A phylogenetic approach to conservation prioritization for Europe's bumblebees (Hymenoptera: Apidae: Bombus)

Biological Conservation 206 (2017) 21–30

Contents lists available at ScienceDirect

Biological Conservation journal homepage: www.elsevier.com/locate/bioc

A phylogenetic approach to conservation prioritization for Europe's bumblebees (Hymenoptera: Apidae: Bombus) Nicolas J. Vereecken Agroecology & Pollination Group, Landscape Ecology & Plant Production Systems Unit, Université Libre de Bruxelles (ULB), Boulevard du Triomphe CP 264/2, Boulevard du Triomphe, B-1050 Brussels, Belgium

a r t i c l e

i n f o

Article history: Received 9 September 2016 Received in revised form 2 December 2016 Accepted 9 December 2016 Available online xxxx Keywords: Phylogeny Bumblebees IUCN Red List Evolutionary distinctiveness Phylogenetic diversity Range size Conservation prioritization

a b s t r a c t Bumblebees are an essential component of our agroecosystems, and their decline represents a major threat for the sexual reproduction — and hence survival — of wild flowers and several important pollinator-dependent crops alike. The EU bumblebee fauna encompasses many highly imperiled species characterized by a relatively narrow range size and often restricted to high elevation mountain habitats where the threats of both current and future global warming are expected to be particularly severe. In this context, identifying how and where limited conservation resources should be targeted is a pressing priority to meet our fundamental biodiversity conservation targets in an economically-efficient way. Because classical taxonomic approaches to conservation can potentially overlook important alternative aspects of biodiversity such as the phylogenetic diversity, a key component for the maintenance of ecosystem processes and services, I used a multi-gene molecular phylogeny encompassing more than 85% of the EU species to combine categories of the IUCN Red List with the evolutionary legacy and range size of EU bumblebees. My results from phylogenetic generalized least squares (PGLS) and phylogenetic independent contrasts (PIC) analyses first indicate that, contrary to theoretical prediction, evolutionary relatedness explains none of the range size similarity or the probability of extinction risk in EU bumblebees. Furthermore, although the extinction of extant threatened EU bumblebee species is unlikely to have a significant effect on the expected phylogenetic dispersion of the remaining Bombus species at the EU scale, my results clearly illustrate that a significantly disproportionate amount of phylogenetic diversity/evolutionary history might be lost if the extant threatened EU bumblebee species would become extinct. Collectively, this study exemplifies the fact that the incorporation of a phylogenetic approach can increase the efficacy of the existing prioritization for the conservation of EU bumblebees (i) by capturing the phylogenetic diversity and its associated functions, as well as (ii) by better targeting species that are both evolutionarily unique (or non-redundant), threatened and restricted in their range size. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction As Darwin (1859) pointed out in his magnum opus “On the origin of species”, the tendency for closely-related species to exhibit similar habits, constitution and structure, is ubiquitous in nature. Evidence from modern-day phylogenetic trait mapping analyses across the tree of life supports — by and large — the idea that evolutionary relatedness explains the trait and sometimes ecological niche similarity between species, a pattern known as phylogenetic niche conservatism (Harvey & Pagel, 1991; Hansen & Martins, 1996; Revell et al., 2008; Wiens et al., 2010). This phenomenon has important implications, since it significantly impacts on a range of ecological and evolutionary processes and patterns, from fundamental aspects of e.g. co-occurrence probabilities of species, to the extent to which closely-related taxa of threatened

E-mail address: [email protected].

http://dx.doi.org/10.1016/j.biocon.2016.12.009 0006-3207/© 2016 Elsevier Ltd. All rights reserved.

species are expected to be more or less often at risk than their more distant relatives (Faith, 1992; Mace et al., 2003). While the IUCN Red List of Threatened Species represents a powerful tool for practical conservation planning (Rodrigues et al., 2006), conservation biologists today face the major challenge of setting conservation priorities in groups of sometimes highly diverse species for which a substantial part of the taxa evaluated are categorised as “Data Deficient (DD)”. In this context, the incorporation of species' contribution to phylogenetic diversity can successfully be taken into account for conservation planning, as it helps using alternative metrics of biodiversity to reach a better understanding of where and what diversity is at risk (Mace & Baillie, 2007; Mace & Purvis, 2008; Cadotte & Davies, 2010; Redding et al., 2010; Winter et al., 2013; Jetz et al., 2014; Nunes et al., 2015; Pollock et al., 2015). With an estimated 20,000 species described worldwide (Ascher & Pickering, 2016) and their key role in the pollination of flowering plants (Ollerton et al., 2011), bees rank among the groups of species that are increasingly the focus of conservation actions following reports of

22

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

large-scale declines. Like many other groups of wild bees, bumblebees have been reported to decline in developed regions such as in North America (Thorp & Shepherd, 2005; Cameron et al., 2011; Bartomeus et al., 2013) and in Western Europe where several authors have reported on apparent cases of widespread decline since the 1970's (Peters, 1972; Alford, 1973) and 1980's (Williams, 1982; Williams, 1986; Rasmont & Mersch, 1988; Goulson, 2003; Williams & Osborne, 2009). With the recent publication of the IUCN European Red List of Bees, Nieto et al. (2014) have provided an assessment of the status and trends of ca. 2000 species of wild bees, including all 68 species of EU bumblebee recorded of which an estimated 46% are thought to be threatened by extinction. Among the major causes of (bumble)bee decline, Nieto et al. (2014) have pointed out the intensification of agriculture, the increase of pollution from agricultural waste, the loss of habitat due to urban development and land use changes in agro-ecosystems such as the loss of permanent and unimproved grasslands as the main threats (see also González-Varo et al. (2013)). Climate change was also evoked as a significant explanatory variable in the equation by Nieto et al. (2014) and Rasmont et al. (2015) have stressed in their recent Climatic Risk and Distribution Atlas of European Bumblebees that the increasing temperatures and long periods of drought are severely threatening many of the EU bumblebee species, particularly those that live in high altitude habitats and characterized by very small range sizes (see also Kerr et al. (2015)). Indeed, although the geographic range of the world bumblebees spans a wide range of latitudes and habitat types, their tendency to be adapted to cold and temperate regions of the Northern Hemisphere (Williams, 1998; Hines, 2008) has led many species to be characterized by a relatively narrow range size, e.g. in high elevation mountain ranges where the threats of current and future global warming are particularly severe (Rasmont et al., 2015; Kerr et al., 2015). A narrow range size has therefore the potential to inflate the vulnerability of bumblebee species, a phenomenon also observed in other terrestrial animal and plant species with a smaller geographic range that are assumed to be comparatively more threatened by extinction, all other factors being equal (Cardillo et al., 2008; Harris & Pimm, 2008; Runge et al., 2015). Bumblebees are an essential component of our agroecosystems, and their decline represents a major threat for the sexual reproduction — and hence survival — of wild flowers and pollinator-dependent crops alike. Furthermore, because Bombus is the only extant genus of the corbiculate bees in the tribe Bombini (Michener, 2007), it is of prime importance to assign priorities to conservation efforts by taking into account not only species as independent units, but also the fact that the extinction of more evolutionarily unique Bombus species would imply a greater loss of phylogenetic diversity (PD) (see also Isaac et al. (2007), Hidasi-Neto et al. (2013), Faith (1662), Forest et al. (2015), and Huang and Roy (2015)). PD is often used as a proxy for functional diversity (FD), an important component for the maintenance of ecosystem processes and services (Faith, 1662; Tilman et al., 1997a; Tilman et al., 1997b; Cadotte et al., 2008; Cadotte et al., 2009; Cadotte et al., 2012; Gravel et al., 2011; Rolland et al., 2011), particularly in the context of pollination services provided by wild bees (Hoehn et al., 2008; Albrecht et al., 2012; Fründ et al., 2013; Martins et al., 2015). Consequently, the conservation of higher PD is likely to safeguard the evolutionary legacy of bumblebees, their potential to face environmental changes, and can be anticipated to be pivotal for the maintenance of present and future ecosystem processes and services (Faith, 1662; Cadotte et al., 2012; McNeely, 1988; Faith et al., 2010). The time is now ripe for the incorporation of PD as an alternative biodiversity metric into conservation planning to avoid worst-case losses of long branches from the bumblebee tree of life (see e.g. Davies (2015) discussing this issue with a focus on the South African Cape flora) and their associated ecological/economic consequences (see also Kleijn et al. (2015) and Potts et al. (2016)). To address this issue and to assess the relevance of a phylogenetically-informed conservation action of EU bumblebees

