Patterns of Species, Phylogenetic and Mimicry Diversity of Clearwing Butterflies

in the Neotropics

Nicolas Chazot, Keith R. Willmott, André V.L. Freitas, Donna Lisa de Silva, Roseli Pellens, and Marianne Elias

Abstract The Neotropical region comprises six of the major biodiversity hotspots of the planet, including the Andean foothills, which harbour the most diverse terrestrial ecosystems. It is also one of those most threatened by habitat destruction and climatic changes, which cause species extirpation and sometimes extinction, resulting in community disassembly and loss of interspecific interactions. The effects of community disassembly can be particularly strong in highly coevolved mutualistic species assemblages, such as Müllerian mimetic species. Conservation strategies should therefore aim at preserving not only evolutionary diversity, but also species interactions. Here we use mimetic ithomiine butterflies (Nymphalidae: Danainae, Ithomiini) as a model to identify areas of both evolutionary and ecological importance, and hence conservation significance. Ithomiine butterflies form a tribe of ca. 380 species that inhabit lowland and montane Neotropical forests. All species engage in Müllerian mimicry, and drive mimicry in other, distantly related, Lepidoptera. We analyse phylogenetic, distribution and mimicry data for three diverse ithomiine genera, Napeogenes, Ithomia and Oleria. We use different metrics to study geographical patterns of diversity. Patterns of species richness, phylogenetic diversity and mimicry diversity are highly congruent within genera but slightly different among genera. Mountainous regions contain the greatest taxonomic and mimetic diversity in ithomiines, with the Andean foothill region being the area of highest diversity, but other regions, such as Central America and the upper Amazon, are also important. Finally, a measure of vulnerability related to mimicry indicates that mutualistic interactions are not distributed evenly across space and genera. We argue that mutualistic interactions should be taken into account in conservation strategies.

Keywords Ithomiini • Müllerian mimicry • Phylogenetic diversity • Amazonia • Andes

Introduction

The Neotropical region extends from Mexico to northern Argentina, including the Amazon basin, the Andean cordillera and the Atlantic Forest. It is the most biologically diverse of the world's major biogeographic regions (Gaston and Hudson 1994; Myers et al. 2000; Hawkins et al. 2007). At least one million species of insects, 40,000 of plants, 3000 of fishes, 1294 of birds, 427 of mammals, 427 of amphibians, and 378 of reptiles inhabit the Amazonian basin (Da Silva et al. 2005a). It is also a region of high endemism, including 6 of the world's 25 biodiversity 'hotspots' (Myers et al. 2000).

Many areas of the Neotropics are under continual threat from deforestation. In 2004, ca. 2.7 million ha of forest were cleared in Amazonia alone (INPE). In Central America, only 1.7 % of the original dry forest remains, and most of this comprises small, isolated patches (Griscom and Ashton 2011). Similarly, the Atlantic forest has been reduced to 12 % of its original area, with astonishing rates of deforestation every year (SOS Mata Atlântica, INPE, ISA 1998; Ribeiro et al. 2009). In Amazonia, wood extraction, industrial logging, cattle pastures, banana plantations and more recently oil palm culture are the main causes of the ongoing deforestation. The Neotropics are also threatened by climatic changes, which are likely to be particularly serious in mountain habitats (e.g., Engler et al. 2009; Chen et al. 2011; Feeley et al. 2011). Habitat destruction and climatic changes may cause species extirpation, displacements and extinction (e.g., Loiselle et al. 2010), which may in turn result in community disassembly, with loss of interspecific interactions (Sheldon et al. 2011). The consequences of community disassembly can be particularly strong in highly coevolved mutualistic species assemblages, such as insect-pollinator networks, plant species engaged in facilitation, or Müllerian mimetic species (Chazot et al. 2014).

To preserve Neotropical biodiversity, given constraints of time and money, it is essential to identify priority areas for conservation (Williams et al. 1996). However, there are several problems in identifying such areas. Firstly, distribution data are not available for all taxa, so attention has focused on indicator taxa, which are expected to reliably indicate patterns of diversity in other, more poorly known groups (e.g., Howard et al. 1998; Lamoreux et al. 2006). Insects constitute at least 70 % of all terrestrial organisms (Samways 1994), and their outstanding evolutionary success in virtually all terrestrial habitats potentially makes them one of the most valuable study groups for understanding the distribution and origins of biodiversity, and for developing efficient means to conserve that biodiversity (Brown 1997). However, because of the diversity of insects and taxonomic difficulties in many groups, some authors have suggested that conservation research be focused on a few suitable taxa, such as butterflies (New 1993; Brown and Freitas 2000; Bonebrake et al. 2010; Basset et al. 2013). Butterflies can be used to monitor ecosystem health (e.g. Warren et al. 2001), reveal broadly applicable patterns of diversity and endemism, and effectively communicate complex scientific ideas to the public and generate popular support for conservation (Sparrow et al. 1994; Boggs et al. 2003).

