A Comparative Study of the Flora and Soils of Great Duck and Little Duck Islands, Maine, USA

Abstract

In Press

Full text

RHODORA, Vol. 118, No. 973, pp. 46–85, 2016
ÓCopyright 2016 by the New England Botanical Club
doi: 10.3119/15-11; first published on-line April 20, 2016.
A COMPARATIVE STUDY OF THE FLORA AND SOILS OF
GREAT DUCK AND LITTLE DUCK ISLANDS,
MAINE, USA
LUKA NEGOITA
Syracuse University, Life Sciences Complex, 107 College Place,
Syracuse, NY 13244
e-mail: lnegoita@syr.edu
MATTHEW DICKINSON
P.O. Box 571, Cape Neddick, ME 03902
GLEN H. MITTELHAUSER
Maine Natural History Observatory, 317 Guzzle Road, Gouldsboro, ME 04607
NISHANTA RAJAKARUNA
College of the Atlantic, 105 Eden Street, Bar Harbor, ME 04609; Unit for
Environmental Sciences and Management, North-West University, Private Bag
X6001, Potchefstroom, 2520, South Africa
ABSTRACT. Strong environmental gradients and varied land-use practices
have generated a mosaic of habitats harboring distinct plant communities on
islands on the coast of Maine. Botanical studies of Maine’s islands, however,
are generally limited in number and scope. Baseline studies of Maine’s islands
are necessary for assessing vegetation dynamics and changes in habitat
conditions in relation to environmental impacts imposed by climate change,
rising sea levels, invasive species, pests and pathogens, introduced herbivores,
and human disturbance. We conducted a survey of the vascular plants and soils
of forest, field, and ocean-side communities of Great Duck and Little Duck
Islands, ME. These islands differ in environmental and land-use features, and in
particular the presence of mammalian herbivores; Great Duck Island has had
over a century of continuous mammalian herbivory while Little Duck Island
has been largely free of mammalian herbivores over the last 100 years. We
recorded 235 vascular plant species in 61 families on the Duck Islands, 106 of
which were common to both islands. The composition, abundances, and
diversity of plant species substantially differed within similar plant communities
between the islands. These differences were particularly evident in the forest
communities where Little Duck Island had significantly greater sapling
regeneration and a more recent peak in tree recruitment. Soil properties also
significantly differed between these islands, with a higher pH in all three
communities and higher P, Ca, and K in field, forest, and ocean-side
communities, respectively, on Little Duck Island, and higher soluble salts in
forest and ocean-side communities of Great Duck Island. Together, our
findings suggest that soil characteristics and the dominance and regeneration of
46

vascular plant species can differ substantially even between adjacent islands
with otherwise similar geologic characteristics and glacial history, and that
mammalian herbivory along with other ecological factors may be important
drivers of these differences.
Key Words: coastal ecology, insular ecology, baseline survey, mammalian
herbivory, Gulf of Maine, vascular plants, edaphic features
The coastline of northeastern North America includes a mosaic of
islands with varied topography, climate, bedrock and surface geology,
and with complex postglacial and post-settlement histories (McMaster
2005; Turcotte and Butler 2006). The strong environmental gradients,
along with historical factors, have produced varied habitats harboring
a wide range of species and vegetation types on islands of the region
(Clayden et al. 2010; Greene et al. 2005). The state of Maine, located at
the intersection of temperate and boreal bioclimatic zones in
northeastern North America, is home to 2103 vascular plant taxa
(Campbell et al. 1995) and 104 natural plant communities (Gawler and
Cutko 2010). Although more than 3000 islands hug the coastline of
Maine, botanical studies of the state’s islands are limited in number and
scope, with the vast majority focusing on floristics (Folger and Wayne
1986; Greene et al. 2005; Lesser 1977; Lewis 1983; Mulligan 1980; Pike
and Hodgdon 1962; Rand 1900; Rappaport and Wesley 1985; Redfield
1885, 1893; Stebbins 1929; Wise 1970), and a few on plant-habitat
relations (Ellis et al. 2006, 2011; Hodgdon and Pike 1969; Nichols and
Nichols 2008; Rajakaruna et al. 2009; Wherry 1926). Long-term and
systematic floristic studies on Maine’s islands are necessary for
assessing long-term vegetation dynamics, including changes in habitat
conditions, especially in light of significant environmental impacts
imposed by climate change, rising sea levels, invasive species, pests and
pathogens, introduced herbivores, and human disturbances on island
ecosystems (Caujap ´
e-Castells et al. 2010; Harris et al. 2012).
The Great Duck and Little Duck Islands, ME, provide an important
setting for establishing a baseline ecological study to assess long-term
changes to plant diversity, community composition, and habitat
conditions. The environmental factors that differ between these islands
may offer some insights into the current and future ecologies of these
islands. For example, Great Duck Island has had over a century of
continuous mammalian herbivory in the form of sheep and hares,
whereas Little Duck Island has been largely free of mammalian
herbivores. Introduced mammalian herbivores are a major conserva-
tion concern, especially in predator-free habitat fragments such as
islands where herbivores can severely limit tree recruitment (Peterson et
2016] 47
Negoita et al.—Flora of Duck Islands, Maine

al. 2005; Terborgh et al. 2001). These islands also differ in topography
and other aspects of land-use history, which may in turn drive plant
communities and edaphic features. In this study, we conducted a survey
to describe the vascular flora and associated soils of the islands’ natural
plant communities. We compared species composition and diversity,
tree demography, and sapling regeneration to assess the potential
impact of long-term herbivory and other environmental and human
factors on plant communities of the two islands.
MATERIALS AND METHODS
Site description. The Duck Islands, ME (44.168N, 68.258W) are
composed of Little Duck Island (LDI, 35 ha) and Great Duck Island
(GDI, 91 ha), located about eight kilometers south of Mount Desert
Island in the Gulf of Maine (Figure 1). The climate of the region is
typically characterized by cool summers and mild winters (McMahon
1990). Average high and low annual temperatures for the nearby town
of Mount Desert are 11.9 and 0.18C, respectively, with an average
annual precipitation of 114.7 cm (between 1981 and 2010; US Climate
Data 2015).
The Duck Islands have similar bedrock geology and glacial history
(Osberg et al. 1985), with parent material typically composed of coarse
Figure 1. Study area, showing location of Great Duck and Little Duck
Islands, Maine, USA.
48 [Vol. 118
Rhodora