compared to a scenario of conservation priorities based solely on the IUCN categories, I combined a molecular phylogeny with recent data on the distribution range (Nieto et al., 2014; Rasmont et al., 2015) and the conservation status of bumblebees (Nieto et al., 2014) in Europe. Using ancestral character estimation, phylogenetic independent contrasts (PIC) and phylogenetic generalized least squares (PGLS), I aim (i) to determine the extent to which the evolutionary legacy of bumblebees explains biogeographic patterns (range size in particular), and (ii) to test if IUCN Red Listed species that are threatened at the EU scale due to population declines are phylogenetically more closely-related or more evolutionarily unique than expected by chance alone. Next, I also computed the EDGE (Evolutionary Distinct, Globally Endangered) metric (iii) to analyse and combine the degree of evolutionary distinctiveness (ED) along with the extinction risk (weighted IUCN categories). Last, I discuss these results to assess conservation priorities by examining the extent to which the current IUCN Red List categories also captures alternative aspects of diversity such as the phylogenetic diversity of EU bumblebees, and how species that are both evolutionarily unique (or non-redundant), threatened and restricted in their range size (i.e., “evolutionary distinctiveness rarities”, EDR) can be better targeted for conservation action. 2. Materials & methods 2.1. Molecular phylogeny of EU bumblebees To prepare a phylogenetic tree for the EU bumblebees whose first branch lengths capture the expected similarities and differences among species, I used the most likely Bayesian tree (Tree ID #Tr2906 on https://treebase.org) produced by (Hines, 2008) based on (Cameron et al., 2007) using mixed-model Bayesian analyses for a study on the historical biogeography, divergence times, and diversification patterns of bumblebees at the worldwide scale, including representatives of all 38 subgenera (Williams, 1998). This multi-gene molecular phylogeny encompasses 229 Bombus species; the sequence data include ~ 3745 amplified nucleotides, including both intron and exon regions from 5 genes: mitochondrial 16S rDNA, elongation factor-1α F2 copy (EF-1⍺), long-wavelength rhodopsin (opsin), arginine kinase (ArgK), and phosphoenolpyruvate carboxykinase (PEPCK) (Hines, 2008; Cameron et al., 2007). I then pruned this tree with the ape package (Paradis et al., 2004), which yielded in a near-complete species-level phylogenetic tree comprising 59 Bombus species out of the 68 species listed at the EU scale, i.e. more than 85% of the EU bumblebee species count (Nieto et al., 2014). This phylogenetic tree was used for all statistical analyses in the present study by adjusting branch lengths with the rho parameter set to 1 using Grafen's (1989) method as implemented in the ape package (Paradis et al., 2004). The final tree had therefore scaled node heights so that the root height was equivalent to 1. 2.2. Phylogenetic diversity (PD) associated to threatened EU bumblebees To estimate the taxonomic (TD) and phylogenetic diversity (PD) decline associated to a loss of Red Listed bumblebee species, I computed the TD and PD associated to 4 distinct communities: (i) all 58 EU bumblebee species, (ii) all 58 EU bumblebee species without species categorised as CR and EN on the IUCN Red List (Nieto et al., 2014), (iii) all 58 EU bumblebee species without species categorised as CR, EN and VU on the IUCN Red List (Nieto et al., 2014), and (iv) all 58 EU bumblebee species without species categorised as CR, EN, VU and NT on the IUCN Red List (Nieto et al., 2014). This analysis was performed with the picante package (version 1.6-2) (Kembel et al., 2010); calculations of PD are measured for each community as the total branch length of a tree linking all species represented in this particular community (Faith, 1992).

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

I then calculated the standardized effect size (SES) of PD to predict if the loss of species from one of the above communities to the next was associated to a decrease of PD similar to what one might expect by chance using n = 9999 permutations. Because the calculation of PD's SES is a two-tailed test, the observed p-value associated to PD's SES was compared to p-values of 0.025 and 0.975 in order to determine if the observed loss of PD from one of the above communities to the next is lower or higher than expected by chance alone. To illustrate these results and the extent to which phylogenetic diversity is at risk, I used a custom R function by José Hidasi Neto's (https://github.com/hidasi/rfunctions) to plot a dendrogram-based extinction curve of Faith's (1992) phylogenetic diversity (PD), analogous to the well-known survival curves. This figure therefore illustrates the degree of PD loss associated to the extinction of EU bumblebee species compared to a null community (n = 100 randomizations), assuming that the EU bumblebee species would become extinct following the order of the IUCN categories, since the latter are expected to indicate how near a species is to extinction. 2.3. The interplay between IUCN range size and phylogeny To test the phylogenetic signal of EU range size — estimated by Rasmont et al. (2015) as the number of 50 km square grids occupied by each EU bumblebee species for the period 1970–2000 at the EU scale — I first had to exclude B. (Pyrobombus) modestus Eversmann, 1852, a species that only lives in a few locations in boreal forests between Moscow and the Ural Mountains (Rasmont et al., 2015) for which the range size was not available. I acknowledge the fact that some species in this dataset have a global range size extending far beyond the EU (e.g. for B. (Bombus) patagiatus Nylander, a species with an Eurasian distribution — see Williams et al. (2012a)—, or B. (Cullumanobombus) cullumanus (Kirby) — see Williams et al. (2012b), but I performed the analyses on the EU range size as a purely pragmatic step, because expertise and data were more readily available for this region and because this geographic framework is relevant for conservation action and policy. For the remaining 58 species, I first tested the phylogenetic signal in range size, i.e. the tendency of closely-related species to be more similar than species drawn at random from the same tree, I first computed Blomberg's K with the phytools package (version 0.5-10) (Revell, 2012) to test the extent to which the distribution of range size on the EU bumblebees phylogeny is equivalent to that expected by Brownian motion of trait evolution (Monte-Carlo tests, n = 999 randomizations). I then used the adephylo package (Jombart & Dray, 2008) to compute Abouheif's Cmean (Abouheif, 1999), with Monte Carlo simulations and 9999 randomizations) and the phytools package (version 0.5-10) (Revell, 2012) to calculate Pagel's λ (Pagel, 1999; Freckleton et al., 2002). These two metrics are known to perform well to help statistically test for a pattern of phylogenetic trait conservatism (Münkemüller et al., 2012). I then tested whether or not threatened species of the IUCN Red List (“CR”, “EN”, and “VU” categories) had similar mean EU range size compared to non-threatened taxa (“NT”, “LC” and “DD” categories). I used the nlme package (version 3.1-121) (Pinheiro et al., 2015; Pinheiro et al., 2016) to perform a phylogenetic generalized least squares (PGLS) analysis to account for the non-independence of the closelyrelated bumblebee species (see Harvey and Pagel (1991) and Felsenstein (1985)) in this analysis. The approach of the PGLS is akin to the general least squares (GLS) method, but for the PGLS any species pair is expected to be correlated owing to their shared evolutionary history. The phylogenetic correlation matrix is used to weight the data from each species in a fitted linear model. 2.4. The interplay between IUCN conservation status and phylogeny I first used the caper package (version 0.5.2) (Orme et al., 2013) to compute the D statistic developed by Fritz and Purvis (2010) to measure

23

the phylogenetic signal in a binary trait. Here, the D statistic was used to characterize the phylogenetic structure (i) of “Data Deficient (DD)” species, and (ii) of threatened species of the IUCN Red List (aggregated “CR”, “EN”, and “VU” categories). I tested the observed D value for significant departure from both random association and the clumping expected under a Brownian evolution threshold model. The value of D can vary from smaller than 0 (phylogenetically clustered pattern) to greater than 1 (phylogenetically overdispersed pattern) and the distributions of scaled D from n = 9999 permutations are used to assess the significance of the observed scaled D compared to 0 and 1, respectively. I measured the amount of phylogenetic diversity (hereafter, PD) (Faith, 1992) captured by species within each IUCN category (DD, LC, NT, VU, EN, and CR) as the sum of all branch lengths within the phylogeny connecting the species belonging to that particular assemblage/ IUCN category. To test whether each IUCN category captured PD than expected by chance, I used null models to compare observed PD values with 999 randomly generated values by shuffling taxa labels across tips of the phylogeny (across all taxa included in phylogeny). These analyses were performed with the picante package (version 1.6-2) (Kembel et al., 2010). To investigate if (i) “Data Deficient (DD)” species, and (ii) threatened species of the IUCN Red List (“CR”, “EN”, and “VU” categories) of bumblebees are more closely-related than expected by chance, I computed the standardized effect size (SES) of the mean pairwise phylogenetic distances (MPD) for species assemblages comprising only the two categories of species (i.e., “DD” species vs. all other species, and species classified in the “CR”, “EN”, and “VU” categories vs. non-threatened categories “NT”, “LC” and “DD”). SES compare the value for a community to the mean expected under a random-draw null model correcting for the standard deviation and allow to evaluate the magnitude of the difference between groups independently of sample size, thereby providing more detailed information than the significant differences highlighted by the p value alone (see e.g. Sullivan and Feinn (2012)). I calculated the Nearest Relative Index (NRI) as the inverse of MPD's SES and because the calculation of this value is a two-tailed test, the observed p-value associated to MPD's SES was compared to p-values of 0.025 and 0.975 in order to determine if the observed NRI value is lower or higher than expected by chance. These analyses were also performed with the picante package (version 1.6-2) (Kembel et al., 2010). A negative value of the NRI metric indicates a trend towards phylogenetic overdispersion such that species are, on average, less closely related to one another than expected by chance with a null model implemented by shuffling the names of species on the EU bumblebees phylogenetic tree 999 times. By contrast, a positive NRI value indicates a trend towards phylogenetic clustering (Webb et al., 2002). I also computed the Mean Nearest Taxon Distance (MNTD), a metric that provides an average of the distances between each species and its nearest phylogenetic neighbor in the community (Webb et al., 2002).

