Plant species richness: the world records.

摘要

Questions: The co-existence of high numbers of species has always fascinated ecologists, but what and where are the communities with the world records for plant species richness? The species-area relationship is among the best-known patterns in community ecology, but does it give a consistent global pattern for the most saturated communities, the global maxima? Location: The world. Methods: We assembled the maximum values recorded for vascular plant species richness for contiguous areas from 1 mm2 up to 1 ha. We applied the power function to relate maximal richness to area and to make extrapolations to the whole Earth. Results: Only two community types contain global plant species maxima. The maxima at smaller spatial grain were from oligo- to meso-trophic, managed, semi-natural, temperate grasslands (e.g. 89 species on 1 m2), those at larger grains were from tropical rain forests (e.g. 942 species on 1 ha). The maximum richness values closely followed a power function with z=0.250: close to Preston's 'canonical' value of 0.262. There was no discernable difference between maxima using rooted presence (i.e. including only plants rooted in the plot) vs shoot presence (i.e. including any plant with physical cover over the plot). However, shoot presence values must logically be greater, with the curves flattening out at very small grain, and there is evidence of this from point quadrats. Extrapolating the curve to the terrestrial surface of the Earth gave a prediction of 219 204 vascular plant species, surprisingly close to a recent estimate of 275 000 actual species. Conclusions: Very high richness at any spatial grain is found only in two particular habitat/community types. Nevertheless, these high richness values form a very strong, consistent pattern, not greatly affected by the method of sampling, and this pattern extrapolates amazingly well. The records challenge ecologists to consider mechanisms of species co-existence, answers to the 'Paradox of the Plankton'.Digital Object Identifier http://dx.doi.org/10.1111/j.1654-1103.2012.01400.x