Pollen input to, and incorporation in, two crater lakes in tropical northeast Australia.

Pollen input to the water surfaces of maar crater lakes Barrine and Eacham (1km(2) and 0.5km(2) respectively, ca. 720m a.s.l.), surrounded by rainforest in northeastern Australia, were measured in floating Tauber traps for a variety of periods spanning the years 1978 to 1985. The mean weekly total pollen catch is 75% of that estimated for the surrounding region of varied vegetation. Regardless of their sources, the influxes of pollen taxa are unevenly distributed over the lakes' surfaces. There is similar variation through time, with some indication that this might be smoothed over periods of 5years or more. There are no inflow streams and pollen input down surrounding slopes is negligible. Mean total pollen influx to seston traps close to the lakes' bottoms (1979 to 1988) is <5% of that to the Tauber traps, a discrepancy attributed to trap design. In contrast, mean annual influx measured by the Tauber traps is similar to that estimated from dated sediment samples younger than 1966 AD at Lake Barrine. The proportions of different pollen taxa are less affected by trap position or period of exposure than are their influxes; their mean percentages for Tauber traps, seston traps and sediment samples are all rather similar.The similarity between the mean contribution of exclusively rainforest pollen to the lakes' surfaces (56%) and to Oldfield traps in rainforest interiors (51%) emphasizes the importance of the rainforests, which clothe the crater walls, as sources of pollen to the lakes. The impact of pollen from other vegetation in the region, although present, is much smaller than at interfaces between rainforest and other vegetation types. All this pollen is mixed within and above the crater and delivered to the lakes' surfaces by gravity and frequent rainfall. Limnological processes redistribute it in the water body before it is incorporated into the deep-lake sediments.The value and limitations of current pollen transport and accumulation theory are noted in relation to sites with morphologies similar to these two crater lakes and to the irregular flowering and floristic inhomogeneity of tropical rainforest. The potential for the use of modern pollen input data in the interpretation of pollen analyses from the sediments of such sites is explored.