Harlequins (Crowley 1994). 



Substrate. In Idaho and Oregon, Harlequins were typically associated with cobble and 

 boulder substrates (Thompson et al. 1993, Cassirer and Groves 1994); there was no difference 

 between substrate chosen by adults and broods (Cassirer and Groves 1994). Substrates on the 

 Rocky Mountain Front of Montana were primarily cobble (62%), boulder (17%), or bedrock 

 (10%) (Diamond and Finnegan 1993). While most Harlequins were observed on gravel and 

 cobble substrates in Grand Teton National Park, use of those substrates was not significantly 

 different than their availability (Wallen 1987). 



Woody material was present at 85% of duck sightings in Oregon, with the most structures 

 being ramps (42%) and drifts (25%) (Thompson et al. 1993). Similarly in Idaho 77% of sightings 

 were near woody debris (Cassirer and Groves 1991). On the Rocky Mountain Front, however, 

 less than 40% of the observation locations contained woody debris and 19% contained only a 

 single piece of woody debris within 10 m (Diamond and Finnegan 1993). 



Loafing sites Loaf sites, places where Harlequins will haul out of the water to rest, are 

 consistently mentioned in the literature and unpublished reports as being important components of 

 Harlequin Duck stream habitat. In Grand Teton National Park, Wyoming, Harlequins 

 significantly selected for sites with >3 loaf sites per m, and avoided areas with no loaf sites 

 (Wallen 1987). On the Rocky Mountain Front, over 70% of Harlequin observations were m areas 

 with 2 or more loaf sites per 10 m (Diamond and Finnegan 1993). Harlequins in Idaho were 

 within 10 m of multiple loaf sites 83% of observations for adults and 94% for juveniles (Cassirer 

 and Groves 1994). Moss covered midstream boulders were used most often in Oregon, with an 

 average of 5 loaf sites per 10 m at observation sights (Thompson et al 1993, 1993). 



In Iceland, Harlequins spent 47% of their time on larger islands, only 24% on rocks 

 protruding above the river, and <1% on the banks of the river (Inglis et al. 1989) 



Quality Streams used by Hariequins in northern Idaho, Montana, and Wyoming (n = 11) 

 were more alkaline than unused streams (x CaCOs = 58 m/1, x CaCO? = 8 m/1, p = 0.004) 

 (Cassirer and Groves 1994). 



There are two records of broods using glacial streams or lakes in British Columbia (Breault 

 andSavard 1991). 



Quantity. Volume discharge was the most important factor separating used from unused 

 streams in Pnnce William Sound, Alaska, volume discharge averaged 3.2 m/s on used streams 

 (Crowley 1994). The largest streams in Prince William Sound were not used by breeding 

 Harlequins, smaller salmon streams were also avoided by nesting females (Crowley 1994). See 

 also Causes of death. 



Food availability. Streams used by Harlequins averaged a higher standing crop of benthic 

 macroinvertebrate biomass in Idaho (x = 0.52 g/m^) than unused streams (x = 0.34 g/m^) (Cassirer 

 and Groves 1994); there was considerable overlap, however, and the differences were not 

 significant. Bengtson (1966) believed that the distribution of Hariequins in Iceland was 

 determined by the availability of suitable food, particularly Simuliidae larvae. In Iceland, 

 Bengtson and Ulfstrand (1971) reported high numbers of non-breeding females on the breeding 

 streams in 1970; this was correlated with a far below normal standing crop of favored invertebrate 

 food species. More research revealed that numbers of Harlequins produced from 1977-85 were 

 correlated with biomass of blackflies {Simulium vittatum) present (Gislason and Gardarsson 1988, 



