Toxicity of single magnesium pulse exposures to tropical freshwater species
Internal Report 608
Supervising Scientist Division
AC Hogan, RA van Dam, MA Trenfield & AJ Harford
Department of Sustainability, Environment, Water, Population and Communities
Concentrations of solutes originating from the Ranger uranium mine do not occur at constant levels in Magela Creek. They can fluctuate quite widely through time due to changes in creek discharge, mine water discharge and mine water source. Continuous monitoring of electrical conductivity (EC) in Magela Creek since 2006 has shown the extent of variability in creek water quality associated with mine water discharges.
Electrical conductivity is the key signature variable for water originating from the minesite, and is dominated by the concentration of magnesium sulfate (MgSO4). Consequently, concentrations of MgSO4 can be confidently predicted based on EC measurements. A large body of research has been undertaken by the Supervising Scientist Division on the toxicity of MgSO4 to local aquatic species. This work has led to the derivation of a site-specific water quality trigger value (TV) for magnesium (Mg) in Magela Creek of 2.5 mg/L (Mg being the primary toxic ion in MgSO4; van Dam et al 2010). However, the Mg toxicity data and associated TV were derived from tests conducted using continuous constant exposures over days (3 to 6 days depending on the species) as recommended by ANZECC/ARMCANZ (2000). This exposure regime is not representative of the majority of conditions in Magela Creek as evidenced by the continuous monitoring data.
This work was started several years ago. At that time comparison of the proposed chronic exposure Mg TV with the EC continuous monitoring data from the 2005-06 to 2008-09 wet seasons revealed 43 exceedances of the TV. The medians for Mg concentration and exceedance duration were 3.4 mg L-1 and 6.1 h, respectively. All except one of the exceedances were of durations shorter than the duration of the chronic toxicity tests. Therefore, they were considered unlikely to be causing detrimental effects downstream of the mine, but there were no quantitative data to support this assumption. Given the high conservation value of the Magela Creek catchment, it was considered necessary to better understand the potential effects of short-term exceedances (pulse exposures) of the Mg TV.
The aims of the study were to (i) assess the toxicity to local freshwater species of Mg pulse exposures relevant to those seen in Magela Creek, and (ii) use the data to develop a model from which Mg or EC TVs could be derived for any given exposure duration.
Six local freshwater species (green alga, Chlorella sp.; duckweed, Lemna aequinoctialis; snail, Amerianna cumingi; cladoceran, Moinodaphnia macleayi; green hydra, Hydra viridissima; and northern-trout gudgeon, Mogurnda mogurnda) were exposed to single Mg pulse exposures of 4 h, 8 h and 24 h duration (at a constant Mg:Ca ratio of 9:1, as per van Dam et al ). At the end of each exposure period the test organisms were transferred to clean water and their responses monitored for the remainder of the standard toxicity test durations (3 to 6 days, depending on species). At least two toxicity tests were undertaken for each combination of species and pulse duration. A limited number of continuous Mg exposure toxicity tests were completed to confirm the responses and toxicity values reported in van Dam et al. (2010).
For all species, Mg toxicity increased as exposure duration increased. However, the extent to which toxicity increased ranged from 2-fold to 40-fold between species. Moreover, the nature of the positive relationship between toxicity and exposure duration differed (linear or exponential) between species. The concentrations of Mg resulting in 10% inhibition of response (IC10) for each species are provided in Table ES.1.
For one species, the cladoceran, M. macleayi, increased sensitivity to Mg was observed following pulse exposures at the onset of reproductive maturity (at ~27-h old) compared with exposure of neonates (<6-h old). This increased sensitivity may be related to the coincidence of the exposure pulse with the physiological processes of moulting and/or reproductive development.
The IC10 data shown in Table ES.1 were used to derive 99% species protection TVs for each exposure duration, based on the assumption of log-logistic species sensitivity distributions. The resultant TVs for each pulse exposure duration are shown in Table ES.2.
The functional relationship between the Mg and EC TVs (Table ES.2) and exposure duration is well described by an exponential function. This model provides an interpolation tool by which a TV for any given pulse duration can be calculated. The model will be incorporated into a Mg/EC TV framework that will enable improved interpretation of transient pulses of EC in Magela Creek downstream of the Ranger mine.
It is recommended that further work be done on assessing organism recovery time and the potential for carry-over toxicity in situations where multiple pulses occur in series to enhance interpretation and application of the Mg/EC TV framework.