Tropical forests: their richness in coleoptera and other arthropod species

TLDR: It is concluded that current estimates of Arthropod species numbers are grossly underestimated; that there could be as many as 30 million species extant globally, not 1.5 million a, usually estimated.

Abstract: Extrapolation from data about canopy insects collected by fogging methods together with estimates of tropical plant host specificity indicate that one hectare of unrich seasonal forest in Panama may have in excess of 41,000 species of arthropods. Further extrapolation of available data based on known relative richness of insect Orders and canopy richness leads to the conclusion that current estimates of Arthropod species numbers are grossly underestimated; that there could be as many as 30 million species extant globally, not 1.5 million a, usually estimated. Since the early days of naturalists, there has been the question of how many species there were in the forests of the tropics. Bates (1892) wrote of collecting more than 700 species of butterflies within an hour's walk of his home in Para, Brazil. Many have guessed that the arthropod fauna of the world today contains between 1.5 to 10 million species. No hard data are available however, and these estimates are less than reliable and as a result misleading. In a recent paper, Erwin and Scott (1980) provided the first hard data with regard to the Coleoptera fauna of a single species of tree in the tropical seasonal forest of Panama. Also recently, Peter Raven of the Missouri Botanical Gardens wrote me with the same inquiry that Bates had pondered-"How many species are there in one acre of rich tropical forest?" With the hard data available from the Panama study, I set out to give as close an estimate as possible and was shocked by my conclusions. The tropical tree Luehea seemannii is a medium-sized seasonal forest evergreen tree with open canopy, large and wide-spaced leaves. The trees sampled (n = 19) had few epiphytes or lianas generally, certainly not the epiphytic load normally thought of as being rich. These 19 trees over a three season sampling regime produced 955+ species of beetles, excluding weevils. In other samples now being processed from Brazil, there are as many weevils as leaf-beetles, usually more, so I added 206 (weevils) to the Luehea count and rounded to 1,200 for convenience. There can be as many as 245 species of trees in one hectare of rich forest in the tropics, often some of these in the same genus. Usually there are between 40 to 100 species and/or genera, so I used 70 as an average number of genus-group trees where host-specificity might play a role with regard to arthropods. No data are available with which to judge the proportion of host-specific arthropods per trophic group anywhere, let alone the tropics. So conservatively, I allowed 20% of the Luehea herbivorous beetles to be host-specific (i.e., must use this tree species in some way for successful reproduction), 5% of the predators (i.e., are tied to one or more of the hostspecific herbivores), 10% of the fungivores (i.e., are tied to fungus associated only with this tree), and 5% of the scavengers (i.e., are associated in some way with only the tree or with the other three trophic groups) (Table 1).