Auke Bay Laboratory

Use of Polynuclear Aromatic Hydrocarbon Weathering Model to Identify Exxon Valdez Oil in Sediments and Tissues

Scientists at the Auke Bay Laboratory (ABL) used a first-order loss-rate kinetic model of polynuclear aromatic hydrocarbon (PAH) weathering to evaluate 7,767 environmental samples for the presence of oil in connection with the Exxon Valdez oil spill (EVOS).  The model was developed from experiments with gravel coated with crude oil and washed continuously with seawater for 6 months.  The modeled PAHs included the 14 most persistent of 31 hydrocarbons analyzed by gas chromatography/mass spectroscopy.  The parameters in the model were loss-rate constants related to the energy required for PAHs to escape from petroleum and a quantitative index of weathering.  The model accounts for 91% of the temporal variability of modeled PAH concentrations.  We compared the discrepancies between measured and model-predicted PAH concentrations of EVOS samples with a probability distribution of these discrepancies derived from the experimental weathering results.  Only 1,541 field samples contained sufficient PAH concentrations for valid application of the model; 75% fit the model at a > 0.01 type I error, 9% fit an alternate model characterized by the absence of weathering, 17% fit neither model, and a few fit both models.  The total of 1,164 samples that fit the weathering model accounted for 86% of the summed PAH concentrations detected in all 7,767 samples.  We conclude that first-order loss-rate kinetics accounts for the dominant PAH weathering processes in the EVOS samples and that the rate of weathering is determined mainly by the ratio of the surface area of sediment to the volume of petroleum in the environment.

The Exxon Valdez Trustee Hydrocarbon Database containing these results has been released to the public.  The database also contains the analytical results for 63 hydrocarbons (including 44 PAHs) deterined in each of the 7,767 environmental samples collected.  These samples were acquired at a cost exceeding $10 million, by far the most comprehensive collection of consistently analyzed samples of any oil spill.

By Jeffrey Short.

 

Effects of Windchill on Tanner Crabs

Commercial crab-pot fisheries catch male and female crabs ranging in size and carapace condition but legally can retain only large males. As a result, many crabs are returned to the sea, which has caused concern about the effects of handling on discarded crabs.  Previous experiments on crabs have shown that exposure to extremely cold air temperatures has detrimental effects, including limb mortality, limb autotomy, and reduced vigor.  During Tanner crab fishing seasons in the Bering Sea from 1974 to 1995, the daily air temperature on St. Paul Island ranged from 10.8° to -21.7° C (mean was -1.2° C ± 5.7 SD), and wind speed ranged from 0.7 to 21.1 meters/second (m/s) (mean 8.1 m/s  ± 3.3 SD). Because the combined effects of wind and cold air temperature on crabs had not been examined, the ABL took part in a joint experiment with the Alaska Department of Fish and Game (ADF&G) to assess the potential impacts of windchill on crab survival and activity.

In November 1997, Tanner crabs were exposed to three wind speeds (0.0, 7.7, and 15.9 m/s) and three air temperatures (0.3° , -10.4° , and -17.1° C) in a two-factor design. Results examined were crab death, duration of time to turn right side up (righting time), and loss of legs (limb autotomy).  Each treatment group contained 10 crabs.  Crabs were exposed to wind for 5 minutes, then immediately returned to water.  Relative humidity ranged from 68% to 100% during aerial exposure. After exposure, crabs were held in aquaria for 30 days.  Mean water temperature during and after exposures was 6.1° ± 0.2° C.  Unexposed crabs served as controls.   Crabs were monitored for righting times at five periods during the 30-day experiment:  1-2 days, 6-7 days, 14-15 days, 21-22 days, and 28-29 days after exposure.  Leg loss was monitored throughout the experiment.  The experiment is considered preliminary because of the large intervals between each temperature and each wind speed, and because some aspects were not replicated (e.g., mortality).

Table 1. Percent crabs dead as a function of wind speed (meters per second) and air temperature (degrees Centigrade). There was only one observation per cell, hence no associated error terms or statistics. Wind speed (m/s) 0 ± 0 m/s 7.7± 0.3 m/s 15.9 ± 0.3 m/s Temperature (°C) Mean % crabs dead Mean % crabs dead Mean % crabs dead no air exposure 0     0.34 ± 0.05 0   0    0 -10.4 ± 0.2 0 90 100 -17.1 ± 0.2 0 100 100

Crab death increased as a function of air temperature and wind speed (Table 1 above).  No crabs died in the 0.3°C exposures or in the -10° and -17°C with no wind.  But when exposed to -10° and -17°C temperatures and also subjected to wind (8 to 16 m/s), 90% to 100% of the crabs died.  All deaths were recorded within 4 days of exposure, and the majority of the mortalities (90%) had taken place within 2 days of exposure.

