Facultative expression of alternative male morphologies is thought to allow individual males to select the phenotype with the highest fitness gain given their competitive status relative to other males with which they compete for females. Choice of, or switching between, morphs commonly relies on developmental threshold responses. Evolutionary changes in developmental threshold responses are thought to provide an important avenue for phenotypic diversification and the evolution of morphological and behavioral novelties. However, the extent to which alternative male phenotypes and their underlying threshold responses actually evolve in natural populations is unclear. Likewise, the ecological factors that shape the evolution of threshold responses in natural populations are unexplored for most organisms, as are the consequences of such modifications for patterns of morphological diversity. I examined the ecological basis of rapid threshold evolution in exotic populations of the horn-polyphenic dung beetle Onthophagus taurus. Male O. taurus vary continuously in body size as a function of larval feeding conditions. Only males that exceed a critical threshold body size develop a pair of long horns on their heads, whereas males below this threshold remain hornless. Populations in two exotic ranges of this species, the eastern United States and western Australia, have diverged in the mean threshold body size, which has resulted in the evolution of highly divergent and novel horn length-body size allometries in these populations. Populations in a third and previously unstudied exotic range of O. taurus in eastern Australia exhibit threshold body sizes roughly intermediate between the eastern U.S. and western Australian populations. I tested three hypothesis to explain how differences in ecological and demographic factors can drive allometric divergences between populations, using data derived from comparative, standardized sampling of a large number of populations in each exotic range. Results suggest that differences in the intensity of both intra-and interspecific competition have contributed to the evolution of divergent thresholds in these populations. My results do not support the hypothesis that shifts in threshold body sizes to larger body sizes are a consequence of increases in the mean body size of competing males. I discuss my results in the context of Onthophagus mating systems and the evolutionary implications of threshold evolution.