Lipid content

The mean lipid content in blubber was low relative to the value of 84–85% reported in apparently healthy sea lions [23]. The condition of the carcass might be responsible for these differences in blubber lipid content among individuals but no significant difference was reported in the lipid content of the blubber between fresh, moderately decomposed and advanced decomposed carcasses. Indeed, some of the extremely low lipid contents were from fresh carcasses. The low lipid content of several animals may have resulted from the mobilization of fat reserves from the blubber prior to death due to starvation or disease, as suggested by Kajiwara et al. [22].

There was a strong negative relationship between lipid content of the blubber and tDDT and PCB concentration in the blubber. This relationship was observed in a previous study on California sea lions [22], which reported that the highest concentrations of both tDDT and PCBs were found in individuals with the lowest blubber lipid content. Similarly, other studies of pinnipeds and cetacea report a negative relation between blubber thickness and concentrations of organochlorines [14, 24–26]. This negative relationship can be explained by the fact that fat-soluble contaminants such as tDDT and PCBs are less easily mobilized from the blubber than lipids during fasting and thus tend to accumulate in the remaining blubber layer [24, 27].

Sex and age categories

The usual pattern observed in marine mammals is an increase of fat-soluble contaminants such as PCBs and tDDT in the tissues of males and immature females, while in mature females, levels decrease or remain unchanged, as they excrete PCBs and tDDT through the fetus and nursing. As a consequence, males are generally more contaminated than females and juveniles [9, 10, 12–14]. This has not been documented specifically in California sea lions and their habits make it difficult to do so. Males migrate north of the area for half the year while lactating females stay in the Southern California Bight all year round in close proximity to the tDDT hotspot near Palos Verdes. Sample sizes in this study were insufficient to adequately test for sex and age differences. Moreover, the females collected in central California may have been non-lactating, not yet parous, or simply younger, having experienced fewer reproductive cycles. More data are needed to address changes in contamination levels as a function of sex or age.

Location

The pattern of contaminant levels as a function of the stranding location observed in the present study was somewhat unexpected. It may, however, be explained in part by the fact that individuals, males in particular, move long distances and feed on a variety of prey, that themselves move significantly. The principal prey of California sea lions are northern anchovy (Engraulis mordax), Pacific whiting (Merluccius productus), mackerel (Trachurus symmetricus), rockfish (Sebastes spp.), and market squid (Loligo opalescens) [28–30], highly mobile, vertically and horizontally migrating, schooling species that are widely dispersed on or near the coastal continental shelf with distributions that vary seasonally and annually [31, 32]. As a consequence, the concentrations reported here may not necessarily reflect the contamination levels of prey in the area where the sea lions stranded.

tDDT/PCB and DDT/DDE ratios

tDDT concentrations in California sea lions were approximately 3 times higher than PCB concentrations, a pattern characteristics of marine mammals from the California coast. This is in contrast to the North Atlantic, Arctic, Gulf of St Lawrence, Mediterranean sea, and the coast of Washington, where PCB concentrations in the blubber of marine mammals are usually as high or higher (up to 1 order of magnitude) than tDDT concentrations [3, 9, 10, 13, 33–41]. Only a few studies, such as those on the contamination of Baikal seal (Pusa sibirica), northern fur seal (Callorhinus ursinus) and killer whales (Orcinus orca) in Oregon, report higher levels of tDDT than PCBs [3, 12, 42]. The scant information on marine mammals from the tropical and equatorial fringe of the Northern Hemisphere generally also report tDDT as the main pollutant, followed closely or equaled by PCBs [43]. Kajiwara et al. [22] observed higher levels of tDDT compared to PCB in the blubber and liver of California sea lions while they found the reverse in the livers of harbor seals (Phoca vitulina) in California. They concluded that California sea lions, stranded in central and northern California probably spent parts of the year feeding in the Southern California Bight while harbor seals, also stranded in central and northern California, remained at the same location throughout the year and thus received exposure only locally [22]. Sea otter (Enhydra lutris) tissues from California also showed higher levels of tDDT compared to PCBs, while in the tissues of the same species from Alaska, the reverse was reported [44, 45]. Thus, the tDDT/PCB ratio appears to be higher in California and may originate, at least partly, from the DDT hotspot in the Southern California Bight, the legacy of DDT production and byproduct dumping over 30 years ago.

