Species-level, but not family-level diet breadth predicts geographic distribution of Sydney butterflies

1. Diet specialisation drives life-history adaptations and is an important factor determining the geographic distribution of species.

2. Previous empirical studies have shown that diet specialists should compose the majority of species in a given location, and theory predicts that generalists should have higher geographic distribution range compared with specialists (the niche breadth-range size hypothesis). Although the evidence in support of these predictions remain scarce for herbivorous insects from isolated regions of the world such as Sydney on the east coast of Australia. 

3. Here, I compiled data from a public database to test these predictions. By measuring the diet breadth and geographic distribution of Sydney butterflies, I showed that species-, genus- and family-level diet breadth agree with worldwide patterns for Lepidoptera, whereby diet breadth is composed largely of specialist species with a long tail of generalists. 

4. Furthermore, species- and genus- (but not family-)level diet breadths were positively correlated with geographic distribution, providing supporting evidence for the niche breadth-range size hypothesis. 

5. A machine-learning algorithm revealed that the positive relationship between diet breadth and geographic distribution was likely to be influenced by a common evolutionary history amongst species. 

6. The findings of this study in Sydney butterflies provide support for an important yet debated ecological hypothesis, thereby contributing to our understanding of ecological and nutritional factors driving butterfly species distribution.