Examination of spotted sand bass (Paralabrax maculatofasciatus) pollutant bioaccumulation in San Diego Bay, San Diego, California

Fish collection

Spotted sand bass from the 2002 non-tissue report (Marty, 2003) were collected via hook and line from central San Diego Bay (Fig. 1), with a total of 253 fish caught from September 25–29, 2002. The spotted sand bass from the 1989 and 2002 tissue studies were collected in 1989 and from 1999–2002, respectively, via trawls and hook and line within various locations in San Diego Bay (Fig. 1). Fish from the 1989 study were collected by the County of San Diego while the remaining fish were collected beginning 10 years later through the State of California’s Coastal Fish Contamination Program (SDCDHS, 1990; Gassel, Brodberg & Roberts, 2002).

Necropsy and lab analysis: tissue reports

Fish for tissue analysis were killed by wet ice immersion. Following capture each fish was weighed, measured, and examined for lesions and abnormalities (e.g., fish hooks). Fork lengths (FL) measurements of the Serranidae in southern California are considered equal to total length (TL) by the California Department of Fish and Wildlife’s State Finfish Management Project, and thus the two terms are used interchangeably. Spotted sand bass for tissue analysis were stored on ice and delivered within 24 h for lab processing. Fish not processed within 24 h were frozen until analysis.

Processing and analysis of 1989 report tissues was conducted by Pacific Analytical, Inc., with methods, quality assurance (QA) and quality control (QC) procedures described in detail in SDCDHS (1990), Appendix H. Processing and analysis of 2002 report tissues was conducted by the State of California’s Water Pollution Control Lab and Moss Landing Marine Laboratory, with QA/QC procedures followed as prescribed in a Quality Assurance Program Plan (Ichikiwa, 2000) in accordance with USEPA requirements and recommendations (USEPA (U.S. Environmental Protection Agency), 2000). Fish from each site were frozen at −20 Celsius until dissection. Fish were dissected into fillets and composited into groups, with each composite containing four to six fish. Fifteen composites were prepared in 1989, and twenty-five prepared from 1999 to 2002. The 1989 report tissue composites within each site were created by size class using weight, with the first sample containing the four smallest fish and last containing the four largest. Composites from the 2002 report targeted fish above the historic minimum take size (305 mm length), and were created using USEPA’s 75 percent rule (USEPA (U.S. Environmental Protection Agency), 2000), where the smallest fish was no less than 75 percent of the length of the largest fish in the composite. Fish for tissue analysis were not aged or sexed. 1989 report tissue composites were analyzed for percent moisture and lipid content, DDTs (4,4′ DDE and DDT), PCBs (Aroclor 1254 and 1260), and mercury. 2002 report tissues were analyzed for a suite of pollutants, including all DDTs, total PCBs, and mercury

Necropsy and lab analysis: non-tissue report

The necropsy procedure for age and sex determination in the 2002 non-tissue study was performed on-vessel in San Diego Bay, with fish held in pumped San Diego Bay seawater for a period no longer than eight hours, though holding times were typically much less. Fish were killed as described in Stehr, Myers & Willis (1994), with minor modifications. Spotted sand bass were killed via an overdose of anesthesia followed by exsanguination during necropsy (Marty, 2003). Tissues for histopathology were then removed as described by Marty (2003) followed by removal of the right otolith. Processing and analysis of 2002 tissues and otoliths for sex and age, respectively, was conducted at the Central Histological Facility in Sacramento, CA, with methods described in detail in Marty (2003) and Marty et al. (1998), Marty et al. (1999).

Data entry and transformation

All data were entered into Microsoft Excel (Microsoft 2012) and exported to R for data summary and statistical analysis (R Core Team, 2013). Data were log-transformed when necessary to meet test assumptions. Assumptions of normality were evaluated visually while those of homogeneity of variances were evaluated using the F-test. Where assumptions were not met, data was evaluated using alternative statistical methods. For statistical comparisons, fish classified by Marty (2003) as female-intermediate or male-intermediate were analyzed as female and male, respectively, unless otherwise specified (Sadovy & Domeier, 2005).

Bioaccumulation

As fish collected for tissue analysis were not aged or sexed, for bioaccumulation linear regression was used to investigate relationships between mean composite size, as weight and fork length, to observed tissue pollutant concentrations. Linear regression was then performed for organic pollutants in each composite to evaluate if there was any relationship between organic pollutant bioaccumulation rates. Subsequent analysis was then conducted to determine if differences in bioaccumulation between sites and years for each pollutant occurred using MANOVA. Pollutant concentrations were normalized to body weight or length where necessary based on linear regression results. For analysis sites were pooled to represent different ecoregions within San Diego Bay (see Komoroske et al., 2012), while years were pooled to represent two classes of 1989 and 1999–2002, as the bulk of samples from the 2002 report were collected in 2001. Analysis of PCBs was limited to Aroclors 1254 and 1260 to allow for comparison of 1989 and 2002 tissues. For MANOVA size classes for analysis were restricted to those that overlapped with previous age ranges analyzed. Lastly, mean composite pollutant concentrations were compared to OEHHA published fish tissue goals and tiered advisory levels for recreational anglers (Klasing & Brodberg, 2008) using Bonferroni adjusted t-tests.