The high proportion of DDE, a DDT metabolite, in tDDT is in the upper range of reports in the literature for other marine mammal species [22, 37, 39, 40, 43, 45–47]. A high proportion of DDE reflects the absence of recent DDT contamination. The ratio DDT/DDE in 2000 (0.0032) however is lower than in the studies of Le Boeuf and Bonnell [19] as well as Lieberg-Clark et al. [21] that reported ratios of 0.023 and 0.032, respectively. This suggests that the composition of tDDT contamination is still changing towards a lower amount of DDT. As the proportion of DDE seems to increase with age in adult males [10], differences between studies may vary in part with the number of older males in each sample.

Temporal trends

To better assess the decline in tDDT levels over the last 30 years and to avoid the potential confounding factor of sex, we compared males in this study with blubber lipid contents higher than 20% with a similar sample of males reported in previous studies [19, 21] (Fig. 3). Blubber tDDT levels were more than 90% lower in 2000 than in 1970. Even so, an accurate estimation of the rate of decrease is elusive because analytical methods for organochlorines have improved in accuracy over the last 30 years [43, 47]. Packed columns, commonly used before the 1980s, underestimated DDT concentrations by 30–40% [43]. In contrast, Addison and Stobo [47] mention DDT-group analyses that were internally consistent from the early 1970s to the early 1990s. In addition, a potential overestimation of tDDT levels due to their co-elution with some PCB congeners was noted for the California sea lion samples in the 1970s [25]. Nevertheless, the main interference would have been the co-elution of p,p'DDT with PCBs, which represents a small proportion of tDDT [[48] in [21]]. Thus, we conclude that the decline of tDDT during the last 30 years was over one order of magnitude.

The decrease in tDDT in California sea lions during the last 30 years, although impressive in magnitude, is in general accordance with declines in other marine mammals [43]. Moreover, although the mean level of tDDT is much lower in 2000 than in 1970, inter-individual variation remains high; several individuals in 2000 exhibited concentrations within the range of concentrations reported in 1970.

It is difficult to draw conclusions concerning the trends of PCBs during the last few decades. The studies in the 1970s reporting PCB concentrations in California sea lions dealt only with lactating females [23, 49]. Moreover, the techniques for analyzing PCBs in the 1970s (packed columns) overestimated the levels by about 50% [43]. PCB concentrations were also expressed as Aroclor mixture and not in a congener-specific approach [43, 47]. In general, a decrease in PCB levels in marine mammal tissues is often reported [43]. The magnitude of the drop seems to vary with time as well as with geographical areas. No systematic trends, or even increases in levels, are observed in some regions such as the Antarctic [43].

Inter-species differences and geographical trends

Despite the marked decline in tDDT in California sea lions over the last 3 decades, levels remain higher in this animal than in most other marine mammals, even when only samples with blubber lipid content >20% were taken into account. Mean tDDT concentrations in the blubber of California sea lions from the present study are higher (up to 2 orders of magnitude) than current levels reported in other marine mammal species, such as harbor seals, harp seals (Phoca groenlandica), grey seals (Halichoerus grypus), white whales (Delphinapterus leuca), fin whales (Balaenoptera physalus), bottlenose dolphins (Tursiops truncatus), harbor porpoises (Phocoena phocoena) from other geographical areas [27, 38–40, 43, 47, 50]. Similarly, PCB levels also appear to be higher than in several other species and geographical areas, although the difference is more moderate. In general, marine mammals from the West Coast of North America display among the highest DDT and PCB loads observed [43].

California sea lions are among the most highly contaminated species along with northeastern Pacific killer whales, St. Lawrence beluga whales, Mediterranean striped dolphin (stenella coeruleoalba), and Mediterranean bottlenose dolphins [42, 43, 51, 52]. It is remarkable that California sea lions exhibit PCB levels that are of the same order of magnitude as those reported in polar bears (Ursus maritimus) from Svalbard, which feed at a higher trophic level, while the tDDT concentrations in California sea lions are 2 to 3 orders of magnitude higher than those of polar bears [36]. Several factors such as sex, age, feeding habits, physiological status (sick, lactating, starving) and analytical techniques exert an influence on organochlorine concentrations and might be partially responsible for the difference between species and geographical areas.

