ficulty in detecting the single hydrocarbon than 

 the more complex WSF is presumptive evidence 

 for the hypothesized analogy. With naphthalene 

 constituting only 0.49^ of the total hydrocarbons 

 in the WSF, the crabs were probably responding 

 primarily to other compounds or, perhaps, to some 

 sort of odor medley. 



One possible explanation for the extreme vari- 

 ability in naphthalene detection is that detection 

 at high naphthalene concentrations was inhibited 

 by some toxic, narcotic, or anesthetic action not 

 present or much reduced at low concentrations. 

 The blocking of chemosensory feeding and mating 

 responses in the crab Pachygrapsus crassipes 

 after 24-h exposure to naphthalene at 10 ~^ nig/1 

 (Takahashi and Kittredge 1973) supports the 

 possibility of such inhibition. If the threshold 

 concentration for chemosensory inhibition was 

 within the range of concentrations we presented, 

 then a sharp increase in the percentage of crabs 

 detecting naphthalene would be expected below 

 the inhibition threshold and would produce the 

 sawtooth-shaped curve seen for naphthalene in 

 Figure 1. A sawtooth-shaped curve would also 

 result if the sensitive antennular chemoreceptors 

 were more impaired than the less sensitive body 

 chemoreceptors on the dactyls, chelae, and mouth- 

 parts. If the antennular chemoreceptors were the 

 more impaired at high naphthalene concentra- 

 tions, detection would occur primarily through 

 body chemoreceptors, and the antennular flicking 

 increases would then derive from a reflex pri- 

 marily involving the body chemoreceptors rather 

 than one involving the antennular chemorecep- 

 tors. If the supposed chemosensory inhibition 

 lessened or disappeared at low naphthalene levels, 

 detection would switch to the more sensitive 

 antennular chemoreceptors from the less sensitive 

 body chemoreceptors. Whatever the explanation, 

 the weak and inconsistent detection of naphtha- 

 lene did not allow estimation of a threshold 

 concentration by the method used here for WSF 

 and elsewhere for food extracts (Pearson and 011a 

 1977; Pearson et al. 1979). Without more evidence 

 concerning the mechanisms producing the partic- 

 ular shape of the naphthalene curve, the use of the 

 apparently real peak in detection at 10~^ rng/1 for 

 estimating thresholds remains an open question. 

 The most conservative approach for now is to 

 consider 10 ~^ rng/1 to be the naphthalene detec- 

 tion threshold. 



For both food extract and petroleum hydro- 

 carbons, the blue crab has exhibited more acute 



chemoreception than the Dungeness crab (Pear- 

 son and 011a 1977, 1979, 1980; Pearson et al. 

 in press). Pearson et al. (1979) hypothesized that 

 the lower detection threshold for clam extract 

 seen in the blue crab was a consequence of the blue 

 crab's greater ability to sample the chemical 

 environment with its higher flicking rate and 

 larger antennules. This hypothesis would apply 

 equally to the differences between the two crabs in 

 the hydrocarbon detection thresholds. 



An important practical question is how the 

 ability of the Dungeness crab to detect petroleum 

 hydrocarbons compares with the range of hydro- 

 carbon concentrations likely to be encountered by 

 the crab. In the water column during an oil spill, 

 McAuliffe et al. (1975) found concentrations of 

 dissolved hydrocarbons ranging from 2 x 10"^ to 

 2 X 10"^ mg/l. Of these dissolved hydrocarbons 

 about one-half were the monoaromatics domi- 

 nating the WSF used here. During a spill from a 

 North Sea platform, Grahl-Nielsen (1978) found 

 petroleum hydrocarbon concentrations ranging 

 up to 4 X 10 ~^ mg/1. In the open sea between Nova 

 Scotia and Bermuda, Gordon et al. (1974) found 

 petroleum hydrocarbon concentrations of 2.04 x 

 10"^ 8 X 10"^ and 4 X 10'^ mg/1 at the surface, 

 1 m, and 5 m. These concentrations roughly agree 

 with those given for relatively uncontaminated 

 oceanic areas by Clark and MacLeod (1977), who 

 also stated that chronically contaminated areas 

 have hydrocarbon concentrations about two orders 

 of magnitude higher than those of the open sea. 

 Unfortunately, analytical difficulties in distin- 

 guishing petrogenic from biogenic hydrocarbons 

 at low environmental concentrations make esti- 

 mates of oil levels in chronically contaminated 

 areas uncertain. For the North Sea, Grahl-Nielsen 

 et al. (1979) found that despite considerable oil 

 production there was no apparent standing crop of 

 petroleum hydrocarbons, but rather petroleum 

 contamination occurred as localized, transient 

 patches. Thus, the petroleum hydrocarbon concen- 

 trations in uncontaminated (10"'* to 10 ~^ mg/1), 

 chronically contaminated (10""* to 10~^ mg/1), and 

 oil spill (10"^ to 10"^ mg/1) situations are all at or 

 above the WSF detection threshold (10"-* mg/1) so 

 that Dungeness crabs can detect hydrocarbons 

 readily at the concentrations found in oil spill 

 situations, probably in chronically contaminated 

 situations, and marginally in uncontaminated 

 situations. In being able to detect the petroleum 

 hydrocarbons at concentrations at and below those 

 found in oil spill situations, Dungeness crabs can 



825 