acidic glacial till (Jordan 1988). Parent material varies in depth, with
occasional exposed bedrock, especially near the ocean. The islands were
connected to the mainland for about 500 y following glacial retreat
around 11,000 YBP; sea-level has increased since, reaching close to
current levels around 4000 YBP (Barnhardt et al. 1995). The islands are
partially covered by forests dominated by Picea spp. and Abies
balsamea and old-fields dominated by a variety of forbs, shrubs, and
graminoids. Each island also harbors a saline wetland with a different
assemblage of species. Redfield (1885, 1893), Rand (1900), Lesser
(1977), Rappaport and Wesley (1985), and Folger and Wayne (1986)
offered preliminary accounts of vascular plants of the Duck Islands.
These islands differ substantially in their history of recent human
use, particularly regarding mammalian herbivory. Although both
islands have been under conservation protection since the 1970s, LDI
has been protected as a bird sanctuary since 1908 (McLane 1989). The
last record of sheep grazing on LDI—a common practice on Maine
islands (Conkling 2011)—was in the late 19th century, and no
permanent populations of grazing mammals have been reported on
LDI in the past 100 y (McLane 1989). Great Duck Island, however,
had a history of permanent human habitation and sheep grazing until
1986, when the lighthouse at the south end of the island became
automated and lighthouse keepers were no longer needed. In addition,
both European hare (Lepus europaeus) and snowshoe hare (L.
americanus) were introduced to GDI in the late 1940s for recreational
hunting. The hare populations have since expanded and were estimated
at 500 individuals or, about 6 hares per ha in 1985 (Folger and Wayne
1986). We made numerous sightings during our fieldwork in 2011,
indicating the hares on GDI were still abundant. This is in contrast to
LDI, where we observed no signs of any mammalian herbivores in
2010. Both GDI and LDI also harbor large populations of nesting
seabirds, including black guillemot (Cepphus grille), common eider
(Somateria mollissima), double-crested cormorant (Phalacrocorax
auritus), great cormorant (P. carbo), great black-backed gull (Larus
marinus), and herring gull (L. argentatus), as well as the threatened
Leach’s storm-petrel (Oceanodroma leucorhoa; Allen et al. 2012).
Vegetation survey. We established and surveyed sixty 20 m
2
plots
on LDI during June–August of 2010 and on GDI during June–August
of 2011. Plots (10 32 m) were randomly located within three strata:
communities dominated by a woody canopy (forest), mostly herba-
ceous vegetation without a woody canopy (field), and the vegetation
found within proximity of the ocean (excluding the rocky berm; ocean-
side). These were the most conspicuous natural communities found on
2016] 49
Negoita et al.—Flora of Duck Islands, Maine

both islands. Prior to fieldwork, we delineated these vegetation
communities from aerial photographs (North American Proficiency
Testing Program 2009) and later refined the maps based on field
observations. The forest community was defined by the presence of a
woody canopy at least 2 m in height and at least 10 m from the rocky
shoreline. The field community was defined by the presence of herbs,
the absence of a woody canopy greater than 2 m in height, and by a
distance of at least 10 m from the rocky shoreline. The ocean-side
community was defined as the vegetation within 10 m of the rocky
shoreline. To randomly select plot locations in the forest and field
vegetation communities, we first generated a geo-referenced map of the
island and overlaid a series of plots based on a 0.40 ha grid. Plots were
excluded from consideration if they occurred within community
transition zones on our delineated maps. Using a GPS unit (Garmin
eTrex Venture HC, Olathe, KS), we navigated to each randomly
selected plot within each stratum. Based on our field interpretations,
plots that occurred at the edges of community transition zones were
moved away from the transition edges. These plots were moved 30 m in
the cardinal direction (north, east, south, or west) that put them
farthest into the community they represented. The long sides of all plots
were oriented north-south. Ocean-side plots were systematically placed
around the perimeter of each island. These ocean-side plots were
oriented perpendicular to the shoreline, beginning at the first
occurrence of 100% vegetation cover from the rocky shore. In total,
we established 60 plots on LDI (29 forest, 20 field, 11 ocean-side), and
60 plots on GDI (29 forest, 19 field, 12 ocean-side).
We subdivided each 10 32 m plot into five 2 32 m subplots to allow
for easier estimation of percent cover of each vascular plant species below
a height of 2 m. Percent cover was estimated for each species within each
subplot to the nearest one percent. Species represented by less than one
percent were recorded as 0.5%, and species represented by one or only a
few seedlings were recorded as 0.1%. Percent cover data from the five
subplots were averaged to represent each plot. All individuals .5cm
diameter at breast height (DBH) in each plot were counted and cored at
breast height, and the age of each tree was estimated in the field by
counting rings with a loupe and adding 10 to conservatively account for
age at breast height. Saplings (,5cmDBH,and.than 20 cm height) of
each tree species were counted within each plot.
Although the plot surveys provided a measure of species abundances
on the islands, this method is likely to miss rare plants. Our plot surveys
also did not include transitional habitats such as between forest and
field, saline wetlands, or the rocky berm where some species were
50 [Vol. 118
Rhodora

exclusively found. Thus, in addition to our plot surveys we traversed the
islands, spending time in other habitats in order to generate complete
species lists for each island. Plants within and outside of plots were
identified in the field to the species level, if possible, or collected and
identified in the lab with a dissecting microscope and taxonomic key
(Haines 2011). Several taxa were identified only to the genus level due to
missing reproductive structures necessary for identification. Infraspecific
taxa were not considered in our study. All vascular nomenclature
follows Haines (2011). A complete list of vascular taxa for each island is
presented in the Appendix. Voucher specimens have been deposited at
the herbarium of College of the Atlantic, Bar Harbor, ME (HCOA). We
compared the list of vascular plant species tallied on the Duck Islands to
those of historic surveys of these islands (Folger and Wayne 1986; Lesser
1977; Rappaport and Wesley 1985; Redfield 1885, 1893) as a preliminary
assessment of historic change in the species composition of these islands.
Soil analyses. Soil samples were collected from the top 10 cm of
mineral soil, in two opposing corners of each 10 32 m plot, and these
were combined to form a single 200 g sample for each plot. Samples
were air dried in the laboratory and then sent to A&L Western
Laboratories Inc. (Modesto, CA) where they were tested for nitrogen
(N, NO
3
-), phosphorus (P, Weak-Bray), potassium (K), calcium (Ca),
magnesium (Mg), sodium (Na), sulfur (S, SO
4
-), pH, percent organic
matter (OM), estimated nitrogen release (ENR), soluble salts (SS), and
cation exchange capacity (CEC). All soil testing procedures followed
the Soil and Plant Analytical Methods of the North American
Proficiency Testing Program (NAPT 2011).
Statistical analyses. Data were analyzed to describe and compare
species diversity, edaphic features, and woody regeneration between
islands within similar vegetation communities. All statistical analyses
were conducted using the R language and environment for statistical
computing (R Core Team 2014).
We compared soil features between the two islands using two methods.
First, we conducted a principal components analysis (PCA) as a way to
visually inspect soil differences in multivariate space and to extract the
soil features most important for describing this variation. The PCA was
calculated using soil data that were log-transformed to aid with
assumptions of normality. For each community, the two primary
axes—those that explained the most multivariate variation—were
plotted and were labeled with the soil features important for driving
this variation (features with loadings greater than 0.3 were included).
Second, we tested for differences in soil features within communities and
2016] 51
Negoita et al.—Flora of Duck Islands, Maine

