The Big Picture: Evolution of Extreme Structures

We study the development and evolution of extreme morphologies in insects, such as the elaborate horns and mandibles of beetles.  After almost two decades studying the evolution of horns in dung beetles (see publications), we have expanded our scope to include the giant rhinoceros beetles and the stag beetles.  For the past 5 years, the majority of our efforts have been focused on the rhinoceros and stag beetles.  Dung beetles, rhinoceros beetles, and stag beetles represent three disparate lineages within the beetle superfamily Scarabaeidae, thought to have diverged from each other approximately 150 million years ago.  Traditional reconstructions of weapon evolution suggest that extreme male weaponry arose independently within each of these three lineages, permitting us to compare how exaggerated male weapons evolved in each instance -- i.e., which genetic and physiological pathways were modified to yield extreme growth of the weapon. [Figure 1]

A great deal is known about the natural histories of each of our focal species, providing a rich ecological context for our current studies of the genetics and development of weapon evolution.  For example, males in the dung beetle Onthophagus nigriventris, like those of many other species in this genus, fight over tunnels that have been excavated beneath dung by females.  Males spar awkwardly with their spear-like thoracic horns as they battle over the entrances to these tunnels, and winners mate repeatedly with the female inside the tunnel.  This species inhabits cool, moist pastures along mountainsides in East Africa such as the rim of the Ngorongoro Crater, and introduced populations thrive in high, cool pastures of Eastern Australia and Hawaii.  The giant Japanese rhinoceros beetle (Trypoxylus dichotomus) flies at night throughout much of East Asia, converging on sap flows on the sides of Fraxinus and other host trees.  Males fight over these feeding sites and mate with females as they feed (see video by Erin McCullough).  The stag beetle Cyclommatus metallifer also lives in East Asia and fights over sap flows on the sides of host trees, and many aspects of their mating system and diet are similar to those of T. dichotomus (yet the weapon, in this case, comprises a massively enlarged pair of mandibles rather than a rigid outgrowth horn).

Mechanisms of Exaggerated Growth: How does horn development differ from that of other structures?

We have been exploring how trait exaggeration arises in each of our three focal beetle lineages.  Because very little is known about the growth and development of horns or mandibles in these species, we are conducting basic exploratory studies of weapon growth [Figure 2].  This involves detailed observation of where in the animal and when during larval development these structures grow, as well as molecular “natural history” – descriptions of the expression domains of patterning genes in the growing appendages of each type of beetle [Figure 3].  In addition, we are comparing the development of several different appendages (weapons, wings, genitalia), so that we may discern the mechanisms specifically responsible for extreme growth in the weapons. 

Exaggerated ornaments and weapons of sexual selection have several properties that distinguish them from other body parts: (1) in the best-condition individuals they attain stunning proportions (exaggerated size), (2) their growth tends to be more sensitive to circumstance – the nutritional histories and physiological conditions of the males who wield them – than is the growth of other body parts (‘heightened’ condition-sensitive expression), and (3) their size differs more dramatically from male to male than do other traits (‘hypervariability’).  For example, big elk have longer legs than small elk, but the difference in leg length between the largest and smallest males pales beside the difference in antler size between those same animals.  Antlers, like most ornaments and weapons of sexual selection, range from enormous to almost non-existent, which is an extraordinary degree of among-individual variability. 

In our research we capitalize on these features of exaggerated male weapons by contrasting the growth of horns (or mandibles) with that of other traits.  For each of our focal species we are exploring patterns of gene expression associated with extreme weapon growth (by comparing gene expression in growing weapons sampled from large, well-fed males and small, poorly-fed males), and we are contrasting this with similar assays of gene expression in two other structures, wings and genitalia.  Horns (or mandibles), wings, and genitalia all develop at the same time in larvae, yet their patterns of growth differ dramatically.  Horns, wings, and genitalia represent a gradient in tissue-sensitivity to nutrition, from exquisite in the case of the weapons, to almost entirely insensitive, in the case of genitalia [Figure 4].  By contrasting gene expression in high-nutrition and low-nutrition individuals across these three traits, we are teasing apart candidate pathways for nutrition-dependent phenotypic plasticity (wings versus genitalia) and for exaggerated growth (horns versus wings).

For each of our focal species (the dung beetle Onthophagus nigriventris, the rhinoceros beetle Trypoxylus dichotomus, and the stag beetle Cyclommatus metallifer) we have sequenced transcriptomes and are now using RNAseq methods to compare transcript abundances across diet treatments and traits.  We also are developing population samples of tissues – weapons, wings, and genitalia dissected from animals at precisely the same stage and in the middle of their burst of growth – for approximately fifty males and thirty females of each species.  RNA extracted from these tissues will permit us to use RT-qPCR to measure transcript abundances for candidate genes in unprecedented numbers of individuals, for a truly ‘population-level’ examination of gene expression.  In essence, we are generating scaling relationships for transcript abundance that will test whether levels of expression of candidate genes correlate with among-individual variation in amounts of trait growth.

This work is a collaborative effort funded by the National Science Foundation, with Laura Corley-Lavine (co-PI), Ian Dworkin (co-PI), Ian Warren, Hiroki Gotoh, and Toru Miura.

A Spectacular Evolutionary Radiation: Weapon evolution in rhinoceros beetles

In order to characterize patterns of diversification of horns, we are constructing a phylogeny for the rhinoceros beetle subfamily Dynastinae.  We are collaborating with Andrew Smith, Dave Hawks, and Matt Paulsen, who have already been working on a molecular phylogeny for Scarabs based on collection material from the University of Nebraska collection, the Smithsonian Scarab collection, and the collections of the Canadian Museum of Nature.  To this we are adding material collected by Prof. Kazuo Kawano, and Jema Rushe is using next-generation sequencing technologies to bulk sequence DNA from these museum specimens (referenced against our transcriptome for T. dichotomus).  All told we plan to include approximately 300 Dynastinae species in this analysis.

Mechanisms of Horn Evolution: Comparing populations that differ in horn length

Much of our work is focused on using among-individual variation and among-trait variation to tease apart underlying mechanisms responsible for nutrition-dependent growth, and, in particular, the exquisitely nutrition-sensitive mechanism responsible for extreme trait growth.  These studies have already implicated the insulin/insulin-like growth factor (IGF) pathway as a critical player (Emlen et al. 2012), and they point to JH signaling as yet another candidate mechanism (see Gotoh et al. 2011).  Because we are conducting these studies in each of three separate lineages of scarab, representing (presumably) three independent evolutionary origins of extreme male weapons, we will be able to compare and contrast the extreme growth of horns to that of mandibles, and we can compare the mechanisms involved with growth of dung beetle horns to that of rhinoceros beetle horns.

However, to directly implicate changes in these (or other) mechanisms in the evolution of weapons, we must focus on a different sort of variation.  Specifically, we need to compare gene expression in horns sampled from one population, to that of horns sampled from another population, and these populations must be chosen so that they differ dramatically in weapon size.  In this case, observed differences in gene expression should point to the mechanisms responsible for recent evolutionary changes in weapon size.  We have already identified divergent populations of the rhinoceros beetle Trypoxylus dichotomus.  Our next phase of research will entail (1) a detailed study of the behavior of these two populations, to measure selection on horns and to try to discern why horn length might have diverged between these populations (in collaboration with Yoshihito Hongo), (2) rigorous characterization of gene expression differences between horns sampled from males in each of these populations (using RNAseq and our transcriptome for this species), and (3) a detailed comparison of the endocrine physiology regulating horn growth in the divergent populations (in collaboration with Adam Dolezal).