Growth modeling

The von Bertalanffy growth curve model was used to develop individual growth curve models for spotted sand bass by sex. Fork lengths by age and body weights were modeled using a common von Bertalanffy growth model (Beverton, 1954; Beverton & Holt, 1957) represented by: ${\displaystyle E\left[L\right|t]=L_{\infty }(1-e^{-K(t-t_0)})}$ where:

• E[L|t] = the expected or average length at time (or age) t,

• L_∞ = the asymptotic average length,

• K = the so-called Brody growth rate coefficient (units are yr⁻¹), and

• t₀ = a modeling artifact that is said to represent the time or age when the average length was zero.

Models were developed for each group using the R package as prescribed by Ogle (2011). Subsequent model comparison by groups was conducted by means of ANOVA after testing assumptions of model generalization for variable parameters.

Where von Bertalanffy models could not be fit, relationships were analyzed using linear regression to test for correlation between age and fork length or weight. Fork length data was log-transformed where necessary and to provide a comparison to the findings of Allen et al. (1995).

Size and weight structure

To examine the size and weight structure of spotted sand bass in San Diego, the 253 individuals from the 2002 non-tissue report were analyzed for differences in mean length and weight by sex utilizing Welch’s t-test with Bonferroni corrections for multiple comparisons. Length frequencies for each sex were subsequently compared to the historic legal size limit for the recreational fishery utilizing t-tests (305 mm, California Department of Fish and Game, 2012), which best represents the size limit for fish taken during the report time frame.

Bioaccumulation

No relationship was found between composite fish weight or total length and pollutant concentration for mercury or DDTs (Fig. 2). A positive relationship was observed when comparing PCB tissue concentration to body weight (df = 1, 37, F = 6.65, r² = 0.1523, p < 0.05, Fig. 3) and fork length (df = 1, 37, F = 7.14, r² = 0.1618, p < 0.05, Fig. 3). Organic pollutants were found to be significantly correlated (df = 1, 12, F = 11.13, r² = 0.336, p < 0.01). Subsequent MANOVAs were attempted for mercury, body-size adjusted PCBs, and DDTs, with no differences found between ecoregions or between the 1989 and 1999–2002 time periods. There was no evidence of a year: ecoregion interaction. PCB and mercury mean composite tissue concentrations were found to exceed California Advisory Tissue Levels for recreational anglers, and multiple individual composites had tissue concentrations exceeding no consumption thresholds (Table 2, Figs. 2 and 3). Mean DDTs were found to be at least an order of magnitude less than advisory tissue levels, with DDE being the only variant detected.

Growth models

Growth models were developed for spotted sand bass utilizing age data for observed fork lengths and body weights. Individual von Bertalanffy models for fork lengths and body weights were attempted for each sex. All models developed failed initial assumptions required for modeling, and for those which start data were able to be calculated, individual growth models were unable to be evaluated statistically or run with accuracy due to model failure or exclusion of portions of the dataset to obtain fit. Visual evaluation of age-weight and fork length-weight relationships (Figs. 4 and 5) indicated that primary growth occurred prior to the first year of age, as indicated by the literature, with subsequent age classes showing a consistent range in variability for weight and fork length for each age class, indicating a very fast approach towards asymptotic growth. However, there was high variability within the population and subsequent analysis of groups by linear regression found no evidence of fish age having a relationship with observed fork length or body weight, making at-age take predictions impossible to calculate.

Size and weight structure

A total of 253 spotted sand bass were sampled for the 2002 non-tissue report. Lengths sampled ranged from 217 to 359 mm fork length (FL), well above the documented lengths of young of year fish (Pondella & Williams, 2009). Fork length differed between the sexes of spotted sand bass, with males having greater lengths, (df = 243.331, t = 6.157, p < 0.0001). Body weight exhibited a similar pattern, with males having a significantly heavier body weight than females (df = 235.798, t = 6.8352, p < 0.0001). Subsequently comparison of mean FL by sex to the 2012 minimum take size for the recreational fishery (30.48 cm TL, California Department of Fish and Game, 2012) found that the mean size of female and male fish were each less than that required for take in the fishery (Female: df = 97, t = 18.8, p < 0.001; Male: df = 154, t = 9.9, p < 0.001, Fig. 4).