec. 2011. Base de donneés des cultures assureés. Guide technique. La Financie`re agricole du Que´bec, St-Romuald, Que´bec, Canada. http://www.fadq.qc.ca/fileadmin/fr/geom/ pour_prof_geom/guide_tech_bdca.pdf Fisher, J. T., S. Boutin, and S. J. Hannon. 2005. The protean relationship between boreal forest landscape structure and red squirrel distribution at multiple spatial scales. Landscape Ecology 20:73– 82. Foley, J. A., R. DeFries, G. P. Asner, C. Barford, G. Bonan, S. R. Carpenter, et al. 2005. Global consequences of land use. Science 309:570–574. Ghilain, A., and M. Be´lisle. 2008. Breeding success of the Tree Swallow along a gradient of agricultural intensification. Ecological Applications 18:1140– 1154. Hassel, M. P. 1982. Patterns of parasitism by insect parasitoids in patchy environments. Ecological Entomology 7:365–377. Holland, J. D., L. Fahrig, and N. Cappuccino. 2005. Body size affects the spatial scale of habitat–beetle interactions. Oikos 110:101–108. Holt, R. D. 1996. Food webs in space: an island biogeographic perspective. Pages 313–323 in G. A. Polis, and K. O. Winemiller, editors. Food webs: Integration of patterns and dynamics. Chapman & Hall, New York, New York, USA. Holt, R. D., J. H. Lawton, G. A. Polis, and N. D. Martinez. 1999. Trophic rank and the species-area relationship. Ecology 80:1495–1505. Holzschuh, A., I. Stefan-Dewenter, and T. Tscharntke. 2010. How do landscape composition and configuration, organic farming and fallow strips affect the diversity of bees, wasps and their parasitoids? Journal of Animal Ecology 79:491–500. Jobin, B., J.-L. DesGranges, and C. Boutin. 1996. Population trends in selected species of farmland birds in relation to recent developments in agriculture in the St. Lawrence Valley. Agriculture, Ecosystems and Environments 57:103–116.

16


DAOUST ET AL. Jobin, B., M. Grenier, and P. Laporte. 2005. Using satellite imagery to assess breeding habitat availability of the endangered loggerhead shrike in Quebec. Biodiversity and Conservation 14:81–95. Jonsen, I. D., and L. Fahrig. 1997. Response of generalist and specialist insect herbivores to landscape spatial structure. Landscape Ecology 12:185– 197. Mackauer, M., and J. P. Michaud. 1996. Host choice by aphid parasitoids (Hymenoptera: Aphididae): host recognition, host quality, and host value. Canadian Entomologist 128:959–980. Marshall, E. J. P., and A. C. Moonen. 2002. Field margins in northern Europe: their functions and interactions with agriculture. Agriculture, Ecosystems and Environment 89:5–21. Matson, P. A., W. J. Parton, A. G. Power, and M. J. Swift. 1997. Agricultural intensification and ecosystem properties. Science 277:504–509. McCarty, J. P., and D. W. Winkler. 1999. Foraging ecology and diet selectivity of Tree Swallows feeding nestlings. Condor 101:246–254. Murphy, M. T. 2003. Avian population trends within the evolving agricultural landscape of eastern and central United States. Auk 120:20–34. R Development Core Team. 2010. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Raychoudhury, R., C. A. Desjardins, J. Buellesbach, D. W. Loehlin, K. Grillenberger, L. Beukeboom, T. Schmitt, and J. H. Werren. 2010. Behavioral and genetic characteristics of a new species of Nasonia. Heredity 104:278–288. Rioux Paquette, S., D. Garant, F. Pelletier, and M. Be´lisle. In press. Seasonal patterns in tree swallow prey (Diptera) abundance are affected by agricultural intensification. Ecological Applications. http:// dx.doi.org/10.1890/12-0068.1 Robinson, R. A., and W. J. Sutherland. 2002. Post-war changes in arable farming and biodiversity in Great Britain. Journal of Applied Ecology 39:157–179. Roby, D. D., K. L. Brink, and K. Wittmann. 1992. Effects of bird blowfly parasitism on Eastern Bluebird and Tree Swallow Nestlings. Wilson Bulletin 104:630–643. Roland, J., and P. D. Taylor. 1997. Insect parasitoid species respond to forest structure at different spatial scales. Nature 386:710–713. Sabrosky, C. W., G. F. Bennett, and T. L. Whitworth. 1989. Bird Blowflies (Protocalliphora) in North America (Diptera: Calliphoridae) with notes on the Palearctic Species. Smithsonian Institute Press, Washington, D.C., USA. Schweiger, O. M., J. P. Maelfait, W. Van Wingerden, F. Hendrickx, R. Billeter, M. Speelmans, et al. 2005. Quantifying the impact of environmental factors on arthropod communities in agricultural landscapes