Table 2. Mean (± SE) percent crabs not righting as a function of wind speed and air temperature. Dead crabs were scored as not righting. Asterisks indicate responses significantly different from unexposed controls: * P < 0.05, ** P < 0.01, *** P < 0.001. Wind speed (m/s) 0 ± 0 m/s 7.7± 0.3 m/s 15.9 ± 0.3 m/s Temperature (°C) Mean % crabs not righting Mean % crabs not righting Mean % crabs not righting no air exposure 0 ± 0     0.34 ± 0.05 2 ± 4   6 ± 5**    0 -10.4 ± 0.2 0 ± 0 100 ± 0*** 100 ± 0*** -17.1 ± 0.2 6 ± 9 100 ± 0*** 100 ± 0***

The percentage of crabs not righting increased significantly  (P < 0.001) as a function of air temperature and wind speed (Table 2 above).  (Because all the crabs that started the experiment were included in this data set, the crabs that died were included with the live crabs that were unable to right themselves.)   Percentage not righting increased in the -10° and -17°C treatments (where most crabs died), but there was no trend at 0.3°C.  Percentages not righting increased in the 8-and 16-m/s treatments, but increases where there was no wind were not significant (P = 0.097).  The nonrighting response was controlled in large part by crab mortality, and responses in remaining cells (all 0.3°C observations and all observations without wind) were likely due to random noise.

Table 3. Mean (±SE) righting times (seconds) as a function of wind speed and air temperature as first observed after exposure; a hyphen (-) indicates that all the crabs in that treatment died. Wind speed (m/s) 0 ± 0 m/s 7.7± 0.3 m/s 15.9 ± 0.3 m/s Temperature (°C) Mean righting times Mean righting times Mean righting times no air exposure 22 ± 10     0.34 ± 0.05 8 ± 2 11 ± 4 10 ± 3 -10.4 ± 0.2 6 ± 2     -17.1 ± 0.2 38 ± 16

In the first measures of righting time, 2 to 3 days after exposure, results were inconclusive (Table 3 above).   There was no significant difference between the exposed groups and the controls (because of the long and variable righting times of the control crabs), but within the exposed groups, righting times increased significantly (P < 0.001) with a decrease in air temperature.  The mean righting time for crabs exposed to -17°C and no wind was significantly greater than in any remaining treatment.

Figure 1. Righting speed as a function of recovery time for each nonfatal treatment combination. Plotted data are means ± SE.
  

Over the 30 days of the experiment, the righting times of surviving crabs varied significantly (P <  0.001) as a function of air temperature (Figure 1 above).  Without any wind, the mean righting time in the -17°C treatment (21.59 seconds) was significantly greater than in the -10°C treatment (5.45 seconds) and the 0.3°C treatment (8.99 seconds).  However, the mean righting time of crabs in the unexposed control group was also high (17.36 sec).  Thus, statistically, the righting times of the treatment groups were not significantly different from that of the control group.

In the only group of surviving crabs (-17°C, no wind) where righting time was elevated significantly above that in controls, the righting time decreased significantly over time (P = 0.027, r = -0.32) following treatment (Figure 1 above).  Again, variation in righting times in the other treatments was high and casts doubt on the interpretation of results, but the magnitude of response at -17°C and uniform change in recovery as a function of time after exposure argues for a bona fide response for that group and argues for additional experiments.

Results of previous experiments support the hypothesis that righting time increases as air temperature decreases.   In this study, only those crabs exposed at 0.3°C survived at each wind speed, and the righting times of the surviving crabs did not vary significantly (P = 0.074) as a function of wind speed.  Thus, these results do not support the hypothesis, but we did not expect a significant response at this exposure temperature.

Table 4. Mean (±SE) percent limb autotomy as a function of wind speed and air temperature. Asterisks indicate responses significantly different from unexposed controls: * P<0.05, **P< P.01, *** P<0.001. Wind speed (m/s) 0 ± 0 m/s 7.7± 0.3 m/s 15.9 ± 0.3 m/s Temperature (°C) Mean autotomy Mean autotomy Mean autotomy no air exposure 0 ± 0     0.34 ± 0.05 0 ± 0     1 ± 1** 0 ± 0 -10.4 ± 0.2 0 ± 0 4 ± 3 4 ± 2 -17.1 ± 0.2   8 ± 3*     21 ± 8*** 4 ± 2

Tanner crabs autotomized limbs as a result of exposure to cold air (P < 0.001) and wind (P = 0.008), but interaction was also significant (P = 0.026) (Table 4 above).  Autotomy increased along the temperature gradient at each wind speed.  Across wind speeds, autotomy was negligible at 0.3°C, was slightly elevated at -10°C, but was greatest at 8 m/s in the -17°C.  It is possible that the -17°C, 16 m/s treatment was so severe that crabs died before the autotomization response could occur.