between islands, using two-tailed Mann-Whitney tests for each
comparison. The Mann-Whitney test is a non-parametric analog to the
t-test, appropriate for comparisons of non-normally distributed data
such as our soil data. We adjusted p-values using the Hochberg method
to reduce the chance of Type I errors in repeated testing, generating more
conservative comparisons (Hochberg 1988).
Finally, we compared the diversity, composition, abundance, and
regeneration of the common species on each island. Understory species
composition (species with vegetation cover within 2 m from the
ground) was assessed using plot mean percent cover data in each
vegetation type. Forest overstory composition was assessed using stem
counts of each species (individuals .5 cm DBH). We also compared
woody regeneration between islands by qualitatively comparing tree
demography. Estimates of tree ages were binned by decade, and plotted
as a density histogram where all bins for an island sum to one. We
tested for differences in sapling numbers (,5 cm DBH, and .than 20
cm height) in forests between islands and accounted for differences in
soil features by using negative binomial models of the form:
Model 1. Sapling count
plot
~soil PCA1
plot
þsoil PCA2
plot
Model 2. Sapling count
plot
~soil PCA1
plot
þsoil PCA2
plot
þisland
in which the primary axes of soil variation from our PCA were used to
generate a null model of sapling count as a function of soil features
(Model 1). Our second model included island as an explanatory
variable for sapling counts (Model 2). We then used a likelihood-ratio
test to assess the importance of including island as an explanatory
variable for sapling count. This approach allowed us to test for
differences in regeneration between islands while accounting for soil
differences. Finally, diversity indices were calculated for each
community on each island. Alpha diversity was calculated as the
species richness within each plot. The Shannon-Wiener diversity index,
calculated using base e, additionally accounted for species evenness as
determined by percent cover of each species in plots. Evenness tests the
extent to which species abundance distributions are skewed towards
few dominant species versus many evenly abundant species. A plot
represented by species with equal relative cover is equivalent to the
natural log of alpha diversity.
RESULTS
In total, we identified 235 plant taxa in 61 families on the Duck
Islands—189 in 56 families on GDI and 151 in 47 families on LDI
52 [Vol. 118
Rhodora

(Appendix). There were 83 species unique to GDI and 46 unique to
LDI, with 106 common between both islands (45%). In the 60
vegetation plots on each island, we found 130 species in 42 families on
GDI and 84 species in 38 families on LDI. Thirty-one species were non-
native, 27% of species on LDI and 24% on GDI. Furthermore, 31
species were newly recorded for GDI, and 44 were newly recorded on
LDI. A total of 62 species were previously recorded on GDI (from
multiple surveys dating between 1885 to 1986), but not found in the
current study. A total of 30 species were previously recorded on LDI,
but not found in the current study.
Forest. The forest community of GDI was dominated by an
overstory of Picea spp. with Betula papyrifera,Sorbus spp., and Abies
balsamea found in only one or two forest plots (Figure 2). The
overstory on LDI was dominated by both A. balsamea and Picea spp.,
and included a greater richness and abundance of broad-leaved
deciduous species such as Acer pensylvanicum,A. spicatum, and B.
papyrifera (Figure 2). We also found a significantly greater number of
regenerating tree saplings per plot on LDI (17.97 64.37 SEM) than on
GDI (2.41 61.78 SEM), even after allowing for differences in soils (p
,0.001; v
2
(1) ¼12.17). This included greater richness of both
coniferous and deciduous sapling species on LDI (Figure 3), including
Abies balsamea and Acer spicatum, with occasional Sorbus spp. and
Acer pensylvanicum. The saplings on GDI mainly consisted of B.
papyrifera and S. americana, though these saplings were only
encountered in one or very few forest plots (Figure 3). The understory
of each island was dominated by Dryopteris spp. (Figure 4). Other than
wood fern, the understory of GDI was mainly composed of
Chamaepericlymenum canadense,Maianthemum canadense, and Ocle-
mena acuminata, whereas A. balsamea and A. spicata dominated the
understory cover on LDI. Clintonia borealis and Streptopus lanceolatus
were also frequently encountered in the forest understory community
of LDI (Figure 4). A mean alpha diversity of 11.85 (60.84) species was
encountered in forest plots on GDI, and a mean alpha diversity of
14.14 (60.64) species was encountered in forest plots on LDI (Table 2).
See Table 2 for the Shannon-Wiener diversity index. Tree recruitment
peaked in the 1940s for GDI, when close to 25% of the trees surveyed
had reestablished (Figure 5). Tree recruitment on LDI, in contrast,
peaked in the 1960s, when over 25% of trees surveyed had reestablished
(Figure 5).
Field. The field community on both islands consisted of a mix of
forbs, graminoids, and shrubs. The field community on GDI was
2016] 53
Negoita et al.—Flora of Duck Islands, Maine

characterized by a dominant layer of Festuca rubra, with occasional
Vaccinium angustifolium,Deschampsia flexuosa, and Rubus hispidus,
and with the less dominant, but frequent occurrence of R. idaeus,
Achillea millefolium,Rumex acetosella, and Fragaria virginiana (Figure
6). The field on LDI was dominated by Poa pratensis. Other abundant
or frequent species on LDI included Rubus idaeus,Elymus repens,F.
virginiana,Moehringia lateriflora,Chamerion angustifolium, and Soli-
Figure 2. Tree composition of the forest community on Great Duck and
Little Duck Islands during the summers of 2010–2011. Relative species
abundance is determined as mean stem count (individuals .5 cm DBH; 6
standard error). Frequency refers to the percent of plots on each island
occupied by stems.
54 [Vol. 118
Rhodora