Table 1 Estimated taxonomic (TD) and phylogenetic diversity (PD) decline associated to a loss of bumblebee species categorised as threatened in the IUCN Red List by Nieto et al. (2014). The threatened categories comprise species listed as “Near Threatened (NT)”, “Vulnerable (VU)”, “Endangered (EN)”, and “Critically Endangered (CR)”. The asterisk indicates that the loss of NT species after the loss of species belonging to the former 3 categories resulted in significantly lower PD than what would be observed in a null community (PD's SES = −1.9179, associated p-value = 0.0209), i.e. a comparatively disproportionate loss of evolutionary history results from the loss of 15 species in the CR, EN, VU and NT categories from the original 58 EU bumblebee species assemblage. Assemblages & IUCN categories All EU Bombus species - Without EN species - Without EN & VU species - Without EN & VU & NT species

Observed Observed TD TD PD lost/(%) 58 54 46 43

9859 9.754 8.438 7.192

– 4/6.89 12/20.69 15/25.86

PD lost/(%) – 0.1052/1.06 1.01/14.41 2.667/27.04⁎

24

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

Table 2 Mean Pairwise Phylogenetic Distance (MPD) and Mean Nearest Taxon Phylogenetic Distance (MNTD) in species assemblages associated to Red Listed EU bumblebee species. The results show that, given the species richness of the tested assemblages, the phylogenetic dispersion of species (i.e., the Net Relatedness Index (NRI) and the Nearest Taxon Index (NTI) computed as the inverse of the standardized effect size of MPD and MNTD) is never significantly different from what is expected in a null community as the EU bumblebee species would become extinct following the order of the IUCN categories. Assemblages & IUCN categories

MPD

All Bombus species

NRI

- Without EN species - Without EN & VU species - Without EN & VU & NT species

−0.5447 0.5607 1.3383

MNTD p-Value

NTI

p-Value

0.6360 0.2600 0.0990

−0.3008 −0.0738 1.0226

0.4330 0.4655 0.1565

In short, whereas MPD is the average distance between two random individuals in a sample, MNTD is the average distance to the closest heterospecific individual for all individuals in the sample. Just as the NRI was computed above, I calculated the Nearest Taxon Index (NTI) as the inverse of MNTD's SES and because the calculation of this value is a two-tailed test, the observed p-value associated to MNTD's SES was compared to p-values of 0.025 and 0.975 in order to determine if the observed NTI value is lower or higher than expected by chance with a null model implemented by shuffling the names of species on the EU bumblebees phylogenetic tree 999 times. These analyses were also performed with the picante package (version 1.6-2) (Kembel et al., 2010). While MPD described above is more strongly influenced by branch lengths at the deepest nodes of the phylogeny, MNTD is necessarily sensitive to replacement of closely related taxa, and both metrics have already been found to be independent from species counts (Swenson, 2009). These two metrics are therefore complementary when discussing patterns of phylogenetic community structure: the associated metric NRI describes a tree-wide pattern of dispersion, whereas NTI is more sensitive to phylogenetic structure towards the tips of the phylogeny. 2.5. Combined analysis of evolutionary distinctiveness and extinction risk Isaac et al. (2007) introduced the EDGE (Evolutionary Distinct, Globally Endangered) metric to combine the analysis of a species' evolutionary distinctiveness (ED) and extinction risk (IUCN categories) following this formula: EDGE ¼ ln ð1 þ EDÞ þ GE  ln ð2Þ where ED is the evolutionary distinctiveness (the degree of a species' isolation on the phylogenetic tree) calculated with the picante package

(version 1.6-2) (Kembel et al., 2010) with “equal splits” (Redding & Mooers, 2006), and GE is the IUCN Red List category weight [“Data Deficient (DD)” = 0, “Least Concern (LC)” = 0, “Near Threatened (NT)” = 1, “Vulnerable (VU)” = 2, “Endangered (EN)” = 3, “Critically Endangered (CR)” = 4]. Finally, I computed the evolutionary distinctness rarity of each individual EU bumblebee species (EDR = ED/species range size) introduced by Jetz et al. (2014) as a metric whose distribution across phylogenetic trees can potentially strongly differ from ED. The advantage of EDR over ED is that the former captures how geographically-restricted ED can be; high-EDR bumblebees in the present study therefore represent species of particular conservation concern under global change, since they are expected to be both evolutionarily unique (or non-redundant) and restricted in their range size. I used the phytools package (version 0.5-10) (Revell, 2012) to map and plot the EDGE and EDR scores as continuous variables on the EU bumblebees phylogeny. The phytools package (version 0.5-10) (Revell, 2012) allows for a fast estimation of the ML ancestral states for these continuous traits, and the interpolating of the states along each edge is done using Eq. (2) of Felsenstein (1985). The matlab-like colour palette used to map the continuous traits on the EU bumblebees phylogeny was retrieved from the colorRamps package (Keitt, 2012). All statistical analyses were performed with RStudio (Version 0.99.489) for R (Core Team, 2015).

3. Results and discussion 3.1. Phylogenetic distribution of conservation statuses in EU bumblebees The analysis of IUCN Red Listed EU bumblebees reveals that out of the 58 species investigated here, 15 species (i.e., 26%) are threatened with extinction. Despite differences in species counts between the EU bumblebees and other groups of terrestrial organisms, it is important to note that the figure obtained here (26%) is slightly lower than the proportion of globally threatened with extinction in zooxanthellate corals (32% of 827 species) (Carpenter et al., 2008) and in amphibians (32% of 2030 species), but is more or less equivalent to the status of mammals (22% of 5488 species) and exceeds that of Mediterranean dragonflies (19% of 165 species) and of birds (12% of 1227 species) (IUCN, 2016). Using the D statistic for phylogenetic signal in a binary trait, I found that both “Data Deficient (DD)” and threatened species on the IUCN Red List are phylogenetically overdispersed, i.e. more distantly related than expected by chance (DD species: D = 0.7258, i.e. significantly different from 0 (p-value = 0.002) but not significantly different from 1 (pvalue = 0.07); Threatened species: D = 0.7279, i.e. significantly different

Table 3 The top 12 ranked species of EU bumblebees to target for conservation efforts based on their EDGE score, classified by decreasing EDGE score. The IUCN category of each species is listed in the first column, followed by the EU range size (number of 50 km square grids occupied by each species for the period 1970–2000 in Europe following Rasmont et al. (2015)), the evolutionary distinctiveness (ED; a measure of the degree of isolation of a species on its phylogenetic tree), the evolutionary distinctness rarity (EDR = ED / species range size), and the EDGE score which encapsulates the amount of unique evolutionary history it represents (i.e., its evolutionary distinctiveness, ED) combined and weighted by its IUCN conservation status (Isaac et al., 2007). Note that the first three species and B. (Alpinobombus) hyperboreus Schönherr, 1809 have a small to large range and high abundances outside of the EU geographical context (Williams et al., 2011). See Materials and methods section for details.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

EU bumblebee species

IUCN category

Range size

ED score

EDR score

EDGE score

B. (Subterraneobombus) fragrans (Pallas, 1771) B. (Thoracobombus) armeniacus Radoszkowski, 1877 B. (Thoracobombus) zonatus Smith, 1854 B. (Pyrobombus) brodmannicus Vogt, 1909 B. (Bombias) confusus Schenck, 1861 B. (Megabombus) gerstaeckeri Morawitz, 1881 B. (Thoracobombus) pomorum (Panzer, 1805) B. (Subterraneobombus) distinguendus Morawitz, 1869 B. (Thoracobombus) muscorum (L., 1758) B. (Alpinobombus) hyperboreus Schönherr, 1809 B. (Alpinobombus) polaris Curtis, 1835 B. (Alpinobombus) alpinus (L., 1758)