Because these preliminary results indicate that windchill leads to more severe effects than indicated by temperature alone, further testing is indicated to collect sufficiently detailed information so that response to wind may be accurately modeled.  These models, in turn, would then be utilized to recommend possible changes in the fishery.  Preliminary results suggest that  exposure of bycatch to severe weather may partially explain historical crab stock declines. Because regulatory changes are costly and have severe allocation effects on some users, there must be solid evidence for the existence of a real problem before the state of Alaska will propose a change in the way the fishery operates.

By Mark Carls.

 

Recovery of Pink Salmon Spawning Habitat after the Exxon Valdez Oil Spill

Researchers at the ABL examined the level of oil contamination in sediment at pink salmon streams in Prince William Sound to document initial levels of oil exposure and subsequent habitat recovery after the Exxon Valdez oil spill.  Results of this study are designed to complement other Exxon Valdez Trustee-funded studies on oil-related embryo mortality in pink salmon.  We analyzed over 300 sediment samples collected from 172 stream deltas in 1989-91 by the ADF&G, and we analyzed an additional 71 samples collected from 12 stream deltas (9 oiled, 3 nonoiled) in 1995 to determine habitat recovery.  Most samples were taken from stream banks immediately adjacent to pink salmon spawning areas.  Samples were fast-screened by ultraviolet fluorescence to measure total concentration of petroleum hydrocarbons (TPHC), and samples with detected oil were analyzed by gas chromatography/mass spectroscopy to determine concentrations of PAHs.

In 1989, TPHC was bimodally distributed: 65 streams were oiled (mean TPHC > 1,000 mg/g), and 85 streams were nonoiled (TPHC below method detection limit of 2 mg/g).  Oil contamination depended on exposure of the streams to ocean currents and the oil trajectory.  In 1995, 6 years after the oil spill, petroleum hydrocarbons were still detected in sediment at 8 of 9 oiled streams: mean TPHC ranged up to 240 g/g, and total PAH concentration in individual samples ranged up to 1,300 ng/g.  The residual PAHs had high molecular weight, and relative abundance was consistent with weathered Exxon Valdez oil.  Interpolation between mean concentrations in 1989 and 1995 indicated that many oiled stream deltas still had a total PAH concentration higher than 3,800 ng/g through the 1992-93 salmon incubation period. (The 3,800 ng/g concentration equals the lowest sediment concentration that has been associated in the laboratory with impaired survival of pink salmon embryos.)  We conclude that leaching of residual weathered oil into incubation substrate could explain persistent elevated embryo mortality in pink salmon from oiled streams through 1993.

By Mike Murphy.

 

Age Validation and Analysis of Ageing Error from Marked and Recaptured Sablefish

In cooperation with the Age and Growth Task of the Center’s Resource Ecology and Fisheries Management (REFM) Division, ABL staff evaluated the accuracy of methods used for estimating the ages of sablefish by comparing the ages determined by two experienced AFSC age readers to known ages. A mark-recapture experiment on sablefish provided a relatively large sample (N = 49) of 2- to 9-year-old sablefish. This sample of known-age fish provided a unique opportunity to evaluate the accuracy of methods used for ageing young sablefish. The study generally confirmed the criteria used to age sablefish. Although the age readers misaged about two-thirds of the fish,  most (81% and 71% for the two readers, respectively) of the misaged fish were misaged by only 1 year. After reexamination of the otoliths, most of the discrepancies between the reader-estimated age and the known age could be resolved. Use of ageing errors derived from known and reader-determined ages should improve the accuracy of recruitment estimates obtained from the stock assessment model for sablefish.

By Jon Heifetz.

 

Sablefish and Rockfish Stock Assessments Completed

Stock assessment reports for the combined Gulf of Alaska (GOA), Bering Sea and Aleutian Island (BSAI) sablefish stock, and slope rockfish and pelagic rockfish in the GOA were completed in November.  The 1998 allowable biological catch (ABC) recommendation for the GOA-and BSAI-combined sablefish stock was 16,800 metric tons (t), down slightly from the 1997 value of 17,200 t, due to a continued decline in the estimated abundance.  The GOA Pacific ocean perch recommendation for 1998 was 12,820 t, down slightly from 1997 due to recalibrations in the  improved computer model program. ABC recommendations for other slope rockfish in the GOA remained the same as for 1997.  Recommendations for pelagic rockfish in the GOA differ slightly from 1997 by treating the nearshore component of the central GOA separately.  These reports were presented to the North Pacific Fishery Management Council GOA and BSAI plan teams in November. Final Total Allowable Catches based on these reports were adopted at the December meeting of the Council.

By Jeff Fujioka.