542 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 123. No. 3. September 2011 
The spatial difference in midge abundance was 
more than twice as large in spring as fall (Fig. 5). 
Phenology was unrelated to midge abundance after 
controlling for influence of date for both spring (/ k 
= -1.87. n - 230, P = 0.061) and fall ( t k = 
— 1.71, n = 180, P = 0.087). Midge abundance 
appeared to be a consequence of proximity to the 
lake, not phenological development. 
DISCUSSION 
We found relatively high numbers of landbirds, 
including those that breed north of the study area, 
occurring in nearshore white cedar dominated 
forests along the northern shore of Lake Huron 
contrary to our hypothesis about the distribution 
of spring migrants. There was a similar, although 
less pronounced effect during fall migration, 
which was in accordance with our hypothesis 
about fall migrant distribution. Migrant landbirds 
concentrated within 0.4 km of the lakcshore 
during both spring and fall. Landscape features 
and associated food supply likely contributed to 
these distributional patterns. 
The northern Lake Huron shoreline includes 
forested islands and peninsulas, coastal wetlands, 
and different forest types, each of which may 
influence migrant distribution (Moore et al. 1990, 
Diehl et al. 2003) or mass gain (Dunn 2000, 
2001). We were primarily interested in assessing 
the influence of northern Lake Huron on migrant 
abundance and attempted to minimize potentially 
confounding effects by sampling the predominant 
forest type, avoiding peninsulas and islands. Wc 
also included transect as a factor within our GLM 
to statistically control for longitudinal and indi¬ 
vidual transect effects on migrant distribution. 
Our analysis revealed a significant transect effect 
for both long- and short-distance migrant distri¬ 
bution in spring and fall, which suggests land¬ 
scape features other than those associated with the 
shoreline influence migrant distribution. Our data 
do not permit us to explore this effect in more 
detail. 
Our results suggest proximity of northern Lake 
Huron influences migrant distribution even with 
the influence of uninvestigated landscape features 
on the spatial distribution of migrants. We found 
both spring and fall migrants differentially 
abundant in shoreline habitats. Concentrations of 
migrating landbirds often occur near large bodies 
of water including the Gulf of Mexico (Lowery 
1945, Gauthreaux 1971, Moore ct al. 1990), Great 
Lakes (Weir 1972, Brock 1992, Bonter et al 
2009), and the Atlantic Coast (Watts and Mabey 
1994). The northern Lake Huron shoreline may be 
attractive to northbound migrants as it is the tin.: 
landfall for birds flying over Lake Huron and lor 
fall migrants which may accumulate at the 
shoreline prior to continuing southbound migra¬ 
tion. We cannot rule out the lake causes migranb 
to accumulate in nearshore habitats, but our data 
suggest other factors, notably presence of midges 
in nearshore habitats during spring, contributed to 
migrant distributional differences. 
Migrants, including those that breed north of 
the study area, concentrated at the shoreline even 
as leaf out was delayed relative to inland areas. A 
positive correlation between stage of leafout and 
abundance of leaf-chewing insects in the same 
area (K. J. Smith and F. R. Moore, unpubl. data' 
suggests abundance of these insects was lower in 
shoreline habitats relative to inland areas, espe¬ 
cially early in the spring migratory period. 
Migrant distribution coincided with midge distri¬ 
bution; more migrants were in shoreline habitats 
during spring and fall, as were midges, and there 
were positive associations between midge abun¬ 
dance and all three categories of migrant abun 
dance in spring. 
The abundance of midges is especially marked 
in spring, suggesting foraging migrants relied on 
aquatic-dependent food resources (Smith 2003) 
disproportionately during spring migration. 
Spring migrants arrived before many leaves bad 
unfurled (Smith et al. 2007; D. N. Ewert andM. J 
Hamas, pers. obs.) during the sampling period and 
in subsequent years. Most terrestrial invertebrate 1 -, 
except spiders, were also scarce on conifers in 
shoreline habitat (Smith et al. 2007: P. L. Hudson, 
pers. comm.). We observed many species in 
shoreline habitats (e.g.. Empidonax flycatchers, 
kinglets, many vireo and warbler species, Wbite- 
throated Sparrow [Zonotrichia albicollis ]. and 
White-crowned Sparrow) feeding on midges from 
foliage, aerially, from rocks along the shoreline 
and from spider webs during our spring surveys 
(Dal I man and Smith 1995). Our results, in 
conjunction with those of Seefelt (1997) and 
Smith ct al. (1998, 2004, 2007), suggest midges 
are a critical early season resource for spring 
migrant landbirds moving through nearshore 
habitats adjacent to northern Lake Huron. 
Fall migrants were more abundant in shoreline 
habitats, as would be predicted if the moderating 
effect of the lake on nearshore habitats resulted in 
increased resource abundance relative to inland 
