Suryan and Harvey: Variability in reaction to disturbance among Phoca vitulina nchardsi 



335 



ANOVA with square root transformation. Initial ver- 

 sus subsequent disturbances were compared within 

 each disturbance categories with ^tests. Randomiza- 

 tion statistics ( Resamphng Stats, Inc., 1995 ) were used 

 in calculation of power to detect significant differences. 



Results 



Clements Reef was the largest haul-out site with the 

 greatest number of seals. Puffin Island, however, had 

 the greatest number of pups, the pups representing 

 a significantly (F=97.6, P<0.01) greater percentage 

 of the total number of seals ashore in comparison 

 with Clements Reef (Table 1 ).The study area on Puf- 

 fin Island appeared to be an important area for fe- 

 male-and-pup pairs and thus may explain some of 

 the observed differences in reactions to disturbance. 

 Harassments of seals occurred during at least 71'7f 

 of the survey days. The primary cause of disturbance 

 was powerboats (Table 1). Most of these (749^, n=96) 

 involved boat operators approaching haul-out sites 

 to view seals. The second most common category of 

 disturbance was unknown (no source was visible or 

 audible to the observer) and was most prevalent at 

 Skipjack Island. Bald eagles were the third most com- 

 mon cause of disturbance and were more frequent at 

 Puffin Island than at other sites. In contrast with 

 powerboats, sea kayaks were uncommon during the 

 study and, therefore, caused fewer harassments ( 1 1% 

 for all sites). Sea kayakers, however, were a greater 

 potential disturbance to seals ashore than operators 

 of powerboats. Fifty-five percent of kayakers {n=ll) 

 within 1 km of a haul-out site harassed harbor seals, 

 whereas only 9'7c of powerboats («=436) within 1 km 

 caused harassment. This finding is a result of seals 

 being less tolerant of kayaks and kayakers who gen- 

 erally travel closer to shore than do powerboaters. 



The extent of recovery following harassments (ex- 

 cluding those caused by eagles) of harbor seals was 

 less at Puffin Island than at Clements Reef and Skip- 

 jack Island (Table 2). Disturbances caused by eagles 

 were excluded from this comparison because eagles 

 were potential predators of pups (eagles were ob- 

 served approaching and disturbing pups) and could 

 affect recovery. Recovery was related to whether the 

 harassment occurred before or after low tide (74% of 

 full recoveries occurred before low tide, whereas 75% 

 of partial and 89% of no recoveries occurred after 

 low tide). This finding corresponds with total num- 

 bers of seals ashore typically decreasing one to two 

 hours after low tide, independent of disturbance. For 

 comparison of recovery among sites, the proportion of 

 disturbances occurring before and after low tide was 

 similar and likely did not influence results (Table 2). 