dago rugosa (Figure 6). A mean alpha diversity of 19 (61.57) species
was encountered in field plots on GDI, and a mean alpha diversity of
12.65 (60.81) species was encountered in field plots on LDI (Table 2).
See Table 2 for the Shannon-Wiener diversity index.
Ocean-side. The GDI ocean-side community was characterized by
its dominant field species, Festuca rubra and Symphyotrichum novi-
belgii, and a generally sparse cover of Agrostis spp. and Calystegia
sepium (Figure 7). In contrast to the ocean-side community on LDI,
several salt-tolerant or wetland species were occasionally encountered,
including Argentina egedii,Bolboschoenus maritimus,Juncus balticus,J.
gerardii,Plantago maritima, and Impatiens capensis (Figure 7). The LDI
ocean-side community was also dominated by the abundant and
frequent species found in its field community, followed by an abundance
Figure 3. Saplings found in each vegetative community on Great Duck and
Little Duck Islands, during the summers of 2010–2011. Saplings are defined as
trees .20 cm in height but ,5 cm DBH. Frequency refers to the percent of
plots on each island occupied by stems. Overall, sapling counts within each plot
differed significantly between islands in the forest community even when
accounting for edaphic differences; p ,0.001; v
2
(1) ¼12.17.
2016] 55
Negoita et al.—Flora of Duck Islands, Maine

of Lathyrus japonicus,Elymus repens,Rubus idaeus,C. sepium, and
Angelica lucida (Figure 7). A mean alpha diversity of 14 (61.34) species
was encountered in ocean-side plots on GDI, and a mean alpha
diversity of 13.27 (61.42) species was encountered in ocean-side plots
on LDI (Table 2). See Table 2 for the Shannon-Wiener diversity index.
Soils. Visual inspection of soil PCA ordinations for each community
indicated that LDI and GDI generally differed in their soil properties
(Figure 8). The greater overall extent of GDI plots in multivariate
Figure 4. Forest community species with the 10 highest abundance and
frequency ranks on Great Duck and Little Duck Islands, during summers 2010–
2011. Abundance was determined as mean percent cover (6SE). Frequency
refers to the percent of plots on each island occupied by each species.
Additional species were included to account for ties in abundance or frequency.
56 [Vol. 118
Rhodora

ordination space suggested that GDI generally had a greater variation in
soil features between plots (Figure 3). The primary two PCA axes
explained 60% of the variation in soil properties in forest soils, 66% of
the variation in field soils, and 65% of the variation in ocean-side soils.
The first axis of variation in forest soils was primarily driven by a
gradient of physical properties related to CEC, organic matter, and
associated nutrient availabilities. The second axis of variation in forest
soils explained more of the soil variation between GDI and LDI, and was
primarily associated with pH, nutrients, and soluble salts. Similar to
forest soils, the first axis of variation in field soils was primarily driven by
physical properties related to CEC, organic matter, and associated
nutrient availabilities, whereas the second axis better differentiated
between the islands and was driven by pH and essential nutrients. The
first axis of variation in the ocean-side soils was similar to the first axis of
other communities, but with the additional variation in soluble salts and
S. The second axis of variation in ocean-side soils was similar to field soils
but with the addition of Mg. Overall, PCA results suggested that axes
related to pH and essential nutrient concentrations best explained the
variation between islands, though salinity and S were also important for
differences between ocean-side soils of the islands. Through pairwise
comparisons, we found pH to be higher on LDI in all three communities.
Figure 5. Tree recruitment history of Great Duck and Little Duck Islands.
Tree ages were determined by coring all trees greater than 5 cm DBH in plots
and adding 10 to ring count to account for age at breast height. Tree age was
then subtracted from 2010 (LDI) or 2011 (GDI) to determine recruitment
decade (n ¼115 for LDI, n ¼63 for GDI).
2016] 57
Negoita et al.—Flora of Duck Islands, Maine

In addition, P was higher on LDI in fields, Ca was higher in LDI forests,
and K was higher in LDI ocean-side sites (Table 1). Soluble salt was
higher in GDI forests and ocean-side communities (Table 1). No other
soil features differed significantly between islands (Table 1).
DISCUSSION
Our study is among the first floristic studies of Maine’s islands to
generate baseline ecological information for vascular plant species
Figure 6. Field community species with the 10 highest abundance and
frequency ranks on Great Duck and Little Duck Islands, during summers 2010–
2011. Abundance is determined as mean percent cover (6SE). Frequency refers
to the percent of plots on each island occupied by each species. Additional
species were included to account for ties in abundance or frequency.
58 [Vol. 118
Rhodora

diversity, abundance, and associated edaphic features. Our findings
suggest that soil characteristics and the dominance and regeneration of
vascular plant species can differ substantially, even between adjacent
islands with otherwise similar geologic characteristics and glacial
history. Differences in vegetation structure were especially apparent in
the forest communities. The overstory on LDI was dominated by Abies
balsamea and had a greater diversity of both coniferous evergreen and
broad-leaf deciduous trees, despite the overall greater diversity of
Figure 7. Ocean-side community species with the 10 highest abundance and
frequency ranks on Great Duck and Little Duck Islands, during summers 2010–
2011. Abundance is determined as mean percent cover (6SE). Frequency refers
to the percent of plots on each island occupied by each species. Additional
species were included to account for ties in abundance or frequency.
2016] 59
Negoita et al.—Flora of Duck Islands, Maine

Figure 8. Principal components analysis of edaphic features within communities on Great Duck and Little Duck Islands, during
summers 2010–2011. Each ordination presents the primary two axes of variation, with the percent variation explained by each axis in
parentheses. Edaphic features with the greatest loadings on each principal component (greater than 0.3) are labeled on each axis. A
minus sign (–) before a feature indicates an inverse relationship with axis arrows.
60 [Vol. 118
Rhodora

species on GDI. The historic recruitment of the overstory on LDI
peaked more recently than on GDI, and the greater abundance of
saplings on LDI suggests that, at the time of this study, the tree species
found on this island were successfully regenerating. In contrast, the
forest on GDI was almost exclusively dominated by Picea spp., which
peaked in recruitment in the 1940s and for which few saplings were
recorded in the current study. The overall significantly lower sapling
count on GDI was evident even when accounting for differences in soil
features.
A number of abiotic and biotic factors may have contributed to the
documented vegetation and soils differences between LDI and GDI.
For example, GDI has a maximum elevation of about 18 m compared
to 27 m on LDI. The lower elevation on GDI could expose inland
habitats to more salt spray. This can explain the significantly greater
soluble salt concentrations in the forest and ocean-side community, and
the greater diversity of halophyte (salt-tolerant) species in the ocean-
side community on GDI. It may also explain why Festuca rubra, a more
salt-tolerant species than Poa pratensis (Torello and Symington 1984),
was the dominant graminoid in the open communities on GDI.
However, sapling counts in GDI forests were lower even when
accounting for the edaphic differences.
Another factor that contributed to the vegetation differences is the
land-use history of these islands. GDI has a long history of introduced
mammalian herbivores, including at least 100 y of sheep and 60 y of
European and Snowshoe hares, compared to LDI, which has been
largely free of mammalian herbivores for at least the last 100 y
(McLane 1989). Mammalian herbivores can have drastic impacts on
plant communities (Crawley 1997; Donlan et al. 2002; McLaren et al.
2004; Nu ˜
nez et al. 2010; Peterson et al. 2005; Terborgh et al. 2001), so it
seems plausible that the long-term history of mammalian habitation on
GDI could have played a role in the vegetation and soil differences
between these islands. Herbivory may shift the composition of plant
species through preferential browsing and grazing of more palatable or
noticeable species. Over time, this can lead to communities composed
of species that are more tolerant, or less palatable, to herbivores (D ´
ıaz
et al. 2001; Gillham 1955; McInnes et al. 1992). For example, Clintonia
borealis and Streptopus lanceolatus have been shown to be particularly
vulnerable to mammalian herbivory (Balgooyen and Waller 1995;
Kraft et al. 2004; Lapointe et al. 2010), which suggests why these
species may be lacking on GDI, though frequently encountered in plots
on LDI.
2016] 61
Negoita et al.—Flora of Duck Islands, Maine

