Environ. Sci. Technol. 2004, 38, 6716-6723
Modeling the Effect of Algal Dynamics on Arsenic Speciation in Lake Biwa FERDI L. HELLWEGER* AND UPMANU LALL Department of Earth and Environmental Engineering, Columbia University, New York, New York 10027
Algae reduce and methylate arsenate and the end product of the reaction is correlated to their growth rate. At slow growth rates, dimethylarsinate (DMA) is produced, and at fast growth rates arsenite (As(III)) is produced. Previous work has linked this phenomenon to the phosphorus luxury uptake mechanism of algae, and a model was developed for the process (Hellweger et al. Limnol. Oceanogr. 2003, 48, 2275). This paper presents the integration of that process model for arsenic transformation by algae into a full ecological model and application to Lake Biwa, Japan. The model application allows for a quantitative analysis of the field data, consistent with the process model and the ecological dynamics of the lake. The newly developed ecological model includes a variable phytoplankton composition, which is needed to simulate luxury uptake. This is in contrast to most existing ecological models, which typically assume a fixed “Redfield” composition. The model adequately reproduces the observed ecology of Lake Biwa, including the rapid uptake of phosphate by phytoplankton without immediate growth (luxury uptake) following lake overturn. The model also reproduces the observed arsenic speciation, including the gradual appearance of DMA during the summer and peaks in As(III) concentration at the onset of spring and fall algal blooms.
Introduction Arsenic Transformation by Algae. In oxygenated surface waters, arsenate (As(V)) is the only thermodynamically stable species, but As(III), methylarsonate (MMA), and DMA are often present as a result of phytoplankton activity (1-3). Algae transform As(V) (AsO(OH)3) because of its similarity to the essential and often growth-limiting nutrient PO4 (PO(OH)3). Algae actively absorb As(V) because they mistake it for PO4. However, inside the algal cell the similarities between As(V) and PO4 break down and As(V) is toxic. In what is thought to be a detoxification mechanism (1), algae reduce As(V) to As(III), methylate it to MMA and DMA, and excrete it, mostly as As(III) and/or DMA. The end product of the transformation reaction is correlated to the growth rate of the algae and to the phosphorus (P) nutrient status of the algae (4) as illustrated in Figure 1. At slow growth rates, under P-limited conditions (Figure 1a) the algae take up As(V), reduce it to As(III), methylate it to MMA and DMA, and then excrete it as DMA. At fast growth rates, under P-replete conditions (Figure 1b) the algae up-regulate their PO4 * Corresponding author present address: Civil & Environmental Engineering Dept, Northeastern University, Boston, MA 02115; phone: (617)373-3992; fax: (617)373-4419; e-mail:
[email protected]. 6716
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004
transport system to assimilate P in excess of their immediate growth requirements (luxury uptake). Since As(V) is taken up by the PO4 transport system, it is also taken up at higher rates. The reduction to As(III) is fast, but the methylation is slower, causing As(III) to build up in the cell and be excreted into the medium. DMA is also produced at this time, but at a much slower rate than As(III). Ecological Models as Research Tools. In many water bodies, As(III) peaks occur preceding or coincident with algal blooms (5), and qualitatively, the mechanism described above is in agreement with this observation. Consider, for example, an As(III) peak preceding a spring bloom in a lake. At the onset of the bloom, the algae are P-replete and are growing fast, causing them to produce As(III). It is reasonable (and convenient) to assume that, at the top and back end of the algal bloom, the specific growth rates are slower due to nutrient (P?) limitation. At that time the As(III) production ceases according to the proposed mechanism, in agreement with the field observation. However, the decrease in the spring phytoplankton bloom could also be due to zooplankton grazing (6), and it is not certain that the algae are P-limited or that their specific growth rate decreases. If the decrease in phytoplankton is due to grazing or limitation by a nutrient other than P, the algae should (according to the proposed mechanism) continue to produce As(III), which would be at odds with the field observations and would contradict the proposed mechanism. The phytoplankton dynamics of laboratory experiments are simple, but those of natural waters are complex and vary in response to a multitude of factors. An ecological model provides the opportunity for an integrated analysis, including and constrained by many of the factors that potentially affect the arsenic speciation (e.g., zooplankton grazing and vertical mixing). Integrating the algae-arsenic process model into an ecological model allows for a quantitative analysis of the proposed mechanism (4) and field data, consistent with the ecological dynamics of the lake. If the integrated model is able to reproduce the major trends in the field observations, with minimal parameter tuning, then the hypothesis of arsenic transformation by phytoplankton is supported. Lake Biwa. Lake Biwa, Japan, was selected for a first fieldscale application of the model because a large database of arsenic concentrations (3) and limnologic parameters (e.g., chlorophyll a and PO4; 7, 8) is available, and the data indicate that arsenic speciation in the lake is in qualitative agreement with the model (e.g., As(III) peaks preceding phytoplankton blooms). A map of the lake can be found in the paper by Sohrin et al. (3) and in the Supporting Information. Lake Biwa consists of a larger north basin and a smaller south basin. The model is applied to the north basin, which has a surface area of 616 km2, average depth of 44 m, and residence time of 5 yr. Physically, the lake is classified as monomictic (T > 6 °C), typically thermally stratified from April through December with the thermocline located at a depth of about 18 m. Ecologically, the north basin is characterized by low to moderate productivity (oligotrophic/mesotrophic, average chlorophyll a ) 4 µg L-1). A subsurface maximum in chlorophyll a is typically present at a depth of about 10 m. Dissolved oxygen (DO) is consumed in the hypolimnion during stratification, but to date, the north basin has not been observed to go anoxic. Total arsenic concentrations average about 8 nmol L-1 (
