the range of concentrations presented, the regres- 

 sion equation relating percentage detection and 

 the logarithm of concentration was not significant 

 (F = 1.3, P = 0.30) (Figure 1). The curve for 

 naphthalene detection was sawtooth-shaped with 

 only four concentrations where the percentages 

 of crabs detecting were above the upper 907c 

 confidence limit about the control value. The 

 sawtooth curve produced three concentrations at 

 which 509^ of the crabs could have detected 

 naphthalene, 10~^ 10~', and 10"** mg/1. Because 

 the factors producing the sawtooth curve are 

 unknown, the most conservative approach is to 

 consider the uppermost concentration, 10 ~^ rng/l, 

 as the threshold for naphthalene detection. In 

 contrast to naphthalene, the percentage of crabs 

 detecting the WSF decreased in a consistent way 

 with the WSF concentration (Figure 1). The re- 

 gression equation was significant (F = 60.4, 

 P<<0.01), and the variability was low iR^ = 

 91.0%). The SO'/f detection threshold from the 

 regression equation was 4 x 10""* mg/1, about 100 

 times lower than that for naphthalene. 



When a crab detected naphthalene or WSF, the 

 response was usually distinct. For crabs meeting 

 the detection criteria, the median ratios of the 

 antennular flicking rates did not vary with con- 

 centration (Median Tests, x^ = 2.38, P = 0.12 for 

 naphthalene; x^ = 9.07, P = 0.75 for WSF), so that 

 what varied with concentration was the percent- 

 age of crabs responding and not the magnitude 



60 



40 



Figure 1. — Percentage of Dungeness crabs detecting naphtha- 

 lene and the water soluble fraction (WSF) of crude oil as a 

 function of the logarithm of concentration (mg/1). The percent- 

 age of crabs detecting a control of membrane-filtered sea water 

 was 28.8% ( n = 66) for naphthalene and 26.8% ( n = 41) for WSF 

 The horizontal dotted line is the 90% confidence limit for 

 the control values for both naphthalene and WSF (38%). 

 The number beside each point is the number of trials at 

 the concentration. 



of the response. Also, the magnitudes of the 

 increase in antennular flicking were the same for 

 both naphthalene and WSF. For naphthalene, the 

 grand median of the antennular flicking rate 

 ratios was 2.04; for the WSF, the grand median 

 was 1.96. 



Discussion 



When presented with naphthalene or WSF of 

 crude oil, Dungeness crabs changed antennular 

 orientation and flicking rate in the same manner 

 as when presented with a clam extract. The blue 

 crab also gives the same detection behaviors 

 for hydrocarbons as for food (Pearson and 011a 

 1977. 1979, 1980; Pearson et al. in press). The 

 similar findings in both species indicate that ' 

 chemoreception by these crustaceans is not re- 

 stricted to chemical cues for food and, thus, agree 

 with Ache's (1975) suggestion that the chemical 

 spectrum sensed by decapod crustaceans is really 

 quite broad. 



While the manner of antennular response to 

 naphthalene and WSF was the same as that to a 

 clam extract, the magnitudes of the flicking in- 

 crease were slightly less and the chemosensory 

 thresholds were 10^ and 10^ times higher than 

 those found for the clam extract (Pearson et al. 

 1979). The grand median ratios of flicking rates 

 for naphthalene and WSF, 2.04 and 1.96, were 

 lower than that for the clam extract, 2.67. Also, the 

 ranges of flicking ratios for the hydrocarbons were 

 <30% of that for the clam extract. The slightly 

 less intense response and much higher thresholds 

 suggest that the petroleum hydrocarbons rank as 

 much less potent chemical cues than sapid chemi- 

 cals from a natural food. 



Previously, Pearson and 011a (1980) had hy- 

 pothesized that the chemical and chemosensory 

 processes producing a higher detection threshold 

 for a single petroleum hydrocarbon, naphthalene, 

 than for a complex mixture of hydrocarbons, the 

 WSF of crude oil, are analogous to the processes 

 producing a similar relationship of thresholds for 

 single amino acids and complex mixtures. Usu- 

 ally, food extracts and complex mixtures of amino 

 acids and other chemicals have a lower detection 

 threshold than that of a single amino acid ( Mackie 

 1973; McLeese 1974). Indeed, with the Dungeness 

 crab the detection threshold for WSF was about 

 100 times lower than that for naphthalene. Also, 

 the variability in detection was much less for WSF 

 than for naphthalene. This apparent greater dif- 



824 



