Ewert et al. • DISTRIBUTION OF MIGRANT LANDBIRDS ALONG LAKE HURON 537 
1986), principally through temperature depression 
in spring and thermal retention and consequent 
warming of these same habitats in fall. These 
effects may influence phenological development 
of vegetation (Ewert and Hamas 1995), as well as 
the emergence, development, and activity of 
potentially important invertebrate prey species. 
We originally hypothesized that (1) migrants 
would avoid the shoreline during spring because 
the cooler nearshore microclimate delays phenol¬ 
ogy of vegetation (Dunn 2001; D. N. Ewert and 
M. J. Hamas, pers. obs.) and appearance of 
terrestrial arthropods (Bonter et al. 2007), and 
i2) migrants would preferentially select shoreline 
areas in autumn as the moderating effect of such a 
large body of water delays early frosts, allowing 
more resources, especially arthropods, to persist 
late into the fall migration period (Bonter et al. 
2009). 
Anecdotal evidence (White 1893. Peet 1908, 
Wetmore 1927, Ewert and Hamas 1996) and 
NEXRAD data (Bonter et al. 2009) suggests 
migrants concentrate along Great Lakes shoreline 
habitats during both spring and fall but we know 
little about the relative abundance of migrants 
along the shoreline and adjacent inland land¬ 
scapes. No work has rigorously examined the 
influence of the northern Great Lakes on distri¬ 
bution of landbirds within the same landscape 
during both spring and fall migration. Our 
objective was to describe migrant landbird 
distribution along and immediately inland from 
the shoreline of northern Lake Huron in an effort 
to evaluate if proximity to northern Lake Huron, 
within a particular habitat type, influences the 
abundance and distribution of birds during spring 
and fall migration. 
METHODS 
Study Site .—The study site encompassed 80 km 
°f northern Lake Huron shoreline extending from 
St. Martin’s Bay, Mackinac County east to 
DeTour Village, Chippewa County. Michigan 
145 57" to 46 04" N. 84 00" W to 84’ 44" W: 
1). Limestone and dolomite peninsulas define 
m uch of the shoreline with sandy coves and 
raarshes occupying borders of intervening bays. 
Small inland lakes and streams are scattered 
throughout the study area. Conifers, especially 
northern while cedar (Thuja occidentalis), balsam 
hr (Abies balsamea). white spruce (Picea gUtuca). 
and white pine (Pinus strobus), and deciduous 
species including paper birch ( Betula papyrifera). 
and quaking aspen ( Populus tremuloides) domi¬ 
nate the forest. Canopy height typically ranges 
from 10 to 15 m. The understory, which is heavily 
browsed by white-tailed deer (Odocoileus virgi- 
nianus), was sparse, consisting mostly of balsam 
fir with lesser amounts of white spruce, quaking 
aspen, striped maple (Acer pensylvanicum), red 
osier dogwood (Coriius stolonifera), and beaked 
hazelnut ( Corylus corvuta). 
Data Collection. —We established 45 perma¬ 
nent 50-m fixed-radius point-count stations along 
nine transects to count birds. Point-count stations 
were placed at 3.2, 1.6, 0.8. and 0.4 km from the 
shoreline and al the forest edge at the shoreline. 
Stations were aligned in transects perpendicular to 
the shoreline; we did not place stations on 
peninsulas to minimize potential biases associated 
with areas where migrants may concentrate. The 
center of each station was at least 50 m from the 
edge of secondary roads or open shoreline. We 
established stations in forest dominated by white 
cedar, balsam fir, paper birch, and quaking aspen 
to control for forest type. Appropriate stations at 
3.2 km were not available in three cases due to 
inaccessibility, and we established stations at 
other nearby sites in appropriate habitat 3.2 km 
inland from the shoreline. 
We recorded all birds detected by sight or 
sound al each station during 5-min count periods. 
We used 5- rather than 10-min counts to sample 
more sites and to ensure that individuals in rapidly 
moving flocks were counted only once. We 
conducted counts between sunrise and 1200 hrs 
EDT from 1 May to 5 June 1993 and 30 April to 5 
June 1994, and from 16 August to 24 September 
1993 and 15 August to 23 September 1994. Only a 
few early spring and late fall migrants, such as 
kinglets. Fox Sparrow (Passerella iliaca), and 
some icterids and eardueline finches largely 
migrated outside the sampling periods (D. N. 
Ewert, M. J. Hamas, and R. J. Smith, unpubl. 
data). We sampled all 45 points on each of the 10 
count days each season. We surveyed birds across 
the entire study area every 3 to 5 days, dependent 
upon the weather. We did not conduct surveys 
during rain or high wind events. We assigned each 
observer a group of adjacent point count stations 
(to minimize travel and maximize sampling 
effort) the night prior to each survey. All 
observers sampled their assigned stations on the 
same day and no observer counted birds at the 
same station on two consecutive sampling dates. 
We classified each bird species as a resident, or 
