Introduction
With growing human pressure on nature, the quality of agricultural landscapes has become increasingly important for protecting biodiversity (Karp et al.
2012). Land use intensification leads to landscape simplification (Landis
2017) with homogenized biological communities as a consequence (Gámez-Virués et al.
2015). Woody pastures, consisting of a variety of grazed semi-natural grassland habitats with some tree cover (Bergmeier et al.
2010), contribute significantly to landscape heterogeneity and biodiversity (Lentini et al.
2012; Wood et al.
2017). Woody pastures have typically been shaped by a long history of silvo-pastoral management, where the combination of grazing and presence of old trees enables co-existence of grassland species and species associated with old growth forests (Bergmeier et al.
2010; Dorresteijn et al.
2013; Garbarino & Bergmeier
2014). Multiple taxonomic groups therefore benefit from the heterogeneous structure of woody pastures, among which species richness and composition of plants, bats and birds have been the most studied (e.g. Fischer et al.
2010; Jakobsson & Lindborg
2015; Wood et al.
2017). However, research suggests diverging responses to pasture tree densities across taxa. For example, plant species richness often peak when canopy cover is 30–50% (e.g. Rolo et al.
2016), but in some contexts plant diversity increases linearly with increasing tree densities (Jakobsson and Lindborg
2015). The trees in woody pastures are also known to enhance bat species richness (Lentini et al.
2012), and the mere presence of a few trees may act as keystone habitats for bats (Tews et al.
2004; Fischer et al.
2010). Species richness of birds in pastures tend to saturate at low to intermediate tree densities (Tellería
2001), although shifts in bird species composition may be more strongly affected compared to species richness (Jakobsson and Lindborg
2017).
Woody pastures have been declining throughout Europe despite their high cultural and natural values (Bergmeier et al.
2010). During the last decade, the European Union (EU) has regulated the number of trees farmers are allowed to have on their pastures to receive subsidies for enhancing biodiversity (European Commission
2014). Therefore, tree density has inadvertently become a main characteristic affecting management of woody pastures in many countries such as Sweden (Beaufoy et al.
2011). Using tree density in woody pastures as an indicator of their quality may be an intuitive way to guide management practices that are easy to manage in practice. However, not only trees but also other vegetation characteristics act as local filters for biological communities (Zobel et al.
1998) and contribute to habitat heterogeneity, which in turn increases niche space and hence species richness (Rosenzweig
1995; Benton et al.
2003). Vegetation attributes in general also tend to have contrasting effects on different organism groups. For example, increasing understory vegetation may decrease bat species richness (Kalda et al.
2015b) and increase bird species richness (Laiolo et al.
2004), and grassland structural complexity in terms of vegetation height and patchiness may positively affect birds and arthropods but not plants (Lengyel et al.
2016). Multi-taxa approaches that consider a variety of structural habitat components are therefore needed to understand how biodiversity responds to vegetation attributes in woody pastures.
In addition to studying multiple taxonomic groups, examining functional rather than taxonomic diversity may help elucidate why species richness in different taxonomic groups show diverging responses to habitat and landscape characteristics in woody pastures. In particular, combining metrics of taxonomic and functional diversity provides a more holistic assessment of the multifaceted nature of biodiversity (Devictor et al.
2010). Whereas taxonomic diversity has been the traditional focus in conservation science, functional diversity of organisms is a better predictor of ecosystem functioning than taxonomic diversity (Gagic et al.
2015). By considering traits that determine how organisms respond to environmental change, a variety of functional diversity metrics can be used to improve the mechanistic understanding of changes in community composition along environmental gradients (Mouillot et al.
2013). Multi-taxa functional approaches have been used in a few recent studies (e.g. Barbaro and van Halder
2008; Pedley and Dolman
2014; Jonason et al.
2017), although none of these have focused on pasture structure and tree density specifically. Furthermore, the above studies have not evaluated whether relationships between functional diversity and the environment are driven by similar traits across multiple taxonomic groups. If similar traits drive patterns in functional diversity metrics consistently across taxonomic groups, it might indicate that functional diversity across taxa is shaped by similar ecological mechanisms. As an example, wing morphology and foraging behavior of bat assemblages relate to landscape characteristics and habitat structure in agroecosystems (Cisneros et al.
2015) and may cause shifts in the functional response of bat communities. Similar responses can be expected concerning birds, where mobility traits and foraging traits could result in a functional shift along gradients of habitat and landscape structure (Karp et al.
