(lASKIN ami WATSON: HARBOR POKCOISK 



Table 1. — Sightings of harbor porpoises per hour in the Simpson's Passage and Fish Harbour areas, southern New Brunswick, 



1970-78. 



ber and May, but probably rarely enter Fish Harbour 

 (see next section). 



Arrival and Departure of Porpoises 



Each Year in Relation to 



Sea Temperatures 



First sightings in Fish Harbour were usually made 

 in mid-late July when surface temperatures attained 

 about 9°C, and never in Lords Cove until about mid- 

 August despite 9°-ll°C being reached at the surface 

 in mid-late July. First regular sightings in Simpson's 

 Passage varied from mid-May to late June. Deter- 

 mining the date of departure of the majority of 

 animals from either sector was difficult because 

 strong autumnal winds invariably interfered with 

 observations from mid-September onwards when 

 relative abundance was still high. In both 1977 and 

 1978, porpoises were still present in Fish Harbour 

 until the last week of September and in Simpson's 

 Passage until at least 15 October. Occasional animals 

 may venture into the latter area in any month of the 

 year since a very small population usually over- 

 winters in the Quoddy region (Gaskin 1977), and one 

 animal was sighted outside Fish Island on 7 Decem- 

 ber 1982 (B. M. Braune^ and D. E. Gaskin, pers. 

 obs.). 



We could find no evidence that the distribution of 

 porpoises was directly or indirectly influenced by the 

 rather small daily local variations in sea tempera- 

 tures within Fish Harbour. The most frequently 



^B. M. Braune, Department of Zoology, University of Guelph, 

 Guelph, Ontario NIG 2W1, Canada. 



observed known animal and her consorts would 

 regularly traverse the width of the area (see Figure. 

 9) and their preferred locations appeared to have 

 specific topographic rather than temperature charac- 

 teristics. 



Estimating Relative Abundance 



Because the species does not make long dives 

 (mean submergence 1 min 44 s, Watson and Gaskin 

 (1983)), the required minimum period of observation 

 needed to search the study area was not excessive. 

 From our records we selected 3 wk in August 

 1972-75, when the probability of animals being pres- 

 ent was high. In a random sample of 40 (i.e., above 

 the minimum size for a "large" statistical sample 

 (Bailey 1959)) search periods in optimum conditions 

 of varying length (5 min to 2 h), the percentage of 

 time that one or more animals was recorded in- 

 creased from 50% for 10-min periods to well over 

 80% for 15-min periods. All observations of < 15 min 

 were therefore discarded from the data set. If we 

 were only interested in presence or absence, as in the 

 case of simple locations at a given time of day or tide, 

 observations from shorter periods or in Beaufort 

 wind force 2 + were still of some value. 



We have already outlined the methods for obtain- 

 ing our "best estimates"; it is worth noting that 

 various characteristics of the animals (e.g., short dive 

 times, stereotyped movements, recognizable individ- 

 uals) and of the study area (limited search area 

 because of shallow water, many landmarks, shelter, 

 and limited entry and exit points for animals) were of 

 great assistance in reducing repeat sightings to a 

 minimum. 



431 



