Ant diversity as a direct and indirect driver of pselaphine rove beetle (Coleoptera: Staphylinidae) functional diversity in tropical rainforests, Sabah, Malaysian Borneo

1 INTRODUCTION

Myrmecophily, a symbiotic association with ants, is an ancient feature of ant ecology (Parker, 2016b; Parker & Grimaldi, 2014) and has arisen in many arthropod groups (Kronauer & Pierce, 2011). However, there is an evolutionary bias toward beetles (Coleoptera), where myrmecophily is most common (Parker, 2016a). Staphylinidae (rove beetles) have the most myrmecophilous lineages, and within this family of beetles, the subfamily Pselaphinae (Parker, 2016a) has many ant-associated species.

The Pselaphines are a species-rich group of predatory leaf litter beetles with 9,854 described living species in 1,247 genera worldwide (Parker, 2016a). They are generally small (∼1–3 mm in body length) and are particularly diverse and abundant in the tropics (Chandler, 2001; Chung et al., 2000; Parker & Maruyama, 2013; Nomura & Mohamed, 2008; Sakchoowong et al., 2008). In Borneo, 123 species have been recorded (Nomura & Mohamed, 2008), but many more are yet to be described. Pselaphines deviate from the typical staphylinid body plan in possessing a more compact, sclerotized integument that reduces abdominal flexibility. Although a relatively invariant body plan would appear to restrict their movements through the substrate, pselaphines have radiated successfully in the leaf litter, particularly in ant-dominated tropical forest floors (Chandler, 2001; Parker, 2016b; Sakchoowong et al., 2008).

Pselaphines have varied levels of social parasitism ranging from facultative or opportunistic nest intruders, a casual association in which the beetle is not wholly reliant on ants, through to obligate myrmecophiles, which are entirely dependent on host colonies (Geiselhardt et al., 2007; Mynhardt, 2013; Park, 1942; Parker, 2016a; Parker & Grimaldi, 2014). Direct observations of pselaphine-ant associations have been recorded (Parker & Maruyama, 2013; Sugaya et al., 2004; Yin & Li, 2013, 2015), but are mostly limited to records of beetles inside nests, with little behavioral data. Nonetheless, insights into their ecological and behavioral interactions can be gained by examining adult morphology (Mynhardt, 2013; Parker, 2016a; Parker & Grimaldi, 2014).

Functional traits associated with myrmecophily include: (a) Trichomes, thick brush-like bunches of setae conducting glandular exudates attractive to the ants; thus initiating trophallaxis, a unique liquid-feeding behavior typically occurring between worker ants or between worker ants and larvae, which the beetles exploit (Cassil & Tschinkel, 1996; Mynhardt, 2013; Park, 1942; Parker & Grimaldi, 2014). (2) Fusion of antennal and abdominal segments, which strengthens the body for protection during handling by worker ants and increases the surface area for secretions to spread (Parker & Grimaldi, 2014). In extreme cases, obligate myrmecophiles (exhibited by the tribe Clavigerini) display deep indentations in the abdomen providing grasping notches for worker ants to pick up and carry the beetle to their nest (Leschen, 1991; Parker, 2016a). (3) Body size enlargement has also been linked to inquilinous lifestyle, although the explanations for this observed trend have not been tested (Parker & Maruyama, 2013).

Pselaphines are habitat specific, occupying moist leaf litter and woody debris microhabitats (Chandler, 2001; Nomura & Mohamed, 2008), and therefore responsive to environmental change. Their ecological and functional importance have made them suitable candidates for conservation studies particularly concerning the effect of rainforest disturbances on biodiversity (Chung et al., 2000; Edwards et al., 2014; Sakchoowong et al., 2008). Ants and beetles, among other taxa, have proven to be effective bioindicators of the response of other taxa to logging and conversion to oil palm in Southeast Asia (Chung et al., 2000; Edwards et al., 2014; Sakchoowong et al., 2008). Ecosystem disturbances greatly influence the abundance and distribution of functionally important species that contribute to maintaining ecosystem functioning and biogeochemical cycles (Hooper et al., 2005; Loreau et al., 2001). Here, we measure the functional traits of pselaphines, specifically the traits putatively associated with myrmecophily, to allow us to better understand (a) how pselaphines interact with individual ants and ant nest colonies and (b) how these traits can be applied as a measure of functional diversity (FD) across varying levels of ecosystem disturbances in tropical rainforests. We use existing ant abundance data, all taken from the same plots, and samples as our pselaphine data, to examine how far our measures of putative myrmecophily match data on ant abundance.

We predict that FD of pselaphine beetles will increase in the presence of high-ant abundance.

2 MATERIALS AND METHODS

The Soil Biodiversity Group, Life Sciences Department, Natural History Museum, London provided the specimens of leaf litter Pselaphinae beetles for analysis. All sampling was conducted between September and October, 2012 in Sabah, Malaysian Borneo. Specimens were on loan from the collections of University Malaysia, Sabah (UMS).

3 RESULTS

A total of 613 Pselaphinae beetles were sampled, across all three sites (Table 5), representing 12 tribes and 42 morphospecies (Table 1).