2012; Trisos et al.
2014), although mobility traits in birds also respond to patterns in migration strategy or home range sizes (Runge et al.
2014; Dufour et al.
2019).
In this study, we evaluated how the structure of woody pastures affects biodiversity, by analyzing taxonomic and functional diversity in three taxonomic groups: plants, bats and birds. Whereas plants and birds are commonly used biodiversity indicators (Wolters et al.
2006), bats are an understudied group of conservation concern (Jones et al.
2009b). Furthermore, these three taxa have different ecological requirements regarding their occurrence within woody pastures, including environmental characteristics at different spatial scales (e.g. Söderström et al.
2001; Kalda et al.
2015a). Bats and in particular birds are mobile and respond to both local habitat quality and landscape context, whereas plants may respond more strongly to local habitat factors. Bats may be more dependent on feeding related aspects compared to birds, which show a larger variation in home-range size and dietary requirements.
Our primary objective was to analyze the relative importance of tree density and other local structural attributes of woody pastures of potential relevance from a management perspective on taxonomic diversity and multiple functional attributes of species assemblages across three different taxa. To estimate functional diversity, we used functional traits from plants, bats and birds that reflect variation between species regarding morphology, dispersal/mobility, height/foraging level, nutrient requirement/feeding, reproduction and life history. We analyzed the diversity metrics in relation to local structural vegetation attributes of the woody pastures by using tree density, shrub density and structural complexity, in parallel with controlling for proximate forest cover (see “
Methods” for details).
We addressed the following three questions: (1) Are species richness and functional diversity metrics correlated among plants, bats and birds in woody pastures? (2) Are plant, bat and bird species richness and functional diversity metrics equally affected by structural vegetation attributes of woody pastures? and (3) How do specific traits and functional trait groups contribute to various functional diversity metrics, and their relations to structural vegetation attributes? We test the following hypotheses: (i) Diversity metrics generally show low congruence among the three taxa (Manning et al.
2015; Birkhofer et al.
2018). (ii) In contrast to species richness (cf. Jakobsson and Lindborg
2015,
2017; Wood et al.
2017), the functional response of bats and birds is more strongly linked to other structural vegetation attributes in and around the woody pastures than to tree density (Jung et al.
2012; Barbaro et al.
2014; Cisneros et al.
2015), whereas the functional response of plants is mainly affected by tree density (Rolo et al.
2016). Finally, (iii) traits belonging to the same functional groups (e.g. morphology) respond similarly to structural vegetation attributes and affect functional diversity metrics similarly across the three taxa (Luck et al.
2012; Dehling et al.
2014).
Discussion
This study demonstrates a general lack of congruence between different diversity measures of plants, bats and birds in woody pastures in mosaic landscapes in southern Sweden. We also found contrasting effects of woody pasture characteristics on different taxa and diversity components, with increasing tree densities mainly benefiting taxonomic diversity whereas other vegetation attributes to a larger extent explained variation in the functional response. Concerning bats and plants, the functional response to woody pasture characteristics was stronger compared to the taxonomic response, whereas this was the opposite for birds, suggesting that a conservation strategy based of tree density in wood pastures fails to benefit both taxonomic and functional diversity in multiple organism groups. Finally, whereas traits directly linked to resource use (height/foraging level and nutrient requirement/feeding) had the strongest impact on the functional response in plants and birds, morphology (body mass) was most important in determining bat functional responses, most likely because bats showed less variation in resource use compared to plants and birds. In the following, we discuss the conservation implications of i) the contrasting taxonomic and functional responses across the three taxa, and ii) how our trait-based approach sheds light on the ecological mechanisms that underpin taxon-specific diversity patterns in woody pastures.
Correlations between diversity metrics
There was only one significant relationship between corresponding diversity metrics across taxa (FRic of plants and bats), confirming our first hypothesis (Maas et al.
2016; Birkhofer et al.
2018). Our results highlight that functional diversity in general does not increase in tandem with increasing taxonomic diversity in woody pastures. Low congruence between different indices of functional diversity is in line with earlier research (Jonason et al.