Table 1. Soil features of Great Duck and Little Duck Islands, Maine during summers 2010–2011. All values are means 6standard
error. Elemental concentrations are reported as ppm (lg/g dry soil). ENR (estimated nitrogen release) is reported as lbs/acre. Cation
exchange capacity (CEC) and hydrogen (H) are reported as meq/100g. Soluble salts are reported as mmhos/com. The soil features that
differed significantly between islands within each community are identified by bold font (Mann-Whitney two-tailed tests; p-values
adjusted with Hochberg method; alpha ¼0.05). Note: n ¼27 for GDI Forest due to missing soil samples from two plots.
Soil Feature
Forest Field Ocean-side
GDI (n¼27) LDI (n¼29) GDI (n¼19) LDI (n¼20) GDI (n¼12) LDI (n¼11)
% Organic Matter 52.3 (65.97) 38.7 (63.29) 33.7 (66.31) 18.6 (61.42) 62.4 (66.79) 38.0 (65.45)
ENR 1076 (6119) 803 (665.9) 704 (6126) 401 (628.5) 1278 (6136) 789 (6109)
P 12.6 (62.71) 14.0 (61.79) 17.5 (64.09) 40.0 (64.19) 9.95 (62.78) 28.3 (64.59)
pH 3.74 (60.05) 3.93 (60.04) 4.37 (60.09) 4.79 (60.07) 4.49 (60.16) 5.14 (60.12)
K 66.1 (64.29) 74.5 (65.1) 69.0 (66.14) 89.9 (67.99) 90.8 (615.7) 176 (627.2)
Mg 241 (619.5) 231 (617.2) 201 (625.0) 245 (631.4) 542 (683.9) 617 (676.4)
Ca 312 (629.5) 371 (624.2) 411 (648.4) 976 (6111) 632 (690.2) 1006 (6158)
Na 181 (625.8) 145 (69.04) 105 (618.4) 115 (618.6) 801 (6270) 372 (665.5)
H 12.9 (61.01) 12.6 (60.75) 8.22 (61.34) 6.63 (60.45) 19.1 (66.43) 7.23 (60.99)
CEC 17.4 (61.34) 17.1 (60.99) 12.5 (61.73) 14.2 (60.99) 30.5 (68.10) 19.4 (62.01)
Nitrate (N) 1.33 (60.35) 1.95 (60.53) 23.5 (65.97) 16.6 (64.37) 20.2 (67.51) 14.4 (63.66)
S 18.9 (63.60) 14.9 (61.01) 16.7 (61.94) 16.6 (62.34) 287 (6177.2) 16.2 (61.43)
Soluble Salts (SS) 0.51 (60.05) 0.30 (60.02) 0.52 (60.05) 0.33 (60.03) 3.96 (62.41) 0.4 (60.06)
62 [Vol. 118
Rhodora

Table 2. Species diversity indices for Great Duck and Little Duck Islands, Maine, during summers 2010–2011. The Shannon-
Wiener diversity index is calculated using base e.
Forest Field Ocean-side
Diversity Index GDI (n¼29) LDI (n¼29) GDI (n¼19) LDI (n¼20) GDI (n¼12) LDI (n¼11)
Alpha (6SD) 11.85 (60.84) 14.14 (60.64) 19 (61.57) 12.65 (60.81) 14 (61.34) 13.27 (61.42)
Shannon-Wiener (6SD) 1.68 (60.08) 1.57 (60.05) 1.94 (60.07) 1.72 (60.05) 1.78 (60.09) 1.58 (60.09)
2016] 63
Negoita et al.—Flora of Duck Islands, Maine

Herbivory can also affect forest succession by, directly or indirectly,
suppressing or supporting the growth of certain woody species (Angell
and Kielland 2009; Heinen and Currey 2000; McInnes et al. 1992;
Peterson et al. 2005). Betula papyrifera was the primary regenerating
species in forests on GDI, congruent with a study that found young B.
papyrifera individuals to be more resistant to snowshoe hare herbivory
by reducing palatability through a high resin content in juvenile twigs
(Bryant et al. 1983). Almost no Picea spp. recruitment was encountered
on GDI, consistent with a study comparing forest regeneration on two
Maine islands that differed in long-term snowshoe hare herbivory
(Peterson et al. 2005). This study concluded that hares were actively
inhibiting the regeneration of northern spruce-fir forest through
seedling browsing, reflected by a decline in tree recruitment following
the introduction of the hares. We found a similar result with our
histogram of tree recruitment on the Duck Islands over the last century,
in which tree recruitment on GDI has declined since the introduction of
hares in the 1940s (Figure 5). Lacking a better alternative hypothesis,
we suggest the reduced sapling regeneration and decline in tree
recruitment on GDI is due to the introduction of hares. However,
future experimental work (i.e., long-term hare exclosures that track
regenerating individuals to maturity) is necessary for directly testing
this hypothesis (Clark et al. 1999). This is because the effect of
herbivory on plant community succession may be highly context
dependent—in some cases accelerating woody succession (Davidson
1993).
Vegetation differences may also be driven by soil processes, in some
cases, to a greater extent than by herbivory (Turkington et al. 2002).
This may be especially important when extrinsic factors such as nesting
seabirds drive essential nutrient and heavy metal concentrations in the
soil (Ellis 2005; Rajakaruna et al. 2009). Essential nutrient concentra-
tions of P and Ca were about two-fold greater in the field community on
LDI, and K was about two-fold greater in the ocean-side community on
LDI. This may reflect a longer history of seabird nesting on LDI, which
would lead to an increase in nutrient concentrations (Ellis 2005).
Finally, the plants are key drivers of soil properties, and feedbacks make
it impossible to distinguish cause from effect without adequate
experimental data (Chapin et al. 2011).
Little Duck Island and GDI also differ in other aspects of their land-
use history. Humans have not inhabited LDI over the last century,
whereas GDI had at least three families with multiple dwellings,
including a schoolhouse for thirty children, at its peak habitation in the
early 20th century (McLane 1989). Although the primary human
64 [Vol. 118
Rhodora

