Limitations of growth-parameters in Lemna gibba bioassays for arsenic and uranium under variable phosphate availability.

Growth behaviour of Lemna gibba L. at different phosphorus supply, and arsenic or uranium exposure levels was investigated in batch culture. Total frond count, total frond area, and dry biomass were used to observe growth at four phosphate, arsenic, and uranium concentrations. L. gibba dry biomass had a linear relationship with the total frond area (r=0.96 at P<0.001), but a non-linear relationship with the total frond count. Using total frond count led to significantly lower growth rates than did the use of total frond area in high phosphate supply, whereas the use of total frond count led to higher growth rates than did total frond area under low phosphorus conditions. The results under uranium loading were more or less a mirror of the results obtained under phosphate variation, except that the growth rates for L. gibba under uranium exposure were lower than in the phosphate experiments. At the lowest arsenic loading, the total frond count was significantly higher than the total frond area, whereas at higher arsenic loading, this was not observed because of arsenic was toxic to L. gibba. The longest roots were observed in lowest phosphate supply, and in highest uranium load. No significant influences on root length were observed with arsenic loading. The individual frond area increased with an increase in phosphate and decreased considerably with uranyl and arsenate loading. L. gibba multiplied rapidly into small and yellowish fronds under low phosphorus supply, and under high levels of arsenic and uranium exposure. L. gibba first developed the individual frond size before multiplying in a favourable environment. The effect of uranium may be attributed to an interaction between uranyl ions and phosphates that results in a precipitate of uranyl phosphates. The results revealed that L. gibba growth behaviour is very much influenced by multiple stresses in a bioassay. Therefore, growth rate calculation models should be chosen according to the parameter used, and toxicity estimation differs considerably with the growth parameter measured. Consequently, the interpretation of L. gibba bioassay results can be significantly influenced.