across organizational levels and spatial scales. Journal of Applied Ecology 42:119–1139. Simon, A., D. W. Thomas, J. R. Speakman, J. Blondel, P. Perret, and M. M. Lambrechts. 2005. Impact of ectoparasitic blowfly larvae (Protocalliphora spp.) on the behavior and energetics of nestling Blue Tits. Journal of Field Ornithology 76:402–410. Steffan-Dewenter, I. 2003. Importance of habitat area and landscape context for species richness of bees and wasps in fragmented orchard meadows. Conservation Biology 17:1036–1044. Streby, H. M., S. M. Peterson, and P. M. Kapfer. 2009. Fledging success is a poor indicator of the effects of bird blowflies on oven bird survival. Condor 111:193–197. Sullivan, B. T., E. M. Petterson, K. C. Seltmann, and C. W. Berisford. 2000. Attraction of the bark beetle parasitoid Roptrocerus xylophagorum (Hymenoptera: Pteromalidae) to host-associated olfactory cues. Environmental Entomology 29:1138–1151. Thies, C., I. Steffan-Dewenter, and T. Tscharntke. 2003. Effets of landscape context on herbivory and parasitism at different spatial scales. Oikos 101:18–25. Thies, C., I. Roschewitz, and T. Tscharntke. 2005. The landscape context of cereal aphid-parasitoid interactions. Proceedings of the Royal Society of London B 272:203–210. Tscharntke, T., and R. Brandl. 2004. Plant-insect interactions in a fragmented landscape. Annual Review of Entomology 49:405–430. Tscharntke, T., A. M. Klein, A. Kruess, I. SteffanDewenter, and C. Thies. 2005. Landscape perspectives on agricultural intensification and biodiversity: ecosystems service management. Ecology Letters 8:857–874. Turlings, T. C. J., F. L. Wa¨ckers, L. E. M. Vet, W. J. Lewis, and J. H. Tumlison. 1993. Learning of hostfinding cues by hymenopterous parasitoids. Pages 51–78 in D. R. Papaj, editor. Insect learning: ecological and evolutionary perspectives. Chapman & Hall, New York, New York, USA. Walde, S. J., and W. W. Murdoch. 1988. Spatial density dependence in parasitoids. Annual Review of Entomology 33:441–466. Werren, J. H., and D. W. Loehlin. 2009. The parasitoid wasp Nasonia: an emerging model system with haploid male genetics. Cold Spring Harbor Protocols 10:1–31. Wheatley, M. 2010. Domains of scale in forestlandscape metrics: implications for species-habitat modeling. Acta Oecologica 36:259–267. Wheatley, M., and C. Johnson. 2009. Factors limiting our understanding of ecological scale. Ecological Complexity 6:150–159. Whitworth, T. L. 2003. A key to the puparia of 27 species of North American Protocalliphora Hough

17


DAOUST ET AL. (Diptera: Calliphoridae) from bird nests and two new puparial descriptions. Proceedings of the Entomological Society of Washington 105:995– 1033. Whitworth, T., and G. F. Bennett. 1992. Pathogenicity of larval Protocalliphora (Diptera: Calliphoridae) parasitizing nestling birds. Canadian Journal of Zoology 70:2184–2191. Wickramasinghe, L. P., S. Harris, G. Jones, and N. V. Jennings. 2004. Abundance and species richness of nocturnal insects on organic and conventional farms: effects of agricultural intensification on bar foraging. Conservation Biology 18:1283–1292.