EN EN EN EN VU VU VU VU VU VU VU VU

36 11 28 4 157 52 166 502 625 33 57 86

0.3942 0.1124 0.0870 0.0569 0.9912 0.3462 0.2072 0.2059 0.1724 0.1003 0.1003 0.1003

0.011 0.010 0.003 0.014 0.006 0.006 0.001 0.0004 0.0002 0.003 0.001 0.001

2.4118 2.1859 2.1628 2.1347 2075 1.6836 1.5746 1.5735 1.5454 1.4818 1.4818 1.4818

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

from 0 (p-value = 0.005) but not significantly different from 1 (pvalue = 0.067)). These results are similar to the situation described by Schachat et al. (2016) who found significantly more phylogenetic signal in the cause of decline than in the IUCN threat ranking of Incilius toads. It seems important to point out that the subgenus Alpinobombus encompasses closelyrelated species with very high to extremely high climate change risk. Small ancestral ranges are likely to give rise to two small descendant ranges for which spatial autocorrelation in the Scandinavian mountains and along the northern tundra (Rasmont et al., 2015) is likely to be a shared cause of decline. By contrast, species like B. fragrans (EN) and B. distinguendus (VU) in the subgenus Subterraneobombus are best described as parapatric or slightly allopatric and are therefore unlikely to co-occur in a given area of Europe; both species have large ranges and higher abundances outside of the EU (Williams et al., 2011). The detection of phylogenetic overdispersion in communities is traditionally interpreted as originating from either a prevalent impact of

25

competition between closely-related species sharing traits, or from habitat filtering where distant relatives share convergent traits. It is important to note that these two scenarios are not mutually exclusive, and that a phylogenetically overdispersed pattern could be the outcome of an interplay between competition and habitat filtering (see e.g. Cooper et al. (2008)). The disentangling of the role of competition vs. habitat filtering in the phylogenetic overdispersion of threatened species on the IUCN Red List of EU bumblebees is beyond the scope of this paper. 3.2. Phylogenetic diversity loss associated to species extinctions Despite the loss of diversity observed as species from different IUCN categories were removed from the original species assemblage, our results show that most of the IUCN categories tested here gathered PD levels similar to what one might expect by chance (Table 1, Fig. 3). Using our dataset that encompasses 58 species of EU bumblebees, I

Fig. 1. Six bumblebee species considered to be of particularly high conservation concern according to the IUCN Red List of European Bees (Nieto et al., 2014) and highlighted in this study. Top Left: Bombus (Subterraneobombus) fragrans (Pallas, 1771), the largest bumblebee in Europe and a species that is restricted to true steppes of Central and Eastern Europe, as well as in the Anatolian plateau (Rasmont et al., 2015). This species is listed as “Vulnerable (VU)” (Nieto et al., 2014). Photo by G Holmström. Top Right: B. (Megabombus) gerstaeckeri Morawitz, 1881, a species characterized by a highly specialised diet — queens and workers collect pollen almost exclusively monkshood flowers (Aconitum spp., Ranunculaceae). This bumblebee usually forms small colonies and is found only in the high mountain ranges of southern Europe: the Pyrenees, the Alps, the Carpathians, and the Caucasus (Rasmont et al., 2015). This species is listed as “Endangered (EN)” (Nieto et al., 2014). Photo by NJ Vereecken. Middle left: B. (Thoracobombus) armeniacus Radoszkowski, 1877, a species found in steppes, forest steppes, and riparian meadows at the altitudes of 1500–2000 m above sea level with a geographic range that includes Hungary, Romania, the Balkan peninsula, Moldova, Ukraine, southern Russia and Turkey to Central Asia (Rasmont et al., 2015). This species is listed as “Endangered (EN)” (Nieto et al., 2014). Photo by JS Ascher. Middle right: B. (Thoracobombus) zonatus Smith, 1854 is a species that occurs in the Balkans, Romania, Moldova, Ukraine, southern Russia, Turkey, Caucasian countries and Iran (Rasmont et al., 2015). This species is listed as “Endangered (EN)” (Nieto et al., 2014). Photo by G Holmström. Bottom left: B. (Alpinobombus) hyperboreus Schönherr, 1809, a socially parasitic species of B. polaris, B. jonellus, and probably other Bombus species found in the Scandinavian mountains and along the northern tundra (Rasmont et al., 2015). This arctic species is listed as “Vulnerable (VU)” (Nieto et al., 2014). Photo by G Holmström. Bottom right: B. (Alpinobombus) alpinus (L., 1758), a species restricted to high altitude habitats of the Alps, the Carpathian and Scandinavian mountains, as well as to Arctic tundra of northern Fennoscandia (Rasmont et al., 2015). This species is listed as “Vulnerable (VU)” (Nieto et al., 2014). Photo by HB Jacobi.

26

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

found no significant decrease in PD after removing the 4 species from the EN category compared to a null community (PD's SES = 0.5428, associated p-value = 0.6924) or after removing the 12 species from the EN and VU categories (PD's SES = − 0.60218, associated p-value = 0.2344). However, when the 15 species in the EN, VU and NT categories were removed from the original 58 species assemblage, I detected a significant lower PD than what would be observed in a null community (PD's SES = −1.9179, associated p-value = 0.0209) (Table 1, Fig. 3). These results clearly illustrate that although a certain degree of PD can persist despite the loss of a small number of the most threatened species in the EN and VU categories, the loss of more species in the NT category leads to a significant “tipping point” outcome (sensu Faith et al. (2010)), with a significantly disproportionate loss of evolutionary history and resilience (Fig. 3). This is arguably because (i) the most threatened species in the EN and VU categories are phylogenetically overdispersed (see also Thuiller et al. (2011) on European birds, mammals and plants threatened by climate change; contra (Purvis et al., 2000)), and because (ii) some threatened species of EU bumblebees like B. mendax are more evolutionarily unique on their tree of life — their extinction would therefore imply a comparatively greater loss of phylogenetic diversity than would be expected by chance alone, resulting in the pattern illustrated on Fig. 3. The examination of the

phylogenetic distribution of threatened EU bumblebee species reveals that entire monophyletic sub-clades in the phylogeny (here, subgenera) are unlikely to become extinct since threatened species often have close relatives that are not threatened (Fig. 2). For example, the loss of B. fragrans and B. distinguendus would not lead to a decline of the subgenus Subterraneobombus since B. subterraneus is listed as NT in the IUCN Red List of EU bumblebees. Likewise, only in the very unlikely event of a decline of B. balteatus and B. mesomelas — two species currently listed as LC — across their whole distribution range, the phylogenetic diversity represented by species in the subgenus Alpinobombus and (part of) the subgenus Thoracobombus is unlikely to be lost as a whole. Nevertheless, this analysis reveals that besides the threatened species on the IUCN Red List of EU bumblebees, other Bombus species that have not (yet) made their way to the Red List might be important to safeguard the overall phylogenetic diversity of a series of subgenera or species groups (Table 2). 3.3. Are geographically-restricted bumblebee species more prone to extinction? It is theoretically assumed that narrow range size can potentially increase the vulnerability of bumblebee species, my results support the

Fig. 2. A near-complete species level phylogeny of EU bumblebee species in “fan” format adapted from Hines (2008). Species marked with black circles are threatened by extinction at the EU scale, i.e. they are listed as either “Critically Endangered (CR)”, “Endangered (EN)” or “Vulnerable (VU)” in the IUCN Red List of European Bees (Nieto et al., 2014). All other species are reported to be either “Near Threatened (NT)”, “Least Concern (LC)” or “Data Deficient (DD)”.

