378 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124. No. 2. June 2012 
depredated. Probability of nest success for nests at 
Porcupine Lake was 51% (95% Cl = 13-80%). 
Probability of nest success for all nests except those 
at Porcupine Lake was 26% (95% Cl = 16-37%). 
Nest density ranged from 0.1 to 0.6 nests/ha 
across the six study sites. Nest density at 
Porcupine Lake was exceptionally high (3.1 
nests/ha). Mean distance among nests was smaller 
in Lumbricus-f ree sites (215 m. 95% C'l = 180— 
250 m) compared to invaded sites (250 m, 95% Cl 
= 236-264 m); however, the difference was not 
significant (t =2.21, df =5 . P = 0.092). Nest 
density did not ditter between Lumbricus-free and 
invaded sites (t = 1.08, df = 5, P = 0.37). 
DISCUSSION 
I observed exceptionally high nesting density 
of Hermit Thrushes in a remnant earth worm-free 
area of northern hardwood forest in northern 
Wisconsin. The 2-year mean Hermit Thrush 
density in the highest density area (3.1 nests/ha) 
was greater than the previous maximum density 
of 74 territories/40 ha (1.9 nests/ha) in Arizona 
(Franzreb and Ohmart 1978) and more than an 
order of magnitude greater than many other 
published estimates (Table 1). 
Evidence from this study for a causal link 
between Lumbricus invasions and Hermit Thrush 
nest density is limited. 1 found no difference in 
nest density between Lumbricus- free and invaded 
sites; however, nests in Lumbricus- free sites 
averaged slightly closer to each other compared 
to nests in invaded sites. A previous study in the 
same study area found significantly reduced 
density of singing Hermit Thrushes in relation to 
Lumbricus invasions (Loss and Blair 2011). It is 
unclear why density of singing Hermit Thrushes is 
not consistently proportional to nesting density 
across invasion categories in my study area, but 
differences in rales ot pairing success and nesting 
attempts between invasion categories may have 
influenced these patterns. 
Other factors could have contributed to the high 
densities observed at Porcupine Lake. The forest 
floor in this area was characterized by cover of 
clubmoss {Lycopodium spp.) that was more 
extensive (17.4% on average) than in other 
Lwnbricus-Wce (12.1% on average) and invaded 
areas (4.3% on average) (SRL, unpubl. data). 
Clubmoss at Hermit Thrush nest sites provides 
more concealment than other nesting substrates 
(Flaspohler et al. 2000); extensive cover of 
clubmoss at Porcupine Lake may have provided 
an abundance of suitable nest sites. A similar 
conclusion was reached in a study in Arizona about 
the importance of small firs (Abies spp.) for 
providing concealment and potential nest sites for 
Hermit Thrushes (Martin and Roper 1988). Further 
study ot relationships between earthworm invasion 
and the presence and abundance of clubmoss is 
needed to clarify whether invasions have a negative 
effect on this plant group. Proximity to forest edges 
may also influence Hermit Thrush density (Flas¬ 
pohler et al. 2001); however, it is unlikely that edge 
effects were present in my study because the sites 
were >0.5 km from forest openings. 
It is possible that individual male Hermit 
Thrushes were paired to more than one female 
and nest, given the proximity of Hennit Thrush 
nests to each other at Porcupine Lake, and that 
two nests in 2009 were only 12.0 m apart. I 
confirmed that each nest was incubated by a 
different (ostensibly female) Hermit Thmsh but 
the secretive nature of males made it difficult to 
know whether each nest was attended by an 
independent pair. Three Hermit Thrushes were 
simultaneously recorded singing on several occa¬ 
sions during point counts conducted at Porcupine 
Lake (Fig. 1) in a previous study (Loss and Blair 
2011). This observation suggests there were 
sufficient males present for each nest to have 
been attended by a separate pair. 
Nest density estimates were based on all nests 
with no differentiation between first nesting 
attempts and re-nesting following abandonment 
or predation. It is possible that nest density 
estimates in Lumbricus-tree sites were inflated by 
higher predation rates resulting in more re-nesting 
attempts. Two lines of evidence from previous 
work in the same sites (Loss and Blair in 201H 
suggest that nest density differences between 
Lumbricus-free and invaded areas are real: (1) nest 
density of another ground-nesting species, tlte 
Oven bird, was higher in Lumbricus-free sites 
compared to invaded sites, and (2) singing male 
density of Ovenbirds and Hennit Thrushes was 
greater in Lumbricus-free sites. The probability of 
Hermit Thrush nest success was higher at Porcu¬ 
pine Lake compared to outside areas, suggesting 
the possibility that re-nesting attempts at Porcupine 
Lake may have actually been less common than in 
other areas. The sample size of nests at Porcupine 
Lake was small [n = 8); therefore, the nest success 
results should be interpreted cautiously. 
Earthworm-free forests have become exceeding¬ 
ly rare throughout northern Wisconsin (Holdsworth 
