16); salinities, 31.8 (±0.4)L (n = 5) and 32.0 

 (±0.0)'Z, (n = 4); and dissolved oxygen 7.6 (±0.7) 

 mg/1 ( n = 16) and 8.0 ( ± 0.3) mg/1 ( n = 9). Clumps 

 of the blue mussel, Mytilus edulis, and the little- 

 neck clam, Protothaca staminea, provided an ad 

 libitum diet. 



Experimental Apparatus 



We coupled the oil delivery system developed by 

 Vanderhorst et al. (1977), and used extensively by 

 Anderson et al. (1979, 1980), to the chemosensory 

 testing apparatus of Pearson et al. (1979). Sea- 

 water contaminated with Prudhoe Bay crude oil 

 was delivered to 20 of the 40 chemosensory testing 

 chambers from dripper arms situated along mani- 

 folds connected to the oil delivery system. Con- 

 taminated water entered each exposure chamber 

 at 0.1 1/min while clean water entered at 0.9 

 1/min. Control chambers received clean water at 

 1.0 1/min. Seawater entered each chamber through 

 a glass funnel connected to a slotted inlet tube 

 within the chamber. Teflon^ tubes carried sea- 

 water solutions of the clam extract to the funnels 

 from burets calibrated to deliver 20 ml within 15 s. 

 Previous dye studies of Pearson et al. (1979) 

 showed that the maximum concentration of an 

 introduced solution within a chamber occurs 10 s 

 after introduction and is 0.011 ( ± 0.003) times the 

 concentration of the introduced solution. 



The delivery system produced oil-contaminated 

 seawater that was largely a water-soluble fraction 

 with some finely dispersed droplets. The chemical 

 composition of this oil-contaminated seawater has 

 been well characterized by Bean et al. (1978) and 

 reported by Anderson et al. (1980). Here we 

 sampled seawater in the testing chambers by the 

 resin column absorption technique of Bean et al. 

 (1978) and analyzed the samples by infrared (IR) 

 spectrophotometry. The data of Bean et al. and 

 Anderson et al. show the correlations between the 

 values determined by IR and the concentration 

 of specific hydrocarbons determined by other 

 methods for the same system. To determine how 

 rapidly hydrocarbon concentrations dropped after 

 stopping the flow of oil-contaminated water in the 

 second experiment, we supplemented IR analyses 

 with analyses for monoaromatic hydrocarbons by 

 a helium gas partitioning technique modified 

 from McAuliffe (1971). 



^Reference to trade names doe.s not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



FISHERY BULLETIN: VOL. 79, NO. 4 



Experimental Solutions 



The experimental solutions were seawater solu- 

 tions of freeze-dried clam extract (FDCE) of little- 

 neck clam prepared following Pearson et al. 

 (1979). Stock solutions averaging 1.89 ( ±0.12) g 

 FDCE/1 (n ^ 6) for the first experiment and 2.06 

 (±0.22) g FDCE/1 in = 5) for the second were 

 refrigerated and used within 5 d. A 10 ^ dilution 

 of the stock FDCE solution was made 1 h before 

 testing with seawater freshly filtered through a 

 0.4 |Ltm membrane. An aliquot of the filtered 

 seawater used for dilution was used as the control 

 solution. All solutions were held in a water bath at 

 ambient seawater temperature. 



Procedures 



After the oil delivery system had been operating 

 for several days and the hydrocarbon concentra- 

 tions measured, a single Dungeness crab was 

 added to each of the 20 exposure and 20 control 

 chambers. Chemosensory testing was synchro- 

 nized to begin and end within either a rising or 

 falling tide and after 24-h exposure to oil-contam- 

 inated seawater. In the first experiment, the 

 FDCE solutions were presented with oil-contam- 

 inated seawater still flowing through the cham- 

 bers. Each crab was presented with either one of 

 two dilutions of FDCE or a control of filtered 

 seawater. After correction for dilution within a 

 chamber, these FDCE concentrations were 10^^ 

 and 10 ~^ g/1. The choice of dilution and the 

 order of presentation were randomized except that 

 active crabs and those with retracted antennules 

 were passed over The observer did not know the 

 identity of any solution. An individual crab was 

 observed for 60 s prior to introduction of experi- 

 mental solution, and the antennular flicking 

 rate and other behavior recorded. The observer 

 depressed a switch of an event counter for each 

 flick of one antennule. The solution (20 ml) was 

 then introduced and observation continued for 

 another 60 s from onset of introduction. 



The criteria of Pearson et al. ( 1979) were used to 

 score the behavior. Detection was indicated when 

 a crab abruptly changed antennular orientation 

 and increased antennular flicking rate so that the 

 ratio of the rate after solution introduction to that 

 before was 1.50 or higher. Previous observations 

 indicate that the a priori probability that such an 

 increase in antennular flicking is spontaneous, 

 rather than in response to the solution, is <5% 



642 