One of the best studied diverse groups of Neotropical butterflies is the tribe Ithomiini (Nymphalidae: Danainae), an exclusively Neotropical group which currently includes about 380 species placed in 47 genera (Lamas 2004; Willmott and Lamas 2007). Ithomiines occur from Mexico to Argentina and are largely restricted to moist forest habitats from sea level up to 3000 m (Beccaloni 1997a). Among the attributes that make the group a potentially useful indicator of conservation priorities for other taxa are its diversity, its broad range of occupied elevations, its abundance in the field and collections, the broad variation in range size among species, and its good level of taxonomic knowledge.

Having selected a potentially suitable indicator taxon, the next issue is to decide what surrogate measure of biodiversity will be used (Williams et al. 1996). Species richness is the most commonly used measure, but it may not represent important aspects of the structure and composition of natural communities. A species richness measure considers all species as equal, ignoring their functional or phylogenetic relationships (e.g., Safi et al. 2011). As an alternative, measures of phylogenetic diversity evaluate species in terms of the amount of unique evolutionary history they represent. The loss of species with no close relatives represents the extinction of an entire lineage, resulting in a greater loss to biodiversity than the loss of a species that shares most of its evolutionary history with another (Mace et al. 2003; Mooers et al. 2005; Maclaurin and Sterelny 2008). During the last two decades several metrics have been developed to assess the phylogenetic diversity of clades and to evaluate and compare communities for conservation based on the phylogenetic diversity of the species they harbour (e.g., Vane-Wright et al. 1991; Faith 1992). Despite the difficulty of defining the most adequate metric (see Redding et al. 2008; Schweiger et al. 2008), and the circumstances where phylogenetics can be useful for conservation (e.g., Rodrigues et al. 2005; Hartmann and Andre 2013), two points emerged from these studies. The first is that conservation strategies based on phylogenetic measures capture more evolutionary diversity than strategies ignoring phylogeny (e.g., Hartmann and Steel 2007; Redding et al. 2008), and the second is that extinctions are not random in the tree of life, but rather are phylogenetically and functionally clumped (Purvis et al. 2000; Yessoufou et al. 2012). In the last few years several phylogenies have become available for the Ithomiini as a whole (Brower et al. 2014; Willmott and Freitas 2006), and also for some speciose clades inside this tribe (Mallarino et al. 2005; Elias et al. 2009; de-Silva et al. 2015). This phylogenetic information thus allows us to consider phylogenetic diversity in an assessment of conservation priorities for Neotropical insects.

In addition to evolutionary diversity, ecological interactions represent another important component of biodiversity that is rarely addressed in conservation prioritization. While difficult to characterize for many insects, ecological interactions among ithomiines have received an unusual amount of attention since these butterflies illustrate some of the most outstanding examples of mutualistic, Müllerian mimicry (Müller 1879). Ithomiines are considered, together with the Heliconiini, the central models in many Neotropical Lepidoptera mimicry rings (Brown and Benson 1974; Beccaloni 1997a). Chemical compounds acquired by adult feeding (Brown 1984, 1985) make ithomiines unpalatable to their predators, which learn to avoid the aposematic wing patterns exhibited by ithomiines. The wing colour patterns of co-existing species are under strong selection for convergence, thereby reducing the individual cost of educating predators (Müller 1879; Mallet 1999), and the resulting 'mimicry rings' contain multiple co-mimetic species which interact mutualistically (Fig. 1). Co-mimetic ithomiines tend to share habitats (Chazot et al. 2014). They also tend to share hostplants (Willmott and Mallet 2004), fly at similar heights above the ground (Beccaloni 1997b; Elias et al. 2008) and fly in similar forest microhabitats (DeVries et al. 1999; Elias et al. 2008; Hill 2010). These tightly-knit webs of interactions may thus be particularly sensitive to community disassembly caused by habitat or climate change (Sheldon et al. 2011), with the potential for cascading co-extinctions due to the loss of a few species whose presence facilitates the existence of other species. Because mutualistic interactions are particularly easy to characterize in ithomiines (species that share the same wing pattern interact mutualistically, species with different wing patterns do not), this group provides a unique opportunity to assess the importance of mutualistic interactions from a conservation perspective.

Studies combining phylogenetic and ecological or functional data to characterize biodiversity patterns and their association with environmental gradients (Devictor et al. 2010; Flynn et al. 2011; Duarte et al. 2012), as well as to test conservationfocused hypotheses (Faith 2008; Safi et al. 2011), are likely to better represent biodiversity and hopefully to guide conservation in a more precise way. Here, we use distribution, phylogenetic and mimicry data for three diverse ithomiine genera, Ithomia, Napeogenes and Oleria, to identify areas of maximal species, phylogenetic and mimicry diversity for each of these genera. We also identify areas of maximum and minimum vulnerability in terms of proportion of potential loss of species due to disruption of mimicry rings. With these three independent replicates we then explore whether different metrics show peaks in the same areas, and whether the different taxa show similar spatial patterns of diversity.