2017), but may also in part be a consequence of our approach to match trait data across taxa that differ in resolution between taxa, which may result in an underestimation of the covariance in functional diversity metrics between plants, bats and birds. As an example, we used Ellenberg indicator values as a proxy for resource use in plants, which only partly reflects the total nutrient requirements of plant species. In contrast to plants, we used the complete feeding gradient for bats and birds. Furthermore, bat species in our study area are 100% insectivorous and thus we used different arthropod taxa to quantify bat feeding traits, in contrast to the wider variation in bird feeding strategies. Whilst we note that alternative approaches to measure diet specialization among bats and birds are available (cf. Barbaro et al.
2019), we believe that our approach provides a good representation of diet preferences for analyzing functional diversity among bats and birds in our study system. However, because of these differences in trait resolution, we acknowledge that the relationships between a larger functional space in plant communities and higher trophic levels in our study should be interpreted with caution (Perner et al.
2005).
Diversity responses to the environment
In line with our second hypothesis, tree density had less pronounced effects on the functional response of bats and birds compared to species richness, whereas the positive effect of tree density on plant species richness extended to the functional response (Rao’s Q). This shows that tree density, more than other vegetation features, determine multi-taxa species richness, but that multiple vegetation features affect the multi-taxa functional response in woody pastures.
The plant communities showed similar functional responses as for species richness, with no decoupling of taxonomic and functional diversity metrics along the tree density gradient (cf. Rolo et al.
2016). Increasing tree density was associated with an increasing total functional space (FRic) in plants in woody pastures. We also show that increasing tree densities resulted in increasing plant functional diversity due to the addition of functionally different species (Rao’s Q) along the tree density gradient. Our results are therefore in line with the expectation that woody pastures create conditions that favor a variety of species that are associated to both open grassland and forests (Garbarino and Bergmeier
2014; Jakobsson and Lindborg
2015), which contributes to the unique conservation values of these habitats (Bergmeier et al.
2010).
Similar to Lentini et al. (
2012), bat species richness increased with increasing tree density, though with rather low explanatory power, which may to some extent be explained by the fact that bat sampling was restricted to one night per site in our study. Notably, the functional response of bats to habitat characteristics was much stronger than the taxonomic response, probably because of expected guild-specific responses to vegetation features (Müller et al.
2012; Kalda et al.
2015a; Barbaro et al.
2019). Increasing FDiv with increasing tree density in woody pastures suggest that bat communities had lower overlap in functional space in conditions with high tree densities. In addition, a higher forest cover in the proximate landscape was associated with higher FRic in bat communities, suggesting that landscape context is important for the range of niches that bat communities span in woody pastures. Thus, our results confirm the importance of vegetation structures at multiple spatial scales on bat diversity in mosaic landscapes (Kalda et al.
2015a).
Bird species richness responded strongly to pasture habitat characteristics, peaking at intermediate tree and shrub densities, consistent with earlier research (Tellería
2001; Laiolo et al.
2004). In contrast, no explanatory variable had a significant effect on any of the bird functional diversity metrics. Birds commonly use multiple habitats scattered across agricultural landscapes, and landscape effects on bird communities are well known (e.g. Karp et al.
2012), and therefore it is not surprising that small-scale habitat structures in woody pastures did not significantly affect bird functional diversity metrics in our study. Although we acknowledge that our analyses may have suffered from low statistical power, studies based on large datasets have shown that landscape patterns even within relatively large extents (25 hectares) may have weak effects on bird functional diversity (Jonason et al.
2017). However, individual traits showed significant responses to the vegetation attributes, which indicate a stronger direct environment-trait link, but might also be an artifact from a larger trait variability gradient among birds compared to the other two taxa.
Because of short-term data collection over different sampling years for each taxon, we had limited precision in interpreting inter-taxa relationships and effects of the environmental variables. However, because weather and local management and habitat conditions were similar across all study years (see “
Methods”), we believe that extant variation attributable to different sampling years had a minor influence on the relative relationship between multi-taxa functional diversity metrics and effects of explanatory variables. Hence, we believe that our analysis of the functional responses accurately reflects the main underlying ecological patterns in the woody pastures in our study. We would expect similar responses in comparable mosaic landscapes including habitats with woody cover (Tews et al.
2004), but in other systems the outcome will depend on the contrast between the focal habitats and properties of the surrounding landscapes.
Our main objective was to implement a holistic approach including taxonomic and functional characteristics of species assemblages to understand the biodiversity effects of pasture management, which are policy-driven practices that typically act at local scales without capturing the ecological context at larger scales (European Commission
2014; Ekroos et al.