impact may be directly linked to grazing by introduced sheep and
browsing by introduced hares, localized agricultural plots, trail
compaction, an airplane landing strip, and timber harvest may have
also had important impacts on the soils and vegetation of GDI. Timber
harvest may have influenced the tree demography by removing certain
age classes and reducing the diversity of hardwoods in the canopy
(Figures 2, 5). Human habitation of GDI has largely declined over the
last century and especially, since 1986, when the lighthouse became
automated. Thus, it seems unlikely that human history had a
confounding influence on the most recent decline in tree recruitment.
Those trees would have likely been too young to be worth harvesting
until at least the 1970s and, by then, human habitation had declined.
Furthermore, any human impact by trail compaction, localized
agriculture, or the airplane landing strip is unlikely to be evident in
our plot data since we avoided placing plots in areas with evident
human disturbance. Overall species lists for the Duck Islands, however,
may reflect some of these human effects. For example, several species of
orchids such as Malaxis unifolia and Platanthera clavellata were found
growing on the airplane landing strip on GDI and were otherwise
absent from the rest of the island. Finally, it is important to
acknowledge prehistoric differences in human-use between GDI and
LDI and the potential impacts of shell middens on soil nutrients and
vegetation structure (Cook-Patton et al. 2014). Future work may
consider available archaeologic history as an additional factor.
Repeated resurveys are important for gauging long-term changes in
communities and ecosystems, though it would be important that a
similar survey effort be made in order to yield comparable results. Most
earlier surveys of the Duck Islands were not as thorough as the current
study (but see Folger and Wayne 1986). For example, efforts by
Redfield (1885, 1893) and Rand (1900) were based on single- or several-
day surveys and, by Rappaport and Wesley (1985), on only five days.
Lesser (1977) did not include any graminoids or other cryptic species.
Thus, the extant surveys are not ideal for drawing conclusions about
factors contributing to any documented changes. Nonetheless, some
interesting anecdotes may be gleaned from historic surveys. For
example, Capsella bursa-pastoris, an otherwise common ruderal species,
was not encountered in the current study of the Duck Islands, yet it was
commonly recorded on both islands in historic surveys of lesser effort.
The disappearance of C. bursa-pastoris suggests the potential role of
stochastic population drift in communities (Vellend 2010), especially as
it may affect short-lived species such as C. bursa-pastoris on islands.
Overall, 73 species were historically recorded but not found in the
2016] 65
Negoita et al.—Flora of Duck Islands, Maine

current survey, which supports the idea of the dynamic nature of island
community turnover as described by MacArthur and Wilson’s (1967)
theory of island biogeography (but see Nilsson and Nilsson 1983).
Future resurveys should also include quantitative records of species
abundances in order to gauge relevant changes in ecosystem properties
that are otherwise lost in the coarse scale of basic floristic surveys.
Limitations of the study. We chose to avoid plant-community
transition zones in our vegetation survey. Although this approach was
necessary for characterizing particular plant communities, it ignores the
unique characteristics of the transition zones. Several species were most
common in the transition zone between forest and field, rather than in
either community. These, and other species missed in plots, were
captured on the total species list (Appendix). The transition between
forest and field may also harbor a greater number of regenerating trees
due to proximity to seed sources and increased light availability,
though such data were unavailable from our study. Although our plot
design does not account for all regeneration, composition, and diversity
of each island, it does offer data that are comparable between these
and, hopefully, future island surveys.
Our study is among the first to incorporate edaphic features and
vascular plant species abundances into a robust baseline description of
island ecosystems in the Gulf of Maine (but see Rajakaruna et al.
2009). Our study provides an important baseline from which to gauge
future changes in coastal Maine habitats, and our causal understanding
of island ecosystems will increase as more surveys are conducted and
their data made available. We suggest that future island inventories
incorporate plot surveys to estimate the abundance of plant species, as
well as to quantify associated edaphic properties. Such baseline data
and future re-surveys will be essential for better understanding the
potential direct and indirect effects of climate change, rising sea levels,
herbivory, and other human impacts.
ACKNOWLEDGMENTS. We thank Kaija Klauder for her field assis-
tance with island surveys. We also thank Chris Petersen for advice and
feedback on the design of this study, and Robert S. Boyd, Robert S.
Capers, and two anonymous reviewers for helpful comments on earlier
drafts of this paper. We thank John Anderson, College of the Atlantic,
and the Alice Eno Research Station on Great Duck Island for lodging
and accommodation during fieldwork. We thank Trent Quimby and
Scott Swann for boat transportation. We also thank Arthur Haines, Jill
Weber, Sal Rooney, Jordan Chalfant, David Werier, Rick Fournier,
Bill Moorhead, and Eric Doucette for help with plant specimen
66 [Vol. 118
Rhodora