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

Fig. 3. Dendrogram-based extinction curve of Faith's (1992) phylogenetic diversity (PD) illustrating the degree of PD loss associated to the extinction of EU bumblebee species when moving from left to right on the species loss axis. The dotted grey curve represents the mean (±S.D.) values of randomized PD, while the colour-coded dotted curve represents the values of observed PD (Red = “Endangered (EN)” species, Orange = “Vulnerable (VU)” species, Yellow = “Near Threatened (NT)” species, Green = “Least Concern (LC)” and “Data Deficient (DD)” species). The results illustrate that the extinction of EN-VU-NT species would result in a more significantly disproportionate PD loss than would be expected by chance alone in a null community (PD's SES = −1.9179, associated p-value = 0.0209). See Materials and methods section for details. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

fact that the distribution of range size on the EU bumblebee phylogeny is equivalent to that expected by Brownian motion of trait evolution (Blomberg's K = 0.0372, p-value = 0.972), i.e. range size values drift randomly over evolutionary time and the phylogenetic relationships of taxa perfectly predict the covariance among taxa for range size

27

values. Most importantly, my results also show that phylogenetic relatedness explains none of the range size similarity among bumblebee species (Abouheif's Cmean = − 0.1322, p-value = 0.9421; Pagel's λ: lambda = 6.7287e-05, log-likelihood = − 433.203, p-value = 1). These results are in line with other studies reporting a lack of range size evolutionary heritability (Gaston, 1998; Blackburn et al., 2004; Waldron, 2007), but they are also in marked contrast with other studies where clear, broad-scale evolutionary patterns of range size have been found (e.g. Jablonski (1987), Webb and Gaston (2003), Webb and Gaston (2005), Hunt et al. (2005), Jones et al. (2005), Dexter and Chave (2016) on Amazonian tree genera). These mixed results in the contemporary literature can be explained by the fact that range size inheritance (or a lack thereof) does not follow the same principles as the evolution of other quantitative traits like body size: indeed, the speciation mode (e.g. allopatric, parapatric, sympatric or vicariant speciation) can theoretically result in any pattern range size divergence or convergence over evolutionary time, through a complex interplay of ecological, physiological or abiotic factors, which challenges any attempt to generalize and/or transpose the observed phylogenetic distribution of range size from one group of organisms to the next. To determine whether or not threatened species of the IUCN Red List (“CR”, “EN”, and “VU” categories) had similar mean EU range size compared to non-threatened taxa (“NT”, “LC” and “DD” categories) I used a Phylogenetic Generalized Least Squares (PGLS) approach. Although Red Listed Bombus species had on average a significantly smaller range size compared to non-threatened species (Kruskal-Wallis rank sum test: chi-squared = 7.4299, df = 1, p-value = 0.006), the relationship became non-significant when considering the lack of evolutionary independence of the bumblebee species investigated (PGLS: AIC = 942.43, BIC = 948.61, log-likelihood = −468.22; p-value = 0.0928). Similar results have been found in recent study by Schachat et al. (2016) who found that smaller range size was not significantly associated with any single IUCN threat ranking in Incilius toads. By contrast, it has been

Fig. 4. Character mapping and ancestral state reconstruction of the EDGE (Evolutionary Distinct, Globally Endangered) (left) and EDR (Evolutionary Distinct Rarities) (right) metrics on a near-complete species level phylogeny of EU bumblebee species adapted from Hines (2008). The EDGE metric combines the analysis of a species' evolutionary distinctiveness (ED, its degree of isolation on the phylogenetic tree) and extinction risk (weighted IUCN Red List categories) (Isaac et al., 2007), whereas the EDR metric combines the analysis of a species' evolutionary distinctiveness (ED) with its EU range size (geographic distribution computed as the number of 50 km square grids occupied by each species for the period 1970–2000), irrespective of its IUCN conservation status (Jetz et al., 2014). The EDGE and the EDR metrics help increase the efficacy of the existing prioritization for the conservation of EU bumblebees (i) by capturing the phylogenetic diversity and its associated functions (EDGE), as well as (ii) by better targeting species that are both evolutionarily unique (or nonredundant), threatened and restricted in their EU range size (EDR). See Materials and methods and Results and discussion sections for details.

28

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

shown elsewhere that IUCN threat ranking can sometimes be predicted by its phylogenetic position (Bielby et al. (2008) on frogs; Isaac et al. (2012) on amphibians). The results highlighted above have important implications for conservation. Indeed, since there seems to be an apparent mismatch between the threatened species on the IUCN Red List of EU bumblebee species and the list of species characterized by a small to very small range size, it can be hypothesized that more phylogenetic diversity is expected to be lost if anything happens to small range size species, as deeper phylogenetic branches than expected by chance are more likely to be lost (Purvis et al., 2000). This situation can be typically illustrated by B. mendax in this study — a species only listed as “Near Threatened (NT)” in the IUCN Red List of EU bumblebees characterized by a range size restricted to the high altitude habitats of the Cantabrian Mountains (where it is very rare), the Pyrenees and the Alps. 3.4. EU bumblebees on the EDGE (and the EDR metrics): setting conservation priorities To reach a better integration the EU bumblebees' evolutionary distinctiveness (ED) and extinction risk (weighted IUCN categories), I mapped the EDGE (Evolutionary Distinct, Globally Endangered) metric (Isaac et al., 2007) on the Bombus phylogeny tips shown in Fig. 2 before estimating the ancestral value at each node using maximum likelihood ancestral state reconstruction (Revell, 2012). The integrative effect of ED and IUCN category on the EDGE metric is apparent in Fig. 3 and has also helped ranking a top 12 of EU Bombus species deserving special conservation attention (Table 3). Collectively, these results indicate that the species in Table 3 encompass all of the most severely threatened species (EN & VU categories) in the IUCN Red List of EU bumblebees which can therefore be regarded as properly representing the evolutionary distinctiveness in EU bumblebees (contra e.g. Fenker et al. (2014) on Neotropical pitvipers; see also Davies (2015) on the South African Cape flora). Yet the results presented in Table 3 go one step further by also providing a rank of the top 12 EU bumblebees to target for conservation action by taking into account the need to preserve ED species (i.e., those lacking close relatives and that are associated with short phylogenetic branches), something lacking altogether in the original IUCN Red List of EU bumblebees assessment. According to my results, species with the highest EDGE scores listed like B. fragrans and B. armeniacus (two species have substantial populations outside of EU, see Williams et al. (2011)), as well as B. zonatus and B. brodmannicus are among those deserving the highest conservation attention (see in Table 3 and also Figs. 1 & 3). Despite the general lack of statistical support for range size evolutionary heritability described above, it is important to note that these species listed in the top #4 are all in the EN IUCN category and are also each characterized by unusually small range size (Table 3). The combined effects of ED and range size were then examined with the EDR metric which I also mapped on the Bombus phylogeny tips before estimating its ancestral value at each node using maximum likelihood ancestral state reconstruction (Revell, 2012). The rationale behind the EDR metric is that coupling ED to range size provides a unique way of reaching a better targeting of species that are both evolutionarily unique (or non-redundant) and restricted in their range size, irrespective of their status on the IUCN Red List of EU bumblebees. Here, the results shown in Table 3 and Fig. 4 pinpoint B. patagiatus among others as a species with a particularly high EDR score. Yet, this species has a geographic distribution that extends far into the Oriental and Palaearctic Regions, and in Europe it is strictly restricted to EFinland. It has therefore been categorised as “Data Deficient (DD)” in the IUCN Red List of European Bees, and its highlighting in Fig. 4 should be considered as an artefact. Conversely, it is interesting to note that B. brodmannicus stands out on the EDR phylogenetic mapping — this species is listed as “Endangered (EN)” in the IUCN Red List of European Bees because of its small range size and its disjunct