3.1 Beetle myrmecophile traits

In this study, 12 morphospecies belonging to six different tribes were assigned as probable myrmecophiles, from a total of 42 total morphospecies recorded (Figure 1-3). The following groups of probable myrmecophile were found. (a) Two morphospecies were recorded from the tribe Clavigerini (Figure 1a,b). (b) Five morphospecies were recorded form the tribe Cyathigerini (Figure 2). (3) Only one morphospecies was recorded form the tribe Arnyliini (Figure 1c). (4) Two morphospecies were recorded from the tribe Tmesiphorini (Figures 1d, 3a), one morphospecies was recorded from the tribe Tyrini (Figure 3c) and one morphospecies was recorded from the tribe Batrisini (Figure 3b).

3.2 Tribe Clavigerini

Beetles within the tribe Clavigerini are exclusively obligate myrmecophiles, exhibiting dramatic external morphological modifications for social parasitism inside ant colonies (Figure 4a,b). They are characterized by having a reduced number of antennomeres with a truncate antennal apex (Figure 4c) and are sparsely pubescent and lack any foveae. There is a distinct broad basal depression “tergal plate” in the abdomen bearing large tufts of trichomes (Figure 4d).

3.3 Tribe Cyathigerini

Morphospecies belonging to the tribe Cyathigerini (Figures 5a–f, 6a,b) exhibits a reduced number of antennomeres (seven antennomeres) and are characterized by unique apical antennal club structures. The antennal clubs of morphospecies 3 (Figure 5b), 20 (Figure 6a) and 22 (Figure 6b) are enlarged, highly pubescent block-shaped terminal segments. In contrast, the antennal clubs of morphospecies 6 (Figure 5e) and morphospecies 8 (Figure 5f) are large and round and broadly excavated on the inner surface. Interestingly, trichomes are present inside the antennal club of morphospecies 6 (Figure 5e) but are absent in morphospecies 8 (Figure 5f). Morphospecies in this tribe also exhibit a reduced number of fovea, lack the presence of trichomes (excluding morphospecies 6, as previously mentioned) and dense punctures across the entire dorsal surface.

3.4 Tribe Tmesiphorini

Both morphospecies 36 (Figure 6e) and 5 (Figure 7a) have a complete set of antennomeres (11 antennomeres), which successively increase in size toward the apical antennomere and are highly pubescent across the entire body including the antenna. Both morphospecies exhibit a reduced number of basal elytral foveae (bef), with only two sets of bef. Interestingly, morphospecies 36 lacks the presence of trichomes compared to morphospecies 5. Antennomeres 6–11 of morphospecies 5 have shallow cavities embedded with impressions and setae on the ventral side (Figure 7c).

3.8 Possible myrmecophile beetles exhibiting myrmecophile traits

Both morphospecies 37 (Figure 8c) and 13 (Figure 8b) have compact antenna, partial fusion of sternites and tergites, and a reduced number of foveae. Morphospecies 32 lacks trichomes but is highly pubescent and has two sets of bef (Figure 8a). Interestingly, antennomere 6 and 7 of morphospecies 32 create a convex V-shape along the antennae, an antennal structure that has not been recorded in any other morphospecies in this study.

3.9 Pselaphinae FD

Old growth rainforest sites Danum and Maliau recorded the highest mean ant abundances (1,193 and 911, respectively, per transect) compared with logged rainforest site SAFE (576). Pselaphinae beetle putative myrmecophile FD is significantly correlated with ant abundance across all three sites, at the plot level (t = 2.45, p = 0.0247). The presence of myrmecophile functional traits increases as ant abundance increases (Figure 9).

4 DISCUSSION

Primary rainforest sites support high abundances of leaf litter ants, which is correlated with a high abundance of probable myrmecophile pselaphine beetles. In accordance with our predictions, putative myrmecophile FD of pselaphine beetles increased as ant abundance increased.

The arrangement and presence or loss of foveae have been recognized as one of the most important morphological structures for taxonomic identification of Pselaphinae (Chandler, 2001; Park, 1942). The functional role of foveae remains unknown. Two probable foveal functions have been proposed by Chandler (2001): (a) thoracic foveae are sensory and (b) abdominal and head foveae provide structural rigidity. Moreover, the trend appears to be toward loss of foveae with basal Pselaphinae lineages exhibiting a more diverse set of foveal patterns across the entire body (Chandler, 2001). Although the functional role of foveae is unrelated to myrmecophily, there is a reduction in this trait with ant-associated Pselaphinae beetles, specifically with regards to the basal elytral foveae sets (Chandler, 2001; Parker & Maruyama, 2013). This is shown here, where nonmyrmecophile morphospecies belonging to the tribe Proterini retain a complete set of basal elytral foveae. Probable myrmecophile Pselaphinae morphospecies recorded in the tribes Tmesiphorini and Tyrini all exhibit a reduction of two sets of basal elytral foveae and tribes Arnyliini and Cyathigerini retain just one set of basal elytral foveae. Obligate myrmecophile Pselaphinae morphospecies 23 and 27 belonging to the tribe Clavigerini show a complete loss of basal elytral foveae.

