Strong and others 



Chapter 32 



Distribution and Population Estimates in Oregon 



stay diving dove flew flying split 

 BEHAVIOR 



1992 (n = 4721) 



1993(n = 3291) 



Figure 7 Behavior of murrelet groups observed while on transect: stay, remained on 

 the water surface; diving, engaged in diving activity; dove, dove in apparent avoidance 

 of boat; flew, flew in apparent response to the boat; flying, flying past when detected; 

 split, group separated in apparent avoidance of the boat (in all other instances group 

 members behaved the same). 



reliability of detection due to proximity, duration of 

 observation, and number of observers. 



The average density of birds seen from air was 33.1 

 percent (1992) and 16.3 percent (1993) of that seen by boat 

 strip transects (table 2), even though they transected the 

 same offshore zone at a similar time of year. The brief 

 scanning time when flying over the transect strip at 90 knots 

 may be the greatest factor affecting detection rates by air. 

 Slight variations in plane altitude and speed, banking on 

 turns, observers checking time and location, and distraction 

 from other species all contributed to further reduce scanning 

 time for murrelets. In addition, on the 1 July 1993 flights, the 

 senior author scanned an area in advance of the plane and 

 noted Marbled Murrelets diving in response to the plane's 

 approach. The extent of this behavior cannot be quantified 

 absolutely, and probably varies with type of plane. On the 1 

 July 1993 north bound survey, at least nine birds dove in 

 front of the plane (8 percent). 



Estimates using counts from shore were in the same 

 general range of those based on aerial surveys, though there 

 was no consistency across years (table 1). Shore counts had 

 the highest variability in numbers with coefficients of 

 variation averaging over 100 (table 3, fig. 4). The high 

 variability resulted from Marbled Murrelets' locally patchy 

 and shifting distribution (Nelson and Hardin 1993b, Strong 

 and others 1993). Low average numbers seen could also be 

 due to their patchy distribution. Difficulty in detecting birds 

 from a low, distant vantage point under variable conditions 

 may also have reduced the number of detections in some 

 cases. Even though we compensated for difficulty in detection 



by halving the calculated area scanned when computing 

 densities, values were still far lower than from the vessel. 

 These results may have occurred because the smallest effort 

 was invested in shore surveys. Increased effort may have 

 reduced variability and improved results. Weighting of high 

 counts in proportion with the patchiness of high density 

 areas could possibly generate average densities more 

 representative of the population. 



Strip and Line Transect Vessel Estimates 



Line transects generated the highest estimates, and they 

 were consistently higher than strip transect estimates using 

 the same data. Strip transect estimates were between 60 

 percent and 88 percent of line estimates, but the difference 

 was only marginally significant in one case (center region, 

 1992, r-test, P = 0.023) and not significant in others where 

 sample size was sufficient. 



Strip transects may be conservative if the assumption 

 that all birds within the strip are detected is not met. This 

 was apparently the case when the strip was 130 m and 

 greater distances from the vessel (fig. 2). Estimates using 

 a 130 m strip width were 90.6 and 92.8 percent of those 

 for a 100 m strip for 1992 and 1993, respectively. We 

 interpreted this as indicating that 7 to 10 percent of the 

 birds were not detected with the larger strip width. The 

 strip width of 80 m resulted in even higher estimates, but 

 1 1.5 percent and 8.7 percent of the birds had avoided the 

 vessel beyond this strip width in 1992 and 1993, 

 respectively (compared with 6.9 percent and 5.2 percent 

 for a 100 m strip). The selection of a 100-m strip, then, 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



347 