Wiens, J. A. 1989. Spatial scaling in ecology. Functional Ecology 3:385–397. Winkler, D. W., and P. E. Allen. 1996. The seasonal decline in Tree Swallow clutch size: physiological constraints or strategic adjustment? Ecology 77:922–932. Winkler, D. W., K. K. Hallinger, D. R. Ardia, R. J. Robertson, B. J. Stutchbury, and R. R. Cohen. 2011. Tree Swallow (Tachycineta bicolor). In A. Poole, editor. The birds of North America. Cornell Lab of Ornithology, Ithaca, New York, USA. http://bna. birds.cornell.edu/bna/

SUPPLEMENTAL MATERIAL APPENDIX A Table A1. Akaike weights of the models explaining number of Tree Swallow hatchlings that fledged from the nest within the experimental nest boxes at the spatial scales of 0.3 km and 5 km along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. AICc weights depict the relative strength of each model for a given spatial scale. Models are generalized mixed models with a log link for Poisson error distribution with farm ID (numerical label given to each of the 40 farms used in the study) as random factor (n ¼ 387; nests were included in the analysis when 1 nestling developed within it). See Table 1 for variable definition and justification. Nest þ Intensive Spatial extent (km) 0.3 5

2

AICc wi

R

0.31 0.12

0.26 0.26

Nest

Nest þ Extensive

Intensive 2

2

2

AICc wi

R

AICc wi

R

AICc wi

R

0.49 0.60

0.22 0.22

0 0

0.16 0.16

0.20 0.28

0.24 0.22

Extensive AICc wi

R2

0 0

0.16 0.16

Notes: List of models and their variables: Nest þ Intensive ¼ number of Protocalliphora sialia pupae þ year þ hatching date þ nest weight þ intensive culture; Nest ¼ number of Protocalliphora sialia pupae þ year þ hatching date þ nest weight; Intensive ¼ intensive culture; Nest þ Extensive ¼ number of Protocalliphora sialia pupae þ year þ hatching date þ nest weight þ extensive culture; Extensive ¼ extensive culture.

APPENDIX B Table B1. Akaike weights of the models explaining the abundance of Protocalliphora sialia pupae within the experimental nest boxes at the spatial scales of 0.2 km and 2 km along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. AICc weights depict the relative strength of each model for a given spatial scale. Models are generalized mixed models with a log link for Poisson error distribution with farm ID (numerical label given to each of the 40 farms used in the study) as random factor (n ¼ 387; nests were included in the analysis when 1 nestling developed within it). See Table 1 for variable definition and justification. Nest þ Intensive Spatial extent (km) 0.2 2

2

AICc wi

R

1 0

0.44 0.40

Nest

Nest þ Extensive

Intensive 2

2

2

AICc wi

R

AICc wi

R

AICc wi

R

0 0

0.39 0.39

0 0

0.16 0.15

0 1

0.41 0.43

Extensive AICc wi

R2

0 0

0.15 0.16

Notes: List of models and their variables: Nest þ Intensive ¼ number of fledglings þ year þ hatching date þ nest weight þ intensive culture; Nest ¼ number of fledglings þ year þ hatching date þ nest weight; Intensive ¼ intensive culture; Nest þ Extensive ¼ number of fledglings þ year þ hatching date þ nest weight þ extensive culture; Extensive ¼ extensive culture.


18


DAOUST ET AL.

