Remarkable fly (Diptera) diversity in a patch of Costa Rican cloud forest: Why inventory is a vital science

Abstract

Study of all flies (Diptera) collected for one year from a four-hectare (150 x 266 meter) patch of cloud forest at 1,600 meters above sea level at Zurqui de Moravia, San Jose Province, Costa Rica (hereafter referred to as Zurqui), revealed an astounding 4,332 species. This amounts to more than half the number of named species of flies for all of Central America. Specimens were collected with two Malaise traps running continuously and with a wide array of supplementary collecting methods for three days of each month. All morphospecies from all 73 families recorded were fully curated by technicians before submission to an international team of 59 taxonomic experts for identification. Overall, a Malaise trap on the forest edge captured 1,988 species or 51% of all collected dipteran taxa (other than of Phoridae, subsampled only from this and one other Malaise trap). A Malaise trap in the forest sampled 906 species. Of other sampling methods, the combination of four other Malaise traps and an intercept trap, aerial/hand collecting, 10 emergence traps, and four CDC light traps added the greatest number of species to our inventory. This complement of sampling methods was an effective combination for retrieving substantial numbers of species of Diptera. Comparison of select sampling methods (considering 3,487 species of non-phorid Diptera) provided further details regarding how many species were sampled by various methods. Comparison of species numbers from each of two permanent Malaise traps from Zurqui with those of single Malaise traps at each of Tapanti and Las Alturas, 40 and 180 km distant from Zurqui respectively, suggested significant species turnover. Comparison of the greater number of species collected in all traps from Zurqui did not markedly change the degree of similarity between the three sites, although the actual number of species shared did increase. Comparisons of the total number of named and unnamed species of Diptera from four hectares at Zurqui is equivalent to 51% of all flies named from Central America, greater than all the named fly fauna of Colombia, equivalent to 14% of named Neotropical species and equal to about 2.7% of all named Diptera worldwide. Clearly the number of species of Diptera in tropical regions has been severely underestimated and the actual number may surpass the number of species of Coleoptera. Various published extrapolations from limited data to estimate total numbers of species of larger taxonomic categories (e.g., Hexapoda, Arthropoda, Eukaryota, etc.) are highly questionable, and certainly will remain uncertain until we have more exhaustive surveys of all and diverse taxa (like Diptera) from multiple tropical sites. Morphological characterization of species in inventories provides identifications placed in the context of taxonomy, phylogeny, form, and ecology. DNA barcoding species is a valuable tool to estimate species numbers but used alone fails to provide a broader context for the species identified.