The tribe Cyathigerini comprising five facultative myrmecophile morphospecies was dominant in the logged (SAFE) rainforest sites. The large excavated clubs are suggestive of morphological structures that may be associated with myrmecophily, in that they potentially provide structures for worker ants to hold onto while transporting the beetle into the ant's nest. There is no direct evidence that antennal clubs have a functional role in myrmecophily; however, pselaphine beetles in this tribe occasionally form a facultative association with ants (Sugaya et al., 2004). Moreover, the antennal clubs are known to be secondary sexual characters unique to the tribe (Chandler, 2001; Sugaya et al., 2004). Male Plagiophorus species in the tribe Cyathigerini recorded from Borneo have enlarged round antennal clubs (Nomura & Mohamed, 2008) exhibited by morphospecies 6 and 8. One female Plagiophorus species has been recorded from Borneo that has a block-like antennal club (Nomura & Mohamed, 2008) exhibited by morphospecies 3 and 22. We speculate such unique structures to be associated with sexual dimorphism and may play an essential role in beetle copulation. The tribe Cyathigerini has been documented for exhibiting remarkable sexual dimorphism in the antennal club and has been suggested that the female block-like antennal club inserts inside the excavated male antennal club, during copulation (Sugaya et al., 2004). Furthermore, internal morphological differences between males (morphospecies 6 and 8) may be an example of species variation within this particular genus. However, behavioral ecology on mating in pselaphine beetles largely remains unknown.

All members of the tribe Clavigerini are believed to be obligate myrmecophiles and are completely dependent on their ant hosts (Chandler, 2001; Park, 1942). Both morphospecies recorded in the current study exhibit morphological modifications including: reduced mouthparts for receiving trophallaxis (mouth-to-mouth liquid-feeding) from the worker ants (Park, 1942; Parker & Grimaldi, 2014), a reduction and compaction of antenna and abdominal segments that supports the beetle when handled by ant workers and provides a larger surface area for glandular secretions to spread and volatize (Cammaerts, 1992; Parker, 2016a; Parker & Grimaldi, 2014); long tufts of trichomes that ornament the base of the abdomen providing an important functional role during trophallaxis feeding by ant workers and social integration inside the ant colonies (Cammaerts, 1992; Parker & Grimaldi, 2014). Trichomes facilitate the dispersion of secretions produced by secretory Wasmann glands that are both attractive to and consumed by the worker ants thus initiating trophallaxis by the ant to the beetle (Hill et al., 1976; Parker & Grimaldi, 2014).

The presence or absence of one or more myrmecophile functional trait does not always define a myrmecophile. Morphospecies 13 and 37 (belonging to the tribe Hybocelaphini genus Apharinodes) recorded in the current study, both exhibit compact antenna and partial fusion of sternites and tergites, typical functional traits of myrmecophile Pselaphinae beetles. As previously discussed, the biology of many pselaphines is simply unknown, and therefore, it may be possible that morphospecies 13 and 37 form some kind of association with ants. Compact antenna and partial fusion of sternites and tergites are defensive modifications that allow pselaphines to withstand aggression from ants in the leaf litter (Parker, 2016a). Many facultative pselaphines also exhibit these defensive modifications indicating morphospecies 13 and 37 may exhibit some degree of myrmecophily, one that enables them to survive a high frequency of encounters with ants in the leaf litter (Parker, 2016a). Furthermore, morphospecies 13 and 37 are densely covered in squamous pubescence, not “true” trichomes, and this pubescence is thought to conduct secretions from nearby glands (Parker, 2016a). This type of pubescence also occurs in myrmecophile pselaphines belonging to the closely related tribe Ctenistini (Parker, 2016b) further suggesting some kind of association with ants. Morphospecies 32 (Batrisini) recorded in the current study, although lacks the obvious functional traits associated with myrmecophily (as previously discussed) exhibits a unique morphological antennal modification, whereby antennomere six and seven create a convex V shape. Although the functional role of such an antennal modification remains unknown, it may act as a “handle” to aid worker ants during transport of the beetle both inside and outside the ant nest, if at all this particular pselaphine morphospecies forms any association with ants. There are few if any, behavioral studies carried out on Pselaphinae beetles, given the difficulty posed by their small sizes and cryptic microhabitats, therefore it is difficult to conclude whether or not morphospecies 13, 32, and 37 form any association with ants. Behavioral studies on pselaphine beetles are needed to complement morphological studies and provide a better understanding for future ecological studies.

The current study presents a link between adult morphological characters and the functional role they possibly play in the behavior of myrmecophilous leaf litter pselaphines. Ant diversity may be a direct driver of pselaphine beetle FD, thus providing an insight into the dynamic ecological interactions pselaphines exhibit within the microhabitat of the tropical forest floor. This offers potential for exploring further the functional roles of pselaphines within leaf litter invertebrate communities and ecosystem functioning, particularly in determining changes in ecosystem processes in vulnerable habitats.

CONFLICT OF INTEREST

The authors have no conflicts of interest to declare.