776 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 124. No. 4. December 2012 
Relatively few items were larger than SC 2 and 
were categorized as >SC 2. We estimated biomass 
ol each prey item. Biomass refers to live mass of an 
item, not the mass delivered to the nest or 
consumed by nestlings. We removed the data 
collected at nests on or within 0.5 km of the 
University of Victoria property (/? = 31 items) 
from biomass estimates due to a site bias; nearly all 
(8 of 9) European rabbits (Oryctolayus cuniculus) 
were recorded at these few nest sites. Each rabbit 
provided about live times the biomass of a single 
avian item, inflating the biomass proportion 
contributed by mammalian items and inaccurately 
representing the extent of mammal consumption at 
the majority of nests in Victoria. 
We identified 189 (46%) of 408 avian prey 
items to species. Adult masses of these items were 
from Dunning (2008). The age of avian prey 
items, when possible, was identified from feather 
sheathing in nestlings and recent fledglings 
distinctive Juvenal plumages, or the color of the' 
tarsi for some species. We estimated mass of 
avian items identified by species and age from the 
.teuuure 0 " growth rates (Clench and Ueberman 
1978). Most species in SC 1 fledge at >90% 0 f 
aduU mass and we ignored age in these cases. 
Altncial, open-nesting species in SC 2 were 
divided into two categories of smaller (28-59 
and large, (fiU-Sig, species. We calculated mass 
at 70. 90, and 95% of adult mass lor smaller 
species al nestling, juvenal. and unknown ages 
respectively; and for larger species al 65, 85, and 
90%. respectively. Juvenal and unidentifiable 
Items larger than SC 2 were calculated a, 90% 
l a ““ 10 If c °nservalive in our estimates. 
We used assumed averages of 25 g and 65 g for 
unidentified birds in SC I and SC 2. respectively 
Only one unidentified partial item was larger than 
SC and was not assigned a mass. Other 
unidentified items that could not he assigned a 
sue class were not included in mass estiuiLs 
We .denufied 27 of 29 mammalian items (91%) 
to genus or species: mosl were Eastern cottontail 
(SyMagus Honda,,"*) or European rabbits, and 
orway rats \Raltus norveeicus) \ mom 
estimate of 350 g for the two rabbi. spec"s 
obtained from measurements of semiTntact prev 
remains ail rabbits were less than or about half 
grown. We estimated adult mass of Norway and 
black, lats [ R ‘ rattu «) from Nowak ( 1999 )- 
juveniles and unidentified individuals were esli' 
mated a, 90 % of adult mass to be conservative' 
Rats only identifted to genus were assumed to be 
less than or about half-grown Norway rats based 
on general observations and we estimated biomass 
of each at 220 g. Adult masses of one deer mouse 
(Peromyscus maniculatus) and one Eastern gray 
squirrel (Sciurns carolinensis) were estimated 
from Wilson and Ruff (1999). 
Drey Remains. -We tabulated remains of 3.231 
prey items from 44 active nest sites during the 
pre-incubation through fledging stages in 1995- 
1998. The items were mainly pluckings. and legs, 
tails, or partial carcasses. We identified prey to 
the lowest taxonomic class by comparing remains 
to a reference collection of locally common prey- 
si/.ed species and the Royal British Columbia 
Museum's vertebrate collection. We included 
items from reoccupied nest sites hut treat the data 
as independent because no single nest site 
contributed >7% of the total items over the 4 
study years. We combined all stages of breeding 
(pie-incubation through fledgling) and pooled 
data from all years because proportions of avian 
prey were similar (94-98%). Avian items identi¬ 
fied to species were categorized, based on adult 
masses (Dunning 2008), into the same size classes 
Irom Siorer (1966), Kennedy and Johnson (1986), 
and Bielefeld! et al. (1992). We classified prey 
items as native or introduced to the study area 
when possible. 
Video Analysis.— We used video cameras to 
record prey deliveries in different years to two 
arbitrarily selected nests >2.5 kin from the 
University of Victoria property. The custom-made 
camera fed into two VHS video recorders and an 
Axis brand video server that recorded a still image 
every minute. We placed cameras at an oblique 
angle. 80 to 1 00 cm from the center of the nest 
platform. The Torquay nest was recorded from 2 
June through 21 July 2001. Recording start time 
varied between 0800 and 0930 hrs from 2 June to 
4 June, 0430 hrs during 6-15 June, and 0500 hrs 
t e rest of the days. Recording continued until 
T “ hrs - w hen the videotape ran out each day. 
vn ol the 50 days had incomplete recordings. 
I he Burnside nest was recorded from 4 June 
lb rough 5 July in 2002. We recorded from 0500 to 
- 00 hrs each day (excluding 6 incomplete 
ays). Complete days for both nests averaged 
6 his of footage and incomplete days ranged 
Iroiri 6 to 12 hrs of footage. We obtained 750 and 
467 hrs of video from the Torquay and Burnside 
nests, respectively. 
identified prey to the lowest taxonomic 
eve possible, age, and categorized them into size 