distribution between the Alps and the Caucasus (Rasmont et al., 2015). It should therefore be considered as an “evolutionarily distinct rarity” (or “EDR species”) and this status should help prioritize its conservation at the EU scale. Another EDR species of interest here and already mentioned above is B. mendax, which is not considered to be threatened by extinction, as it is listed as “Near Threatened (NT)” in the IUCN Red List of European Bees. Yet, its comparatively higher EDR score (see Table 3, Fig. 4) suggests that this species should receive particular conservation attention as it is both remarkably evolutionarily unique and highly restricted in its range size limited to high alpine and subalpine areas as mentioned above. The goal of the research presented in this paper was to have an explicit focus on the EU bumblebee fauna in order to make recommendations that would be valid within the particular biogeographical and political context of the EU. A worldwide analysis using all bumblebee species based on a complete phylogeny and a thorough evaluation of each species with the IUCN criteria would be both desirable and relevant to see if the patterns shown in this study on the EU species are consistent with the results from such a worldwide analysis. 4. Concluding remarks and conservation recommendations The incorporation of phylogenetic approaches in conservation has yielded a significant body of literature in recent years (Faith, 1662; Rolland et al., 2011), including in the context of IUCN threat rankings. This research has also highlighted that the IUCN criteria will not systematically preserve the highest levels of PD or put a higher emphasis on species characterized by small to very small range size. By combining the EDGE and EDR metrics in a robust molecular phylogenetic context, I hope I was able to demonstrate that a substantial improvement of the conservation prioritization can be achieved to better preserve our evolutionary heritage of the EU bumblebee fauna. This is essential as we are experiencing a state of shift in the planet's environment driven primarily by human activities within the “Anthropocene” (Barnosky et al., 2012). Following these analyses and based on my results, I suggest the following recommendations to help prioritize conservation actions of bumblebees in particular, but also of bees in general, by developing a more phylogenetically-informed conservation scheme. Recommendation #1. In order to ensure an optimal allocation of resources for conservation actions targeting threatened EU bumblebees, the use of species distribution models (SDMs) for the identification of key biodiversity areas and important evolutionary refugia or regions with higher phylogenetic endemism is an essential next step and the way forward (Souto et al., 2014; Brooks et al., 2015). This, in turn, will have to be done by carefully keeping a balance between more conventional, species-based but nevertheless successful conservation approaches like the IUCN Red List (Rodrigues et al., 2006), and the functional- or phylogenetically-informed alternatives to conservation prioritization that are increasingly developed in the modern literature (Faith, 1662). Recommendation #2. I support Faith et al.'s (2004) view that any “unique PD contribution”, here observed for B. mendax (NT on the IUCN Red List), should be incorporated in conservation planning, either by upscaling its IUCN threat rank accordingly, or by taking this assessment into account for the identification of key biodiversity areas. Indeed, this species listed here both as an “EDGE and EDR species” (Table 3, Fig. 4) is highly evolutionarily unique, and its extinction would imply a comparatively greater loss of phylogenetic diversity, leaving no close relative with a more recent common ancestor in the EU, although other close relatives are obviously present in Turkey, the Caucasus, and further east in Asia. Other species like B. brodmannicus (EN on the IUCN Red List), also highlighted as both an “EDGE and EDR species” (Table 3, Fig. 4), should attract as much conservation attention as B. fragrans (top #1 EDGE species listed in Table 3) owing to its small EU range size and evolutionary distinctiveness.

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

Recommendation #3. With the recent publication of the IUCN Red List of EU bees, it came as a shock that more than 50% of the EU wild bee species (estimated to ca. 2000 species) were evaluated as “Data Deficient (DD)”. This figure is the direct outcome of a dramatic shortage of biogeographical data collected consistently across countries, years or even decades. Not surprisingly, and as highlighted in the IUCN Red List of EU bees (Nieto et al., 2014), this is in turn the result of a dramatic lack of investment from public and private institutions and stakeholders alike into the science of taxonomy and other collections-based research with an integrative perspective on specimen identification, taxonomy, nomenclature, systematics, ecology, and biogeography. If tomorrow's field-based conservationists are to develop action plans across regional and national borders to avoid worst-case losses of long branches from the bee tree of life and their associated ecological/economic consequences, considerable efforts should be made by universities, museums, institutes and the private sector alike to financially support fundamental research. The short- and mid-term goals should be (i) to systematically revise all groups of wild bees with morphological and molecular methods, (ii) to encourage field surveys with standardized protocols across habitats and regions, and (iii) to develop scientific outreach and education programmes that would encourage the general public and amateur naturalists to be involved in surveying our (partly) vanishing fauna of wild bees. The computation of PD and alternative biodiversity metrics and their analysis in a larger interdisciplinary framework require high-quality biogeographical and ecological data, along with molecular phylogenies based on multiple gene and traits calibrated with a molecular clock. Only this way will we reach a better measure of the contributions of each species to the multiple facets of wild bee biodiversity, and only this way will we be able to develop more fine-grained species distribution models (SDMs) and more scientifically-sound conservation action programmes. Acknowledgements I am grateful to G Hollström and HB Jacobi for sharing their superb photos of rare bumblebee species used for the Fig. 1 of this article. I also thank L Revell for his kind help with the use of the phytools package, as well as A Dorchin, PH Williams and three anonymous referees for their comments on an earlier version of this manuscript. References Abouheif, E., 1999. A method for testing the assumption of phylogenetic independence in comparative data. Evol. Ecol. Res. 1, 895–909. Albrecht, M., Schmid, B., Hautier, Y., Müller, C.B., 2012. Diverse pollinator communities enhance plant reproductive success. Proc. R. Soc. B 279, 4845–4852. Alford, D.V., 1973. Bumblebee Distribution Maps Scheme Guide to the British Species. Bee Research Association, London UK (48 pp. London). Ascher, J.S., Pickering, A., 2016. Discover Life bee species guide and world checklist (Hymenoptera: Apoidea: Anthophila) (draft 45 of 2016). http://www.discoverlife.org/ mp/20q?search=Apoidea. Barnosky, A.D., Hadly, E.A., Bascompte, J., et al., 2012. Approaching a state-shift in the biosphere. Nature 486, 52–56. Bartomeus, I., Ascher, J.S., Gibbs, J., Danforth, B.N., Wagner, D.L., Hedtke, S.M., Winfree, R., 2013. Historical changes in northeastern US bee pollinators related to shared ecological traits. Proc. Natl. Acad. Sci. U. S. A. 110 (12), 4656–4660. Bielby, J., Cooper, N., Cunningham, A.A., Garner, T.W.J., Purvis, A., 2008. Predicting susceptibility to future declines in the world's frogs. Conserv. Lett. 1, 82–90. Blackburn, T.M., Jones, K.E., Cassey, P., Losin, N., 2004. The influence of spatial resolution on macroecological patterns of range size variation: a case study using parrots (Aves: Psittaciformes) of the world. J. Biogeogr. 31, 285–293. Brooks, T.M., Cuttelod, A., Faith, D.P., Garcia-Moreno, J., Langhammer, P., Pérez-Espona, S., 2015. Why and how might genetic and phylogenetic diversity be reflected in the identification of key biodiversity areas? Phil. Trans. R. Soc. B 370, 20140019. Cadotte, M.W., Davies, T.J., 2010. Rarest of the rare: advances in combining evolutionary distinctiveness and scarcity to inform conservation at biogeographical scales. Divers. Distrib. 16, 376–385. Cadotte, M.W., Cardinale, B.J., Oakley, T.H., 2008. Evolutionary history and the effect of biodiversity on plant productivity. Proc. Natl. Acad. Sci. U. S. A. 105, 17012–17017. Cadotte, M.W., Cavender-Bares, J., Tilman, D., Oakley, T.H., 2009. Using phylogenetic, functional and trait diversity to understand patterns of plant community productivity. PLoS One 4, e5695. Cadotte, M.W., Dinnage, R., Tilman, D., 2012. Phylogenetic diversity promotes ecosystem stability. Ecology 93, 223–233.

