Greenlaw and Post • SALTMARSH SPARROW SEXUAL BEHAVIOR 
255 
and two (1978) 2-in platforms with the aid of 
binoculars and a telescope. The platforms allowed 
us to observe birds within 25-35 m of our vantage 
point and witness local travel up to 100 m. 
Platforms were placed 125-1,025 m apart in the 
upper intertidal zone where breeding Saltmarsh 
Sparrows were numerous. No platform site was 
used in more than one breeding season. Each day 
(0500-1000 hrs) we randomly chose one or two 
platforms from which to observe sparrows; the four 
quadrants around each platform were sampled in 
random sequence (3-5 on each platform daily). 
Observation periods were 30 min in each quadrant 
sampled. A unit observation was a bird at one 
location within a given quadrant. Information on 
bird identity (band combination if known), location 
in relation to nearest gnd marker, and type and 
duration of behavior were recorded on audiotapes 
and transcribed later; unidentified individuals were 
numbered sequentially during each observation 
period. Each revisit to a quadrant by an individual 
sparrow within a 30-min period was regarded as an 
independent observation. The total number of 
identified males around each platform varied front 
28 to 40 during the season. 
We used blinds on the ground to observe 
behaviors near nests during 1979-1985. We 
followed several males with wool yarn wing 
streamers for timed periods in 1985 to examine 
patrolling behavior of males. The yarn, which 
frayed and produced it visible patch of color, was 
tied loosely around the inner wing next to the body 
and later removed. We Were able to identify birds 
directly by band combinations in -40% of out- 
observations. Sex was ascertained indirectly in 
other cases by the behavior of participants involved 
in interactions. Males sang, performed sequences 
of lookout behavior, and initiated meetings with 
females, while females uttered a specific call (Tuc 
call) when they left nests containing young or 
during some interspecific interactions with males. 
We tested the efficacy of direct and indirect 
methods of sex identification to yield similar 
Irequency estimates by comparing the frequency 
of male-male versus male-female classifications 
across the same set of observation periods by each 
method. Classifications based on behavioral iden¬ 
tification as male or female were statistically 
indistinguishable from those derived from known 
band combinations (2 x 2 contingency test with 
Yates' correction, two-sided: P = 0.88, n = 209). 
A male morning activity budget was obtained 
for the breeding season (mid-May through Jul) 
from the frequencies of observed behaviors 
obtained from the timed platform observations 
(n = 339, 30-min periods in 2 yrs, distributed in 
30-94 periods at each of 6 platforms). We 
classified activities into singing, ‘lookout’, flying, 
social (male-male or male-female) interaction, 
and unknown activities. We did not monitor 
foraging behavior during limed periods because 
we often did not know whether a bird that 
disappeared into the grass was foraging or 
performing some other behavior. Our impression 
was that males during the early morning from 
sunrise to —0900 to 1000 hrs spent most of their 
time roaming their home ranges and engaging in 
social interactions. The roaming behavior, which 
involved perched watchfulness, flying between 
perches, and singing from perches or in flight 
between perches, is termed ‘patrolling’. We saw 
little oven foraging by males until later in the day. 
Roaming males that intermittently perched silent¬ 
ly in an erect, watchful posture on exposed 
perches are described as performing ‘lookout’ 
behavior. The term ‘forced mounting’ refers to 
cases in which males jumped on females without 
any preliminary behavior and successfully sup¬ 
pressed the active aggressive resistance of a 
female, and assumed a copulatory position on 
her back while she crouched on the ground, often 
accompanied by ‘feather-pulling’ (female’s head 
leathers grasped in male's beak). Male attempts 
(whether successful or not) during male-initiated 
meetings to mount a resisting female on the ground 
and assume a mating position are categorized as 
‘pounces'. These terms are descriptive of male 
behavior that we observed- The term ‘forced 
mating’, which implies successful or potentially 
successful insemination of an uncooperative fe¬ 
male following forced mounts, is avoided until the 
concept is evaluated later in relation to behavior of 
males during pounce events. Female control in this 
study involved females successfully thwarting 
forced mounting attempts of some males, either 
by threatening them or by fending off their efforts 
to mount. Males that were repelled gave up and left 
the area. We often did not know which males 
successfully suppressed female aggression, or 
which ones were thwarted, because we could not 
see band combinations during fast-moving or 
partly obscured interactions. Solicited copulations 
were uncontested and initiated by females at the 
beginning of her nest cycle, and invoked a pre- 
copulatory visual display typical of emberizids 
(Andrew 1956). 
