Implications of gut purging for tissue residues determined in bioaccumulation testing of sediment with Lumbriculus variegatus

INTRODUCTION

Many environmental contaminants commonly found in sediments have the potential to accumulate through the aquatic food chain, thus providing a pathway for exposure of higher trophic levels, including humans. Bioaccumulation tests have been developed as one mechanism for assessing the extent of bioaccumulation that may occur from contaminated sediments [1-3], and these procedures have been incorporated into certain regulatory programs [4].

For freshwater sediments, the oligochaete Lumbriculus variegatus is an organism commonly used for bioaccumulation testing of sediments. Conceptually, this is a simple procedure in which oligochaetes are exposed to test sediments for 28 d, then sieved from the sediments and analyzed for contaminants of interest. The U.S. EPA and ASTM [2, 3] protocols recommend a 24-h holding period in clean water at the end of the sediment exposure to allow organisms to purge their gut of sediment prior to analysis; this purging was included to prevent sediment-bound chemicals in the gut from being measured as a part of the tissue burden.

Inclusion of a 24-h purging period in the test protocol has been questioned by some because of concerns that tissuebound chemicals will depurate from the organisms during holding in clean water, thereby underrepresenting the steadystate concentration in the organism residing in the sediment [3, 5]. Some research has suggested that shorter purging periods, such as 10 to 12 h [6, 7], might be sufficient to clear sediment from the gut of oligochaetes. Although this would reduce depuration, it also presents logistical problems to conducting the test (e.g., test manipulations required late at night), which is a serious practical consideration for a procedure to be conducted, in large part, by contract laboratories.

The purpose of this paper is to provide a quantitative examination of several alternatives for ending bioaccumulation tests with L. variegatus and to discuss their consequences for interpretation of sediment bioaccumulation data. Literature data were used to estimate depuration of nonionic organic chemicals from tissues as a function of time and K_ow. A purging experiment was conducted to quantify rates of sediment elimination from the gut of L. variegatus. The goal in this analysis was not to develop a rigorous kinetic model for L. variegatus but rather to use available data and reasonable assumptions to develop practical guidance for the implementation of bioaccumulation testing.

METHODS

Predicting depuration of nonionic organic chemicals

Projected concentrations of organic chemicals in L. variegatus were based on two compartments, (1) sediment-associated chemicals present in the gut, which are eliminated during purging; and (2) tissue-bound chemicals, which depurate during purging in clean water. Thus, the expected wholebody concentration (c_{whole body}) at any time is equal to the sum of these compartments, such that

(1)

where w_sed(gut) is the proportion of the total body mass comprised of sediment, w_tissue is the complementary proportion comprised of tissue (w_sed(gut) + w_tissue = 1), and c_sed(gut) and c_tissue are the chemical concentrations in the respective compartments (all calculations are dry weight unless otherwise noted). The proportion of organism weight comprised of sediment in the gut was varied as part of the analyses and declined with time as calculated from the sediment purging experiments (see Results). Literature data suggest that L. variegatus do not preferentially ingest organic-rich particles [10], so the concentration of organic carbon in gut-borne sediment was assumed to be the same as in bulk sediment. For nonionic organic chemicals, sediment organic carbon was presumed to be the primary partitioning phase, so that

(2)

where c_OC is the concentration of chemicals in organic carbon and f_OC is the fraction organic carbon. Concentrations of chemicals in tissue were related to the concentration in lipid (c_lipid) by the fraction of lipid in the organism (f_lipid),

(3)

The relative concentration of chemicals in lipid and in sediment organic carbon is expressed by the biota-sediment accumulation factor (BSAF), which is the ratio of these values, that is

(4)

Solving Equation 4 for c_lipid and substituting into Equation 3 yields

(5)

For initial analyses, we assumed that, at the end of the uptake phase of a bioaccumulation study (t = 0 of the purging period), the BSAF was one (c_OC = c_lipid); later analyses evaluated BSAF as an experimental variable. A tissue lipid content of 1.0% (wet weight) for L. variegatus [11] and a moisture content of 80% [7] yields a dry weight lipid content of 5% (f_lipid = 0.05). Substituting these values into Equation 5, the concentration in tissue at the beginning of the purge period would be

(6)

Depuration from tissues during purging in clean water was estimated using an empirical relationship between chemical log K_ow and the elimination rate constant (k_T) developed by Gobas et al. [12] such that

(7)

where k_T (per day) represents total elimination by all routes. Chemicals remaining in the tissue (c_tissue) at time t (in days) is then predicted as

(8)

The Gobas model was developed from measured depuration of halogenated organic compounds from 100-mg guppies rather than oligochaetes. To determine whether this model provided a reasonable representation of the depuration that could be expected in oligochaetes, literature data for depuration from oligochaetes were compared to model predictions (see Results). Depuration rates were assumed to be independent of starting concentration. Concentrations were therefore expressed relative to the initial tissue concentration c_tissue,t=0, which is presumed to be the target endpoint for a bioaccumulation test.