29

Cameron, S.A., Hines, H.M., Williams, P.H., 2007. A comprehensive phylogeny of the bumble bees (Bombus). Biol. J. Linn. Soc. 91, 161–188. Cameron, S.A., Lozier, J.D., Strange, J.P., Koch, J.B., Cordes, N., Solter, L.F., Griswold, T.L., 2011. Patterns of widespread decline in North American bumble bees. Proc. Natl. Acad. Sci. U. S. A. 108 (2), 662–667. Cardillo, M., Mace, G.M., Gittleman, J.L., Jones, K.E., Bielby, J., Purvis, A., 2008. The predictability of extinction: biological and external correlates of decline in mammals. Proc. R. Soc. Lond. B 275 (1641), 1441–1448. Carpenter, K.E., et al., 2008. One-third of reef-building corals face elevated extinction risk from climate change and local impacts. Science 321, 560–563. Cooper, N., Rodríguez, J., Purvis, A., 2008. A common tendency for phylogenetic overdispersion in mammalian assemblages. Proc. R. Soc. B 275, 2031–2037. Core Team, R., 2015. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Darwin, C., 1859. On the Origin of Species by Means of Natural Selection, or, the Preservation of Favoured Races in the Struggle for Life. J. Murray, London. Davies, T.J., 2015. Losing history: how extinctions prune features from the tree of life. Phil. Trans. R. Soc. B 370, 20140006. Dexter, K., Chave, J., 2016. Evolutionary patterns of range size, abundance and species richness in Amazonian angiosperm trees. PeerJ 4, e2402. Faith, D.P., 1992. Conservation evaluation and phylogenetic diversity. Biol. Conserv. 61 (1), 1–10. Faith, D.P., 2015. Phylogenetic diversity, functional trait diversity and extinction: avoiding tipping points and worst-case losses. Philos. Trans. R. Soc. Lond. B Biol. Sci. 370 (1662), 20140011. Faith, D.P., Magallón, S., Hendry, A.P., Conti, E., Yahara, T., Donoghue, M.J., 2010. Evosystem services: an evolutionary perspective on the links between biodiversity and human well-being. Curr. Opin. Environ. Sustain. 2, 66–74. Faith, D.P., Reid, C.A.M., Hunter, J., 2004. Integrating phylogenetic diversity, complementarity, and endemism for conservation assessment. Conserv. Biol. 18, 255–261. Felsenstein, J., 1985. Phylogenies and the comparative method. Am. Nat. 125, 1–15. Fenker, J., Tedeschi, L.G., Pyron, R.A., de C. Nogueira, C., 2014. Phylogenetic diversity, habitat loss and conservation in South American pitvipers (Crotalinae: Bothrops and Bothrocophias). Divers.0 Distrib. 20 (10), 1108–1119. Forest, F., Crandall, K.A., Chase, M.W., Faith, D.P., 2015. Phylogeny, extinction and conservation: embracing uncertainties in a time of urgency. Phil. Trans. R. Soc. B 370, 20140002. Freckleton, R.P., Harvey, P.H., Pagel, M., 2002. Phylogenetic analysis and comparative data: a test and review of evidence. Am. Nat. 160, 712–726. Fritz, S.A., Purvis, A., 2010. Selectivity in mammalian extinction risk and threat types: a new measure of phylogenetic signal strength in binary traits. Conserv. Biol. 24, 1042–1051. Fründ, J., Dormann, C.F., Holzschuh, A., Tscharntke, T., 2013. Bee diversity effects on pollination depend on functional complementarity and niche shifts. Ecology 94 (9), 2042–2054. Gaston, K.J., 1998. Species–range size distributions: products of speciation, extinction and transformation. Phil. Trans. R. Soc. Lond. B 353, 219–230. González-Varo, J.P., Biesmeijer, J.C., Bommarco, R., Potts, S.G., Schweiger, O., Smith, H.G., Steffan-Dewenter, I., Szentgyörgyi, H., Woyciechowski, M., Vilá, M., 2013. Combined effects of global change pressures on animal-mediated pollination. Trends Ecol. Evol. 28, 524–530. Goulson, D., 2003. Bumblebees: Their Behaviour and Ecology. Oxford University Press, Oxford. Grafen, A., 1989. The phylogenetic regression. Phil. Trans. R. Soc. B 326, 119–157. Gravel, D., Bell, T., Barbera, C., Bouvier, T., Pommier, T., Venail, P., Mouquet, N., 2011. Experimental niche evolution alters the strength of the diversity–productivity relationship. Nature 469, 89–92. Hansen, T.F., Martins, E.P., 1996. Translating between microevolutionary process and macroevolutionary patterns: the correlation structure of interspecific data. Evolution 50, 1404–1417. Harris, G., Pimm, S.L., 2008. Range size and extinction risk in forest birds. Conserv. Biol. 22 (1), 163–171. Harvey, P.H., Pagel, M.D., 1991. The Comparative Method in Evolutionary Biology. Oxford University Press Inc, New York, NY, USA. Hidasi-Neto, J., Loyola, R.D., Cianciaruso, M.V., 2013. Conservation actions based on red lists do not capture the functional and phylogenetic diversity of birds in Brazil. PLoS One 8 (9), e73431. Hines, H., 2008. Historical biogeography, divergence times, and diversification patterns of bumble bees (Hymenoptera: Apidae: Bombus). Syst. Biol. 57 (1), 58–75. Hoehn, P., Tscharntke, T., Tylianakis, J.M., Steffan-Dewenter, I., 2008. Functional group diversity of bee pollinators increases crop yield. Proc. R. Soc. B 275, 2283–2291. Huang, D., Roy, K., 2015. The future of evolutionary diversity in reef corals. Phil. Trans. R. Soc. B 370, 20140010. Hunt, G., Roy, K., Jablonski, D., 2005. Species-level heritability reaffirmed: a comment on “On the heritability of geographic range sizes.”. Am. Nat. 166, 129–135. Isaac, N.J.B., Redding, D.W., Meredith, H.M., Safi, K., 2012. Phylogenetically-informed priorities for amphibian conservation. PLoS One 7 (8), e43912. Isaac, N.J., Turvey, S.T., Collen, B., Waterman, C., Baillie, J.E., 2007. Mammals on the EDGE: conservation priorities based on threat and phylogeny. PLoS One 2 (3), e296. IUCN, 2016. The IUCN red list of threatened species. http://www.iucnredlist.org/ technical-documents/red-list-documents (visited June 15, 2016). Jablonski, D., 1987. Heritability at the species level: analysis of geographic ranges of Cretaceous mollusks. Science 238, 360–363. Jetz, W., Thomas, G., Joy, J., Redding, D., Hartmann, K., Mooers, A., 2014. Global distribution and conservation of evolutionary distinctness in birds. Curr. Biol. 24, 1–12. Jombart, T., Dray, S., 2008. adephylo: exploratory analyses for the phylogenetic comparative method. Bioinformatics 26, 1907–1909. Jones, K.E., Sechrest, W., Gittleman, J.L., 2005. Age and area revisited: identifying global patterns and implications for conservation. In: Purvis, A., Gittleman, J.L., Brooks, T.M. (Eds.), Phylogeny and Conservation. Cambridge University Press, Cambridge, pp. 141–165. Keitt, T., 2012. colorRamps: builds color tables. R package version 2.3. https://CRAN.Rproject.org/package=colorRamps.

30

N.J. Vereecken / Biological Conservation 206 (2017) 21–30

Kembel, S.W., Cowan, P.D., Helmus, M.R., Cornwell, W.K., Morlon, H., Ackerly, D.D., Blomberg, S.P., Webb, C.O., 2010. Picante: R tools for integrating phylogenies and ecology. Bioinformatics 26, 1463–1464. Kerr, J.T., Pindar, A., Galpern, P., Packer, L., Potts, S.G., Roberts, S.P.M., Rasmont, P., Schweiger, O., Colla, S.R., Richardson, L.L., Wagner, D.L., Gall, L.F., Sikes, D.S., Pantoja, A., 2015. Climate change impacts on bumblebees converge across continents. Science 349 (6244), 177–180. Kleijn, D., Winfree, R., Bartomeus, I., Carvalheiro, L.G., Henry, M., Isaacs, R., et al., 2015. Delivery of crop pollination services is an insufficient argument for wild pollinator conservation. Nat. Commun. 6, 7414. Mace, G.M., Baillie, J.E.M., 2007. The 2010 biodiversity indicators: challenges for science and policy. Conserv. Biol. 21, 1406–1413. Mace, G.M., Purvis, A., 2008. Evolutionary biology and practical conservation: bridging a widening gap. Mol. Ecol. 17, 9–19. Mace, G.M., Gittleman, J.L., Purvis, A., 2003. Preserving the tree of life. Science 300, 1707–1709. Martins, K.T., Gonzales, A., Lechowicz, M.J., 2015. Pollination services are mediated by bee functional diversity and landscape context. Agric. Ecosyst. Environ. 200, 12–20. McNeely, J.A., 1988. Economics and Biological Diversity: Developing and Using Economic Incentives To Conserve Biological Resources. IUCN, Gland, Switzerland. Michener, C.D., 2007. The Bees of the World. second ed. Johns Hopkins University Press. Münkemüller, T., Lavergne, S., Bzeznik, B., Dray, S., Jombart, T., Schiffers, K., Thuiller, W., 2012. How to measure and test phylogenetic signal. Methods Ecol. Evol. 3, 743–756. Nieto, A., Roberts, S.P.M., Kemp, J., Rasmont, P., Kuhlmann, M., García Criado, M., Biesmeijer, J.C., Bogusch, P., Dathe, H.H., De la Rúa, P., De Meulemeester, T., Dehon, M., Dewulf, A., Ortiz-Sánchez, F.J., Lhomme, P., Pauly, A., Potts, S.G., Praz, C., Quaranta, M., Radchenko, V.G., Scheuchl, E., Smit, J., Straka, J., Terzo, M., Tomozii, B., Window, J., Michez, D., 2014. European Red List of Bees. Publication Office of the European Union, Luxembourg. Nunes, L.A., Turvey, S.T., Rosindell, J., 2015. The price of conserving avian phylogenetic diversity: a global prioritization approach. Phil. Trans. R. Soc. B 370, 20140004. Ollerton, J., Winfree, R., Tarrant, S., 2011. How many flowering plants are pollinated by animals? Oikos 120 (3), 321–326. Orme, D., Freckleton, R., Thomas, G., Petzoldt, T., Fritz, S., Isaac, N., Pearse, W., 2013. caper: comparative analyses of phylogenetics and evolution in R. R package version 0.5.2. https://CRAN.R-project.org/package=caper. Pagel, M., 1999. Inferring the historical patterns of biological evolution. Nature 401, 877–884. Paradis, E., Claude, J., Strimmer, K., 2004. APE: analyses of phylogenetics and evolution in R language. Bioinformatics 20, 289–290. Peters, G., 1972. Ursachen für den Rückgang der seltenen heimishchen Hummelarten (Hym., Bombus et Psithyrus). Entomol. Ber. 1972, 85–90. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., R Core Team, 2015. nlme: linear and nonlinear mixed effects models. R package version 3.1-121. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., R Core Team, 2016. nlme: linear and nonlinear mixed effects models. http://CRAN.R-project.org/package=nlme. Pollock, L.J., Rosauer, D.F., Thornhill, A.H., Kujala, H., Crisp, M.D., Miller, J.T., McCarthy, M.A., 2015. Phylogenetic diversity meets conservation policy: small areas are key to preserving eucalypt lineages. Phil. Trans. R. Soc. B 370, 20140007. Potts, S.G., Imperatriz-Fonseca, V.L., Ngo, H.T., Biesmeijer, J.C., Breeze, T.D., Dicks, L.V., Garibaldi, L.A., Hill, R., Settele, J., Vanbergen, A.J., Aizen, M.A., Cunningham, S.A., Eardley, C., Freitas, B.M., Gallai, N., Kevan, P.G., Kovács-Hostyánszki, A., Kwapong, P.K., Li, J., Li, X., Martins, D.G., Nates-Parra, G., Pettis, J.S., Rader, R., Viana, B.F., 2016. IPBES, 2016. Summary for Policymakers of the Assessment Report of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services on Pollinators, Pollination and Food Production. Secretariat of the Intergovernmental SciencePolicy Platform on Biodiversity and Ecosystem Services, Bonn, Germany (36 pages). Purvis, A., Agapow, P.M., Gittleman, J.L., Mace, G.M., 2000. Nonrandom extinction and the loss of evolutionary history. Science 288, 328–330. Rasmont, P., Mersch, P., 1988. Première estimation de la dérive faunique chez les Bourdons de la Belgique (Hymenoptera, Apidae). Ann. Soc. R. Zool. Belg. 118 (2), 141–147. Rasmont, P., Franzén, M., Lecocq, T., Harpke, A., Roberts, S.P.M., Biesmeijer, J.C., Castro, L., Cederberg, B., Dvorák, L., Fitzpatrick, Ú., Gonseth, Y., Haubruge, E., Mahé, G., Manino, A., Michez, D., Neumayer, J., Ødegaard, F., Paukkunen, J., Pawlikowski, T., Potts, S.G., Reemer, M., Settele, J., Straka, J., Schweiger, O., 2015. Climatic Risk and Distribution Atlas of European Bumblebees. Biorisk 10 (Special Issue, 246 pp).