identification. Finally, we are grateful to the National Audubon
Society, Acadia National Park, College of the Atlantic, and The Nature
Conservancy for permission to conduct fieldwork on the Duck Islands.
This study was supported by Maine Space Grant Consortium summer
fellowships and Garden Club of America field botany grants to M.
Dickinson and L. Negoita, and was completed in fulfillment of the
College of the Atlantic senior project graduation requirement by L.
Negoita.
LITERATURE CITED
ALLEN, B., G. H. MITTELHAUSER,L.WELCH,R.HOUSTON,R.SCHAUFFLER,AND
M. LANGLOIS. 2012. Maine atlas of breeding seabird and coastal wading
bird colonies – 1960 to 2011. Maine Natural History Observatory,
Gouldsboro, ME.
ANGELL,A.C.AND K. KIELLAND. 2009. Establishment and growth of white
spruce on a boreal forest floodplain: Interactions between microclimate
and mammalian herbivory. Forest Ecol. Managem. 258: 2475–2480.
BALGOOYEN,C.P.AND D. M. WALLER. 1995. The use of Clintonia borealis and
other indicators to gauge impacts of white-tailed deer on plant
communities in northern Wisconsin, USA. Nat. Areas J. 15: 308–318.
BARNHARDT, W. A., W. R. GEHREIS,D.F.BELKNAP,AND J. T. KELLEY. 1995.
Late Quaternary relative sea-level change in the western Gulf of Maine:
Evidence for a migrating glacial forebulge. Geology 23: 317–320.
BRYANT, J., F. S. CHAPIN III, AND D. KLEIN. 1983. Carbon/nutrient balance of
boreal plants in relation to vertebrate herbivory. Oikos 40: 357–368.
CAMPBELL, C. S., H. P. ADAMS,P.ADAMS,A.C.DIBBLE,L.M.EASTMAN,S.C.
GAWLER,ET AL. 1995. Checklist of vascular plants of Maine, 3rd revis. Bull.
844, Maine Agricultural and Forest Experiment Station, Orono, ME.
CAUJAP ´
E-CASTELLS, J., A. TYE,D.J.CRAWFORD,A.SANTOS-GUERRA,A.SAKAI,
K. BEAVER,ET AL. 2010. Conservation of oceanic island floras: Present and
future global challenges. Perspect. Pl. Ecol. Evol. Syst. 12: 107–129.
CHAPIN, F. S. III, P. A. MATSON,AND P. VITOUSEK. 2011. Principles of
Terrestrial Ecosystem Ecology. Springer-Verlag, New York, NY.
CLARK, J., B. BECKAGE,P.CAMILL,B.CLEVELAND,J.HILLERISLAMBERS,AND J.
LICHTER. 1999. Interpreting recruitment limitation in forests. Amer. J. Bot.
86: 1–16.
CLAYDEN, S. R., M. C. MUNRO,C.S.BLANEY,AND S. P. VANDER KLOET. 2010.
Vascular flora of the Atlantic maritime ecozone: Some new perspectives,
pp. 197–213. In: D. F. McAlpine and I. M. Smith., eds., Assessment of
Species Diversity in the Atlantic Maritime Ecozone. NRC Research Press,
Ottawa, Canada.
CONKLING, P. 2011. Islands in Time: A Natural and Cultural History of Islands
in the Gulf of Maine, 3rd ed. Island Institute, Rockland, ME.
COOK-PATTON, S. C., D. WELLER,T.C.RICK,AND J. D. PARKER. 2014. Ancient
2016] 67
Negoita et al.—Flora of Duck Islands, Maine

experiments: Forest biodiversity and soil nutrients enhanced by Native
American middens. Landscape Ecol. 29: 979–987.
CRAWLEY, M. J. 1997. Plant-herbivore dynamics, pp. 401–474. In:M.J.
Crawley, ed., Plant Ecology, 2nd ed. Wiley-Blackwell, Oxford, UK.
DAVIDSON, D. A. 1993. The effects of herbivory and granivory on terrestrial
plant succession. Oikos 68: 23–35.
DAVIS, R. B. R. 1966. Spruce-fir forests of the coast of Maine. Ecol. Monogr.
36: 79–94.
D´
IAZ, S., I. NOY-MEIR,AND M. CABIDO. 2001. Can grazing response of
herbaceous plants be predicted from simple vegetative traits? J. Appl. Ecol.
38: 497–508.
DONLAN, C. J., B. R. TERSHY,AND D. CROLL. 2002. Islands and introduced
herbivores: Conservation action as ecosystem experimentation. J. Appl.
Ecol. 39: 235–246.
ELLIS, J. 2005. Marine birds on land: A review of plant biomass, species
richness, and community composition in seabird colonies. Pl. Ecol. 181:
227–241.
ELLIS, J. C., J. M. FARI ˜
NA,AND J. D. WITMAN. 2006. Nutrient transfer from sea
to land: The case of gulls and cormorants in the Gulf of Maine. J. Anim.
Ecol. 75: 565–574.
———, P. J. BELLINGHAM,E.K.CAMERON,D.A.CROLL,G.S.KOLB,C.
KUEFFER,G.H.MITTELHAUSER,S.SCHMIDT,E.VIDAL,AND D. A. WAIT.
2011. Effects of seabirds on plant communities. pp. 177–211. In:C.P.H.
Mulder, W. B. Anderson, D. R. Towns, and P. J. Bellingham, eds., Seabird
Islands: Ecology, Invasion, and Restoration. Oxford University Press,
Oxford, UK.
FOLGER,D.C.AND P. M. WAYNE. 1986. A biological inventory of Great Duck
Island: Hancock County, Maine. The Nature Conservancy, Arlington,
VA.
GAWLER,S.AND A. CUTKO. 2010. Natural Landscapes of Maine: A Guide to
Natural Communities and Ecosystems. Maine Natural Areas Program,
Department of Conservation, Augusta, ME.
GILLHAM, M. E. M. 1955. Ecology of the Pembrokeshire Islands. III. The effect
of grazing on the vegetation. J. Ecol. 43: 172–206.
GREENE, C. W., L. L. GREGORY,G.H.MITTELHAUSER,S.C.ROONEY,AND J. E.
WEBER. 2005. Vascular flora of the Acadia National Park region, Maine.
Rhodora 107: 117–185.
HAINES, A. 2011. Flora Novae Angliae. Yale University Press, New Haven, CT.
HARRIS, T. B., N. RAJAKARUNA,S.J.NELSON,AND P. D. VAUX. 2012. Stressors
and threats to the flora of Acadia National Park, Maine: Current
knowledge, information gaps, and future directions. J. Torrey Bot. Soc.
139: 323–344.
HEINEN,J.T.AND R. C. D. CURREY. 2000. A 22-year study on the effects of
mammalian browsing on forest succession following a clearcut in northern
Lower Michigan. Amer. Midl. Naturalist 144: 243–252.
68 [Vol. 118
Rhodora