RESULTS AND DISCUSSION

Literature estimates of the mass contribution of sediment to overall body mass of oligochaetes are typically in the range of 10 to 20% [7, 10, 13]; Norberg-King and Dawson [14] compared dry weight and ash-free dry weight in oligochaetes reared in four sediments and estimated that sediment contributed from 7 to 37% of total mass, with an average of 20%. Errors in dry weight-based chemical concentrations were estimated for gut sediment weights of 5 to 25% of total dry weight. When test animals are not purged, the predicted errors in chemical concentrations are independent of K_ow but are dependent on both the proportion of organic carbon in the sediment and the proportion of organism mass comprised of sediment (Fig. 1a). Because of the assumed condition that BSAF = 1, there is no error when f_OC = 0.05, which is the same as the assumed lipid content of the oligochaetes. Interestingly, errors were not particularly large over this range of conditions, with most values falling between 80 and 120% of the true tissue concentration. When lipid-normalized chemical concentrations are calculated for oligochaetes that have not been purged, errors tend uniformly toward overestimation and are greater in magnitude than for mass-based concentrations (Fig. 1b).

The gut purging experiment showed a clear and rapid decline in ash content of oligochaetes upon transfer to clean water without substrate (Fig. 2). Within the limits of data variability, ash content was stable after purging for 6 to 24 h with an average of 4.2% of dry weight. This mean residual ash content was subtracted from the mean ash content for hours 0 to 4, then both linear and exponential regression models were fit to these data. The exponential model showed a better fit (r = 0.977 vs 0.915) and described the ash content of the oligochaetes at time t (in hours) as

(9)

The resulting elimination rate constant of 0.7060/h projects that sediment purging is 76% complete in 2 h, 94% in 4 h, and 98.4% in 6 h; half-life is 0.98 h.

Given that our sediment purging study involved a single sediment, it is unknown whether sediments with differing physical characteristics or chemical contamination would show markedly different clearance rates. However, Kukkonen and Landrum [6] and Brooke et al. [7] evaluated several sediments with varying composition and contamination and showed that, despite the varying characteristics, gut purging in all sediments was complete after 10 to 12 h. While there were no published observations for purging times as short as 6 h, these studies provide evidence that gut purging is not delayed to a large extent by these factors.

In addition to providing time to clear sediment from the gut, holding organisms in clean water also allows depuration of tissue-bound chemicals. For nonionic organic chemicals, the depuration rate is related to K_ow. As comprehensive data were not available specifically for L. variegatus or oligochaetes in general, the relationship developed by Gobas et al. [12] was used to estimate depuration constants as a function of K_ow. Depuration rates predicted from this relationship were evaluated using clean water depuration data for L. variegatus exposed to fluoranthene [15], pyrene [6, 16], anthracene [16], fluorene [5; personal communication, B.R. Sheedy, U.S. EPA, Duluth, MN, USA], and benzo(a)pyrene [10] (Fig. 3). In addition, data for depuration of a variety of chlorinated organic compounds by other oligochaetes (Limnodrilus hoffmeisteri and Tubifex tubifex) were included [13, 17; as analyzed in 18] (these latter experiments involved purging in clean sediment rather than water only, which is thought to increase elimination rate [6]). Overall, these data suggest that the depuration constants derived by Gobas et al. [12] provide a reasonable representation of depuration from L. variegatus. It is worth noting, however, that the analysis would be strengthened by additional data specifically for L. variegatus and chemicals between log K_ow three and five, for which depuration is greatest in the first 24 h.

Using the depuration constants from Equation 7, depuration as a function of K_ow was plotted for a series of purging periods (Fig. 4) assuming no contribution from sediment in the gut. For compounds with a log K_ow greater than five, all values are projected to be within 10% of the initial concentration for depuration times up to 24 h. For compounds with a log K_ow less than five, however, substantial differences exist. To retain 90% of the initial concentration of compounds with log K_ow > 4, depuration for 8 h or less is required. Even after just 4 h of depuration, projected concentration of chemicals with log K_ow of three are only 80% of the initial concentration.

Given the opposing influences of gut purging and chemical depuration, selecting the appropriate procedure for bioaccumulation testing involves balancing the diminishing bias from gut contents against the growing bias from depuration. This issue does not have a single solution because it depends on the chemicals of interest, the required accuracy, and other variables that may vary from study to study. Many chemicals typically of interest in bioaccumulation studies (e.g., DDT, endrin, dieldrin, many PCBs, higher molecular weight PAHs) have log K_ow values greater than five, in which case 90% accuracy with little or no bias from gut contents can be achieved with longer purging periods such as 24 h. Nonetheless, results of the gut purging experiment suggest that elimination of sediment is essentially complete within 6 h, which may make this shorter depuration preferable in many instances.