Toxicological effects

Despite relatively high concentrations of tDDT and PCBs, there is no evidence that population growth or the health of individual California sea lions have been compromised. The population has increased throughout the century, including the period when DDT was being manufactured, used, and its wastes discharged off southern California. Few attempts to measure toxicological endpoints in California sea lions have been made [23, 49]. The increase in their numbers suggests a healthy population but does not necessarily mean that the contamination by these organochlorines has not exerted an effect on individuals. The concentrations of PCBs reported in the present study are above the 17 mg/kg lipid weight threshold that elicits effects at physiological levels on endocrine and immune systems [53, 54]. DDT and its metabolites have been shown to be hormone-disrupters, and to induce estrogenic as well as anti-androgenic effects on several species [55–57]. Premature pupping in the early seventies was attributed to the contamination by tDDT [23, 49], however, the role of DDT and its metabolites in this reproductive failure was not ascertained and remains unclear [25, 58]. A real cause-effect relationship between contaminant concentrations and factors of the immune or endocrine system is difficult to determine due to confounding factors. Nevertheless, concentrations reported here are sufficiently high to encourage continued investigation.

Sample collection

Blubber tissue samples were collected from 36 dead California sea lions stranded on islands and the coast of California between April and November 2000 (Table 1). The location, sex, standard length, estimated age category, and condition of the stranded pinnipeds were recorded. The condition of carcasses was characterized as fresh, moderately decomposed or less than two weeks dead, and in advanced decomposition or estimated to be more than two weeks dead. Animals from southern California were found on the Channel Islands, those from central California at Año Nuevo Island, and those from northern California at Marin, Mendocino and Humboldt counties. A tissue sample weighing approximately 150 grams was excised from the medial ventrum region at the level of the axilla, from outer skin to muscle. Samples were deposited in clean jars and refrigerated until organochlorine analysis was conducted.

Chemical analyses

The tDDT and PCB analyses were conducted as described elsewhere [22]. The PCB standard used for quantification was an equivalent mixture of Kanechlor preparations (KC-300, KC-400, KC-500, KC-600). Concentrations of total PCBs were based on the sum 84 peaks, representing 117 major congeners. These congeners represent all those that could be found in technical PCB preparations and in the environment.

Data analysis

All statistical analyses were conducted using the GLM procedure (SYSTAT 10.0 windows version). In order to normalize the data, we took the square roots of tDDT and PCB concentrations in blubber lipids. Two outliers (individuals C110400–06 and C110400–07, Table 1) were removed from the data analysis because of their extremely high PCB and tDDT levels (Table 1) which were far outside the range of the whole data set.

The effects of carcass condition (fresh, moderately decomposed, advanced decomposed) on blubber lipid content as well as on the PCB or tDDT levels in blubber lipids were first examined using a one-way fixed-model analysis of variance (ANOVA) since decomposition of the carcass and a resulting change in blubber lipid content can affect PCB or DDT concentration. However, carcass condition had no significant effect on either blubber lipid content or PCB/tDDT concentrations in blubber lipids (see Results section). For this reason and because of the relatively imprecise character of this factor, it was removed from further analyses.

The effect of blubber lipid content of the blubber (lipid), sex and age category (sex-age category), and the location where the sample was collected (location) on tDDT or PCB concentrations in blubber lipids were examined using a three-way fixed-model analysis of covariance (ANCOVA), with lipid as a continuous variable and sex-age category and location as categorical variables. In a preliminary analysis, we tested the homogeneity of slopes. As the interactions with the continuous variable (lipid) were non-significant in both PCB and tDDT analyses (p > 0.05), we assumed parallelism and pooled these interactions with the model mean square error. The effects of lipid, sex-age category and location were then re-tested under the reduced linear model (see Results section).