2017). To this end, our study highlights how simplified management goals linked to what is regarded best management practice based on one taxonomic group (in many cases the plant community) (e.g. Minden et al.
2016; Tälle et al.
2016), or one habitat characteristic, may fail to enhance the broad majority of biological values (see also Söderström et al.
2001). We also note that including taxonomic groups from complementary trophic levels, with contrasting mobility and degree of habitat structure specialization (e.g. arthropods), would be needed to show the full potential of the effects of different management practices in woody pastures on biodiversity.
Drivers of functional diversity responses
In contrast to our third hypothesis, effects of corresponding trait groups on functional diversity metrics or specific traits’ relationship to the environment showed little congruence across taxa, suggesting that the ecological mechanisms determining community structure may differ substantially between the studied taxonomical groups. The trait exclusion approach showed that functionally distinct plant species regarding light and nutrient requirements contributed strongly to increasing functional richness (FRic) in plant communities as tree density increases. Trees create a heterogeneous environment where resource competition during plant community establishment enables a rich functional space (Belsky
1994; Martens et al.
2000; see also Jakobsson and Lindborg
2015). In addition, the same traits contributed to a parallel increase in plant Rao’s Q with increasing structural complexity, which emphasizes how vegetation complexity has a positive impact on plant functional diversity also at this small spatial scale (cf. Ma and Herzon
2014). However, only in the case of plants did the same traits that contributed most strongly to functional responses also respond individually to environmental factors. In contrast, bats’ functional responses were mainly linked to morphology (body mass), with additional effects of mobility (forearm length) and longevity. Variation in body mass contributed to increases in bat FRic and FDiv with increasing forest cover and tree density, respectively, indicating morphologically driven niche differentiation as a response to tree cover at local and proximate scales. We note that there is a relatively clear division of smaller (
Myotis,
Pipistrellus and
Plecotus sp.) and larger (
Eptesicus,
Nyctalus and
Verspertilio sp.) species in the study area, which might explain this link to morphology. The functional space in bat communities becomes more evenly distributed with increasing proximate forest cover because of (i) increases in interface specialists foraging within the semi-open pastures due to additional vegetation features in the surroundings (cf. Wood et al.
2017), (ii) increases in bat species with relatively short forearms, adapted to denser vegetation, without simultaneous losses of bats with long forearms, and (iii) increases in bats with higher longevity, because most long-lived species in our study area prefer forests. Our results confirm that heterogeneous vegetation structures are important for bats in semi-open habitats (Adams et al.
2009) by moderating the functional space of bat communities through species’ differences in general habitat preferences and maneuverability (Jung et al.
2012; Müller et al.
2012). In line with Ikin et al. (
2012), variation in food preferences and foraging strategies among birds responded significantly to vegetation characteristics, especially increasing tree density, driving decreases in the number of ground foragers and seed eaters whilst favoring mid-high foraging birds (Table S7). However, among birds, functional diversity metrics remained unaffected, highlighting how individual traits of birds may be influenced by local habitat characteristics in woody pastures without resulting in an overall functional response.
Conclusions
Our study showed that there is little congruence between corresponding taxonomic and functional diversity metrics of plants, bats and birds in woody pastures. Further, we found that increasing tree densities in woody pastures benefitted species richness of all three taxa, whereas functional diversity metrics responded to a broader set of vegetation characteristics, including increasing shrub density, structural complexity, proximate forest cover and tree density. The results highlight the need to account for a broader suite of vegetation attributes in management prescriptions, rather than simply limiting high tree densities to preserve high functional diversity in woody pastures, especially regarding communities of mobile organisms like bats and birds. Furthermore, by using a trait group exclusion approach we were able to quantify the underlying ecological effects that contributed most to the observed functional responses. Notably, we found that the functional responses of plants and birds were driven by traits related to resource use, whereas body mass was the main driver of bat functional responses. Our study also shows that trait-specific responses are not necessarily directly linked to functional diversity responses (as for bats and birds), which may to some extent reflect the fact that birds and bats utilize multiple habitat types in mosaic landscapes. We therefore encourage further studies including similar approaches to ours, but in different habitat and landscape contexts. The absent links between diversity of these trophic levels and the complex responses to habitat structure show that multi-taxa functional approaches investigating effects of pasture attributes are needed to inform woody pasture management. By increasing the mechanistic understanding of multi-taxa functional diversity responses there are even better chances to preserve the high biological values in woody pastures in mosaic agricultural landscapes.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.