APPENDIX C Table C1. Akaike weights of the models explaining the level of parasitism by Nasonia sp. on Protocalliphora sialia within the experimental nest boxes at the spatial scales of 0.05 km, 0.2 km and 0.3 km along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. AICc weights depict the relative strength of each model for a given spatial scale. Models are generalized mixed models with a log link for Poisson error distribution with farm ID (numerical label given to each of the 40 farms used in the study) as random factor (n ¼ 207; nests where 1 P. sialia developed within it). See Table 1 for variable definition and justification. Nest þ Intensive Spatial extent (km) 0.05 0.20 0.30

2

Nest

Nest þ Extensive

Intensive 2

AICc wi

R

AICc wi

R

0.01 0.34 0.60

0.81 0.84 0.84

0 0 0

0.78 0.78 0.78

2

Extensive

2

AICc wi

R

AICc wi

R

0 0 0

0.20 0.19 0.19

0.99 0.66 0.40

0.81 0.81 0.81

AICc wi

R2

0 0 0

0.20 0.20 0.45

Notes: List of models and their variables: Nest þ Intensive ¼ number of fledglings þ number of P. sialia pupae þ year þ hatching date þ nest weight þ intensive culture; Nest ¼ number of fledglings þ number of P. sialia pupae þ year þ hatching date þ nest weight; Intensive ¼ intensive culture; Nest þ Extensive ¼ number of fledglings þ number of P. sialia pupae þ year þ hatching date þ nest weight; Extensive ¼ extensive culture.

APPENDIX D Table D1. Pearson (top) and Spearman (bottom) correlations of the percentages of extensive and intensive agriculture between spatial scales within 40 farms along a gradient of agricultural intensification in southern Que´bec, Canada, 2008–2009. Spatial scale 0.05 0.1 0.2 0.3 0.4 0.5 1 2 3 4 5 10 20

0.05 0.913 0.880 0.794 0.689 0.613 0.573 0.400 0.396 0.300 0.285 0.256 0.320 0.320

0.05

0.1

0.2

0.3

0.4

0.5

1

2

3

4

5

10

20

0.925 0.1 0.911 0.863 0.767 0.701 0.675 0.494 0.496 0.415 0.407 0.385 0.434 0.434

0.904 0.919 0.2 0.883 0.817 0.759 0.737 0.546 0.363 0.498 0.495 0.485 0.526 0.526

0.804 0.860 0.878 0.3 0.838 0.794 0.773 0.572 0.589 0.531 0.529 0.522 0.564 0.564

0.698 0.764 0.818 0.829 0.4 0.814 0.797 0.605 0.612 0.551 0.552 0.550 0.600 0.600

0.620 0.695 0.759 0.789 0.791 0.5 0.777 0.595 0.602 0.554 0.561 0.575 0.634 0.634

0.586 0.665 0.731 0.763 0.771 0.753 1 0.353 0.396 0.349 0.393 0.440 0.547 0.548

0.412 0.509 0.570 0.593 0.560 0.584 0.481 2 0.432 0.362 0.391 0.426 0.517 0.517

0.370 0.471 0.352 0.580 0.590 0.583 0.468 0.485 3 0.934 0.913 0.871 0.869 0.869

0.301 0.410 0.500 0.545 0.558 0.558 0.443 0.453 0.985 4 0.901 0.873 0.863 0.863

0.281 0.397 0.492 0.534 0.548 0.549 0.461 0.469 0.983 0.984 5 0.861 0.860 0.860

0.268 0.389 0.490 0.535 0.553 0.561 0.490 0.493 0.970 0.980 0.982 10 0.886 0.886

0.304 0.415 0.515 0.558 0.584 0.604 0.548 0.543 0.943 0.952 0.960 0.971 20 0.886

0.304 0.415 0.515 0.560 0.584 0.604 0.548 0.542 0.943 0.952 0.960 0.971 0.971


19