A critical factor in this analysis is the sensitivity of modeled concentrations to model assumptions. The most critical of these appears to be the assumption that the BSAF = 1. Chemicals with very high K_ow may not reach equilibrium between sediment and tissue within the 28-d duration of the L. variegatus bioaccumulation test and hence would show BSAF values of less than one [3]. Growth dilution could also lower BSAF values. Perhaps more significantly, studies by Mc-Groddy and Farrington [19] and others have indicated that, in certain sediments, the apparent partitioning of PAHs between sediment organic carbon and pore water leads to much higher K_OC values than would be calculated from literature relationships [20], with resulting “apparent” K_OC values as much as three orders of magnitude higher. On the presumption that the interstitial water is indicative of the true chemical activity of the PAH in sediment [20], this suggests that BSAF values could be as much as three orders of magnitude below the value of one assumed in the analyses presented above.

These deviations have profound influences on the concentrations predicted to occur in sediment bioaccumulation tests. Because a BSAF substantially less than one means that the concentration of chemical will be much higher in the organic carbon of the sediment than in tissue lipid, the influence of gut-borne sediment on whole-body concentrations increases dramatically (Fig. 5a). Whereas whole-body concentrations of chemicals deviated from the true value by only about 15% for a gut sediment content of 15% and a BSAF of one (Fig. 1a), changing the BSAF to 0.01 increases the projected error to a more than 30-fold error in high TOC sediments (Fig. 5a). Purging for as little as 6 h eliminates the majority of the bias from low BSAFs (Fig. 5b). Assuming a gut sediment weight of 15% and an f_OC of 0.05, errors in whole-body residue after 6 h of purging would be expected to exceed 10% only for BSAFs less than 0.02.

Biota-sediment accumulation factors slightly higher than one (e.g., 1.7) have been reported and are thought to represent an upper bound for BSAF values for nonionic organic chemicals accumulated directly by benthic invertebrates, at least where active uptake is not involved [3]. As indicated by Figure 5, increasing the assumed BSAF from one to two has relatively little effect on predicted responses.

Metabolism was not explicitly considered in this analysis, although oligochaetes are not thought to possess a large capacity for metabolism of most xenobiotic chemicals [21, 22]. Metabolism of chemicals in the tissue would act to decrease the BSAF, thereby increasing the importance of purging, although it would simultaneously speed elimination of chemical from the tissue during the purging period. For chemicals/sediments showing small BSAF values, measurement of chemical concentration in sediment pore water would likely provide insights as to whether low accumulation rates were caused by unusual partitioning behavior in the sediment or by metabolism.

Bioaccumulation tests can also be used to assess the accumulation of metals from sediment. Several authors have reported errors in analyses of metal bioaccumulation by invertebrates stemming from gut contents [23-26]. In addition, there is evidence that bulk sediment concentrations of metals do not correlate directly with the bioaccumulative fraction of metals in sediment [27], which would further confound the interpretation of tissue analyses that included sediment from the gut. Although the evaluation we have presented is oriented toward nonionic organic chemicals, the resultant purging recommendations should be applicable to studies focusing on accumulation of sediment-associated metals as well.

The impetus for this analysis was the uncertainty surrounding the consequences of alternative methods for terminating bioaccumulation tests with L. variegatus. The original guidance published by the U.S. EPA and ASTM [2, 3] prescribed a 24-h period for elimination of gut contents. Studies by Kukkonen and Landrum [10] and Brooke et al. [7] measured elimination of sediment after 10 to 12 h and found that purging was essentially completed, indicating that this shorter period of gut purging would be sufficient and perhaps preferable in order to reduce depuration of chemicals from tissue. Although technically preferable to a 24-h period, purging periods of 10 to 12 h would be logistically difficult, particularly for a test protocol that is used widely in a regulatory context. In the present study, we evaluated sediment purging more rigorously between 0 and 12 h and found that only 6 h were required for >98% purging of gut contents. Shortening the purge period to 6 h further reduces potential errors from chemical depuration, particularly for compounds with log K_ow < 5.

As illustrated in our analyses, the appropriateness of any protocol for ending the L. variegatus bioaccumulation test will depend on the accuracy required, the chemicals being measured, and the characteristics of the test sediments. In general, however, we believe that the use of a 6-h purging period in clean water will provide a reasonable balance among the biases presented by chemical depuration and sediment in the gut. From a practical standpoint, a 6-h purging period has the added advantage of allowing bioaccumulation tests to be ended within the confines of a normal workday. Nonetheless, where chemicals of interest have log K_ow > 5, longer purging periods, up to 24 h, can probably be used with only minimal errors. For chemicals with log K_ow substantially less than four or which may be rapidly metabolized, kinetic studies may be required to derive accurate estimates of tissue burdens.