prawns. Based on our observations, we further note 

 that prawns appear to be less tolerant to low DO2 

 levels than many of the other species found in the 

 same depth range. However, published data on 

 tolerance to hypoxic conditions for the species pres- 

 ent exists only for munids, which have been shown 

 to tolerate hypoxic conditions as low as 0.1-0.15 mL 

 • L-i (Burd 1983; Burd and Brinkhurst 1984, 

 1985). 



All benthic species observed, except prawn, had 

 a relatively large depth range over which individuals 

 were found in abundance. Those species apparent- 

 ly more tolerant to low DO2 levels were found from 

 about 85-210 m. It is unknown why in contrast to 

 other species observed, prawns were concentrated 

 in a narrow depth range at 70-85 m water depth, 

 so close to lethal water conditions. Most prawns 

 were apparently prevented from going deeper by 

 intolerance to low DO2 concentrations, although a 

 few individuals were below this low DO2 layer and, 

 for the short term at least, were apparently sur- 

 viving. At night, there was little change in the ob- 

 served general depth preference of the main prawn 

 concentration, although more prawns were observed 

 at shallower depths. Prawns were not observed 

 moving vertically on the cliffs in a directed manner, 

 and so prawns observed at shallower depths at night 

 may have been hidden there during the day. 



With the sudden movement of low DO2 water 

 into the depth range occupied by prawns, it is 

 unknown why prawns did not simply walk upwards 

 on the cliff, away from the low DO2 area, and stay 

 in a tolerable environment, as did the pink shrimp. 

 The distance prawns would have had to travel was 

 <10 m in the 70-75 m depth range. Some vertical 

 movement of prawns may have occurred, since in 

 the afternoon dive prior to the evening dive in which 

 dead prawns were first observed, the depth range 

 in which prawns were abundant was narrow (5 m) 

 and at its shallowest depth (70-75 m). 



Two oceanographic factors apparently caused the 

 observed prawn mortality: the existence of a low 

 DO2 water mass in close proximity to the prawns 

 and some event which caused this water mass to in- 

 trude suddenly into the prawn habitat. As indicated 

 earlier, the presence of oxyclines in Saanich Inlet 

 is well documented, although the close proximity of 

 prawns to this lethal environment had not previously 

 been described. 



We offer two possible explanations that could 

 account for sudden intrusion of the anoxic layer: 

 1) change in the amplitude of oscillations of the 

 oxycline, or 2) an overall change in level of the mean 

 oxygen surfaces, perhaps related to a change in 



subsurface properties. With respect to the first, 

 Thomson et al. (in press) showed that in Saanich 

 Inlet, there are regular, peak-to-peak oscillations in 

 DO2 level of the order of 2.5 mL • L' at 100 m 

 depth. These oscillations were found to occur over 

 a period of hours, with a standard deviation and 

 range of effective vertical isopycnal displacement 

 estimated to be 2.0 and 9.6 m, respectively. Thom- 

 son et al. (in press) collected their data in April 1987, 

 when the mean DO2 level at 100 m was 4.8 mL  

 L"^ If similar oscillations of the oxycline occurred 

 during our observations, when DO2 levels were 

 much lower, then with a moderate change in oscilla- 

 tion amplitude, prawn could suddenly experience 

 lethal DO2 levels for time periods up to approx- 

 imately 6 hours. The causal mechanism generating 

 the oscillations and changes in amplitude of oscil- 

 lation of the pycnocline, and hence oxycline, is 

 currently unknown, but is probably due to internal 

 gravity waves propagating within the inlet (R. 

 Thomson^). 



The second explanation involves a rapid change 

 in average depth of the oxycline caused by changes 

 in vertical density profile of the water column. Inter- 

 mediate depth waters outside Saanich Inlet are most 

 dense in the fall, and intrusion of this denser water 

 over the sill into the inlet typically occurs at this time 

 (Pickard 1975). Such intrusions are often caused by 

 strong tidal influxes, and fluctuations in depth of 

 the pycnocline and oxycline subsequently propagate 

 down the inlet as a density intrusion (Holbrook and 

 Halpern 1982). Any intrusion has the potential of 

 suddenly altering oxygen concentrations at various 

 depths. The observation of mortality beginning in 

 late afternoon on 8 October is in agreement with 

 that expected based on the daily and hourly timing 

 of tidal action seen during the study period. 



Prawn mortality as described would thus appear 

 to be an episodic, but perhaps not an uncommon, 

 event in Saanich Inlet. It is probably a fall phenom- 

 enon, for the oceanographic reasons described above 

 and since in other years, this was when the hypoxic 

 layer was shallowest (Richards 1965; Tunnicliffe 

 1981; Burd and Brinkhurst 1984). Our observations 

 clearly demonstrate that sudden catastrophic mor- 

 tality can occur on a scale which may noticeably 

 affect species abundance in an area. If undocu- 

 mented, such episodic mass mortality may confound 

 an understanding of species population dynamics. 

 For fishermen in the area, mortality of prawn is a 

 concern and may explain why seasonal landings may 



2R. Thomson, Institute of Ocean Sciences, Sidney, B.C. V8L 4B2, 

 pers. commun. May 1987. 



606 



