Diversity and Distribution of Tropical Lepidoptera: a bit of cross-purposes

This is the first of the commentaries I promised yesterday. While the rest of the world seems to have become bizarrely fixated on some fossil find from some minor mammalian clade, yesterday's Nature also included two far more interesting papers on the distribution of herbivorous insects in tropical rainforests.

"Short-range endemics" is a bit of a buzzword here in Australia at the moment, referring to the pattern in a number of taxa, especially invertebrates, of large numbers of closely-related species of exceedingly restricted distributions (a study one of my supervisors recently conducted of subterranean arachnids called schizomids found that almost each individual mesa that housed schizomids housed its own individual species). The current papers could be very interesting in light of short-range endemism. They are also very interesting in light of the overall question of why the tropics are so hyperdiverse compared to higher latitudes.

As I said yesterday, the two papers differed somewhat in their conclusions (but more on that later). First off, the paper by Novotny et al. looked at diversity within a 75,000 square kilometre area of lowland rainforest in Papua New Guinea. While the area of rainforest was continuous, the Sepik River does cut through it, and some of the plant species compared had quite restricted distributions. Novotny et al. looked at Lepidoptera (caterpillars), ambrosia beetles (Scolytinae and Platypodinae) and Tephritidae (fruitflies) and compared the species found on each host plant genus investigated between eight sites. They found that there did not appear to be a significant change in species composition from one area to another - the species that were found on Ficus at one site were pretty much the same as those found on Ficus at another, over the entire area investigated. The Sepik River did not appear to be a major barrier to dispersal.

At the same time, Dyer et al. looked at average host specificity of herbivorous caterpillars at different latitudes in the Americas. They found that tropical species tend to have much higher host specificity than temperate species. This is in direct contrast to a paper Novotny et al. published last year, that found no significant difference in host specificity between taxa in Papua New Guinea and Europe. Instead, Novotny et al. attributed the increase in insect diversity in the tropics to the shear increase in number of potential host plant species.

So on the one hand we have a paper that seems to argue for wide distributions of tropical taxa, on the other we have one that argues for high host specificity (and hence, one suspects by implication, more restricted distributions). After reading through the papers, I don't think the conflict is actually that strong, as I'll explain in a moment.

Dyer et al. do offer some suggestions for why their results were different to Novotny et al.'s last year. One is that there may be actual difference between the Old World and the Americas. I just can't see that being significant - while there are some differences in which families are dominant in each hemisphere, there are many families that are present in both, and the latitudinal influences are still similar in each - far north it's still colder. The other factor that I think is far more likely to be significant is that Dyer et al. looked at a far greater range of host species than Novotny et al - the latter looked at 18 species in each area, while Dyer et al. looked at up to a maximum of 281 species in Costa Rica. Most significant of all, though, is that Dyer et al. looked at only one potential host species per genus per area. This would tend to bias their results towards higher measurements of host specificity, but is arguably more informative. If you compare a tropical species that feeds on three species of Ficus to a temperate species that is recorded feeding on one species each of Euphorbia, Quercus and Fagus, the temperate species should obviously be regarded as far less host-specific in light of the far greater phylogenetic distance separating its hosts. Unfortunately, a solely numerical metric will not distinguish the two.

Which brings us back to my point that the two Nature papers are not as contradictory as they first appear. The paper from Papua New Guinea compared species from different areas of the same genus. It looked at a different level of resolution than the Dyer et al. paper. As for the implications of the Novotny et al. paper for short-range endemism, the obvious point seems to be that most short-range endemics appear in taxa such as arachnids, myriapods and troglobites - taxa with relatively low dispersal capabilities. In contrast, Novotny et al. looked at insects - winged, and therefore one would expect able to disperse over greater distances more easily, so long as a suitable host plant was present when it got there. In support of this, spiders that disperse by a 'ballooning' stage when young (such as Nephila, the golden orb weaver) tend to be far less diverse with individual species found over much greater areas.

Of course, the number of potential host species in the tropics is still doubtlessly a factor. But this just begs a further question. If insects are so much more diverse because the plants are so much more diverse - then why are the plants more diverse?

PS. I really feel that I should mention that the study by Novotny et al. had a large proportion of the fieldwork conducted by locally trained staff, a number of whom are in the author list below. With the low levels of scientific education available in third world countries, the organisers of this study are to be commended on this front.

REFERENCES

Dyer, L. A., M. S. Singer, J. T. Lill, J. O. Stireman, G. L. Gentry, R. J. Marquis, R. E. Ricklefs, H. F. Greeney, D. L. Wagner, H. C. Morais, I. R. Diniz, T. A. Kursar & P. D. Coley. 2007. Host specificity of Lepidoptera in tropical and temperate forests. Nature 448: 696-699.

Novotny, V., P. Drozd, S. E. Miller, M. Kulfan, M. Janda, Y. Basset & G. D. Weiblen. 2006. Why are there so many species of herbivorous insects in tropical rainforests? Science 313: 1115-1118.

Novotny, V., S. E. Miller, J. Hulcr, R. A. I. Drew, Y. Basset, M. Janda, G. P. Setliff, K. Darrow, A. J. A. Stewart, J. Auga, B. Isua, K. Molem, M. Manumbor, E. Tamtiai, M. Mogia & G. D. Weiblen. 2007. Low beta diversity of herbivorous insects in tropical forests. Nature 448: 692-695.