Redding, D.W., Mooers, A.O., 2006. Incorporating evolutionary measures into conservation prioritisation. Conserv. Biol. 20, 1670–1678. Redding, D.W., DeWolff, C.V., Mooers, A.O., 2010. Evolutionary distinctiveness, threat status and ecological oddity in primates. Conserv. Biol. 24, 1052–1058. Revell, L.J., 2012. phytools: an R package for phylogenetic comparative biology (and other things). Methods Ecol. Evol. 3, 217–223. Revell, L.J., Harmon, L.J., Collar, D.C., 2008. Phylogenetic signal, evolutionary process, and rate. Syst. Biol. 57 (4), 591–601. Rodrigues, A.S.L., Pilgrim, J.D., Lamoreux, J.F., Hoffmann, M., Brooks, T.M., 2006. The value of the IUCN Red List for conservation. Trends Ecol. Evol. 21 (2), 71–76. Rolland, J., Cadotte, M.W., Davies, J., Devictor, V., Lavergne, S., Pavoine, S., Rodrigues, A., Thuiller, W., Turcati, L., Winter, M., Zupan, L., Jabot, F., Morlon, H., 2011. Using phylogenies in conservation: new perspectives. Biol. Lett. 8, 692–694. Runge, C.A., Tulloch, A., Hammill, E., Possingham, H.P., Fuller, R.A., 2015. Geographic range size and extinction risk assessment in nomadic species. Conserv. Biol. 29, 865–876. Schachat, S.R., Mulcahy, D.G., Mendelson, J.R., 2016. Conservation threats and the phylogenetic utility of IUCN Red List rankings in Incilius toads. Conserv. Biol. 30 (1), 72–81. Souto, C.P., Mathiasen, P., Acosta, M.C., Quiroga, M.P., Vidal-Russell, R., Echeverría, C., Premoli, A.C., 2014. Identifying genetic hotspots by mapping molecular diversity of widespread trees: when commonness matters. J. Hered. 106, 537–545. Sullivan, G.M., Feinn, R., 2012. Using effect size - or why p value is not enough. J. Grad. Med. Educ. 4, 279–282. Swenson, N.G., 2009. Phylogenetic resolution and quantifying the phylogenetic diversity and dispersion of communities. PLoS One 4 (2), e4390. Thorp, R.W., Shepherd, M.D., 2005. Profile: subgenus Bombus. In: Shepherd, M.D., Vaughan, D.M., Black, S.H. (Eds.), Red List of Pollinator Insects of North America. The Xerces Society for Invertebrate Conservation, Portland, Oregon CD-ROM Version 1. Thuiller, W., Lavergne, S., Roquet, C., Boulangeat, I., Lafourcade, B., Araujo, M.B., 2011. Consequences of climate change on the tree of life in Europe. Nature 470, 531–534. Tilman, D., Knops, J., Wedin, D., Reich, P., Ritchie, M., Siemann, E., 1997a. The influence of functional diversity and composition on ecosystem processes. Science 277, 1300–1302. Tilman, D., Lehman, C., Thomson, K.T., 1997b. Plant diversity and ecosystem productivity: theoretical considerations. Proc. Natl. Acad. Sci. U. S. A. 94, 1857–1861. Waldron, A., 2007. Null models of geographic range size evolution reaffirm its heritability. Am. Nat. 170, 221–231. Webb, T.J., Gaston, K.J., 2003. On the heritability of geographic range sizes. Am. Nat. 161, 553–566. Webb, T.J., Gaston, K.J., 2005. Heritability of geographic range sizes revisited: a reply to Hunt et al. Am. Nat. 166, 136–143. Webb, C.O., Ackerly, D.D., McPeek, M.A., Donoghue, M.J., 2002. Phylogenies and community ecology. Annu. Rev. Ecol. Syst. 33, 475–505. Wiens, J.J., Ackerly, D.D., Allen, A.P., Anacker, B.L., Buckley, L.B., Cornell, H.V., Damschen, E.I., Davies, J.T., Grytnes, J.A., Harrison, S.P., Hawkins, B.A., Holt, R.D., McCain, C.M., Stephens, P.R., 2010. Niche conservatism as an emerging principle in ecology and conservation biology. Ecol. Lett. 13 (10), 1310–1324. Williams, P.H., 1982. The distribution and decline of British bumble bees (Bombus Latr.). J. Apic. Res. 21, 236–245. Williams, P.H., 1986. Environmental change and the distributions of British bumble bees (Bombus Latr.). Bee World 67, 50–61. Williams, P.H., 1998. An annotated checklist of bumble bees with an analysis of patterns of description (Hymenoptera: Apidae, Bombini). Bull. Nat. Hist. Mus. (Ent.) 67, 79–152. Williams, P.H., Osborne, J.L., 2009. Bumblebee vulnerability and conservation world-wide. Apidologie 40, 367–387. Williams, P.H., Brown, M.J.F., Carolan, J.C., An, J.-D., Goulson, D., Aytekin, A.M., Best, L.R., Byvaltsev, A.M., Cederberg, B., Dawson, R., Huang, J.-X., Ito, M., Monfared, A., Raina, R.H., Schmid-Hempel, P., Sheffield, C.S., Sima, P., Xie, Z.-H., 2012a. Unveiling cryptic species of the bumblebee subgenus Bombus s. str. world-wide with COI barcodes (Hymenoptera: Apidae). Syst. Biodivers. 10, 21–56. Williams, P.H., Byvaltsev, A.M., Sheffield, C.S., Rasmont, P., 2012b. Bombus cullumanus – an extinct European bumblebee species? Apidologie 44, 121–132. Williams, P.H., An, J.-D., Huang, J.-X., 2011. The bumblebees of the subgenus Subterraneobombus: integrating evidence from morphology and DNA barcodes (Hymenoptera, Apidae, Bombus). Zool. J. Linn. Soc. 163, 813–862. Winter, M., Devictor, V., Schweiger, O., 2013. Phylogenetic diversity and nature conservation: where are we? Trends Ecol. Evol. 28, 199–204.