HOCHBERG, Y. 1988. A sharper bonferroni procedure for multiple tests of
significance. Biometrika 75: 800–802.
HODGDON,A.R.AND R. B. PIKE. 1969. Floristic comparison of three bird
islands in the Gulf of Maine. Rhodora 71: 510–523.
IRLAND, L. C. 1997. Maine’s forest vegetation regions: Selected maps 858–1993
N. E. Naturalist 4: 241–260.
JORDAN, G. B. 1988. Soil Survey of Hancock County Area, Maine. U.S.
Department of Agriculture, Natural Resource Conservation Service.
Augusta, ME.
KRAFT, L. S., T. R. CROW,D.S.BUCKLEY,E.NAUERTZ,AND J. C. ZASADA. 2004.
Effects of harvesting and deer browsing on attributes of understory plants
in northern hardwood forests, Upper Michigan, USA. Forest Ecol.
Managem. 199: 219–230.
LAPOINTE, L., J. BUSSI `
ERES,M.CRˆ
ETE,AND J. P. OUELLET. 2010. Impact of
growth form and carbohydrate reserves on tolerance to simulated deer
herbivory and subsequent recovery in Liliaceae. Amer. J. Bot. 97: 913–924.
LESSER, E. A. 1977. A study of maritime birds on Little Duck Island, Maine.
Senior project report. College of the Atlanic, Bar Harbor, ME.
LEWIS, A. J. 1983. Distribution of three rare plants on islands in Machias Bay,
ME. Rhodora 85: 385–388.
MACARTHUR,R.H.AND E. O. WILSON. 1967. The Theory of Island
Biogeography. Monographs in Population Biology. Princeton University
Press, Princeton, NJ.
MCINNES, P., R. NAIMAN,AND J. PASTOR,AND Y. COHEN. 1992. Effects of moose
browsing on vegetation and litter of the boreal forest, Isle Royale,
Michigan, USA. Ecology 73: 2059–2075.
MCLANE, C. B. 1989. Islands of the Mid-Maine Coast. Vol. 2. Mount Desert to
Machias Bay. The Kennebec River Press, Inc., Falmouth, ME.
MCLAREN, B. E., B. A. ROBERTS,N.DJAN-CHEKAR,AND K. P. LEWIS. 2004.
Effects of overabundant moose on the Newfoundland landscape. Alces 40:
45–59.
MCMAHON J. S. 1990. Biophysical Regions of Maine. State Planning Office,
Augusta, ME.
MCMASTER, R. T. 2005. Factors influencing vascular plant diversity on 22
islands off the coast of eastern North America. J. Biogeogr. 32: 475–492.
MULLIGAN, A. D. 1980. The flora, vegetation, and phytogeographic
relationships of Whaleboat Island, Casco Bay, Maine. Rhodora 82:441–
459.
NICHOLS,W.F.AND V. C. NICHOLS. 2008. The land use history, flora, and
natural communities of the Isles of Shoals, Rye, New Hampshire and
Kittery, Maine. Rhodora 110: 245–295.
NILSSON,S.G.AND I. N. NILSSON. 1983. Are estimated species turnover rates on
islands largely sampling errors? Amer. Naturalist 121: 595–597.
NORTH AMERICAN PROFICIENCY TESTING PROGRAM (NAPT). 2011. Recommended
Soil Testing Procedures for the Northeastern United States, 3rd ed. Soil
2016] 69
Negoita et al.—Flora of Duck Islands, Maine

Science Society of America, Madison, WI. Website (www.naptprogram.
org/methods). Accessed February 2015.
NU˜
NEZ, M., J. K. BAILEY,AND J. SCHWEITZER. 2010. Population, community,
and ecosystem effects of exotic herbivores: A growing global concern. Biol.
Invas. 12: 297–301.
OSBERG, P. H., A. M. HUSSEY II, AND G. M. BOONE. 1985. Bedrock geologic
map of Maine. Maine Geological Survey, Augusta, ME.
PETERSON, T. S., A. UESUGI,AND J. LICHTER. 2005. Tree recruitment limitation
by introduced Snowshoe Hares, Lepus americanus, on Kent Island, New
Brunswick. Canad. Field-Naturalist 119: 569–572.
PIKE,R.B.AND A. R. HODGDON. 1962. Changes in flora of the Machias Seal
Islands. Rhodora 64: 340–346.
RAJAKARUNA, N., N. POPE,J.PEREZ-OROZCO,AND T. B. HARRIS. 2009.
Ornithocoprophilous plants of Mount Desert Rock, a remote bird-
nesting island in the Gulf of Maine, U.S.A. Rhodora 111: 417–447.
RAND, E. L. 1900. Plants from the Duck Islands, Maine. Rhodora 2: 207–209.
RAPPAPORT, N. R., AND F. R. WESLEY. 1985. A vegetation study of nine bird
islands off the Coast of Maine. National Audubon Society Report,
Falmouth, ME.
RD
EVELOPMENT CORE TEAM. 2014. R: A language and environment for
statistical computing. R Foundation for Statistical Computing, Vienna,
Austria.
REDFIELD, J. H. 1885. Insular vegetation. Bull. Torrey Bot. Club 12: 103.
———. 1893. Botanical notes. Bull. Torrey Bot. Club 20: 409–410.
STEBBINS, G. L., JR. 1929. Further additions to the Mt. Desert Flora. Rhodora
31: 81–87.
TERBORGH, J., L. LOPEZ,P.NU˜
NEZ,M.RAO,G.SHAHABUDDIN,G.ORIHUELA,ET
AL. 2001. Ecological meltdown in predator-free forest fragments. Science
294: 1923–1926.
TORELLO,W.A.AND A. G. SYMINGTON. 1984. Screening of turfgrass species and
cultivars for NaCl tolerance. Pl. & Soil 82: 155–161.
TURCOTTE,D.E.AND T. H. BUTLER JR. 2006. Folists and humods off the coast
of eastern Maine. Soil Survey Horizons 47: 5–9.
TURKINGTON, R., E. JOHN,S.WATSON,AND P. SECCOMBE-HETT. 2002. The effects
of fertilization and herbivory on the herbaceous vegetation of the boreal
forest in north-western Canada: A 10-year study. J. Ecol. 90: 325–337.
US CLIMATE DATA. 2015. Historic climate data, Mount Desert, Maine. US
Climate Data, version 2.2 beta. Website (http://www.usclimatedata.com/
climate/mount-desert/maine/united-states/usme0688/2015/1). Accessed
July 2015.
VELLEND, M. 2010. Conceptual synthesis in community ecology. Quart. Rev.
Biol. 85: 183–206.
WHERRY, E. T. 1926. Nitrogen as a factor in plant distribution on Mt. Desert
Island, Maine. Ecology 7: 140–142.
WISE, D. A. 1970. The flora of Isle au Haut, Maine. Rhodora 72: 505–532.
70 [Vol. 118
Rhodora

APPENDIX
CHECKLIST OF THE VASCULAR FLORA OF THE DUCK ISLANDS,MAINE,DURING SUMMERS OF 2010–2011
For each species, the table shows presence (X), absence (–), and frequency of species recorded during the 2010 (Little
Duck Island; LDI) and 2011 (Greater Duck Island; GDI) surveys of the two islands. Nativity (Nativ.) is in reference to
eastern North America (N ¼native, E ¼exotic). X* ¼species not previously recorded; X ¼species recorded both
previously and in the current study; O ¼species previously recorded but not encountered in the current study (Folger and
Wayne 1986; Lesser 1977; Rappaport and Wesley 1985; Redfield 1885, 1893). Plot frequency refers to the number of plots
occupied by each species of vascular flora on each island (N ¼60 plots for each island). Voucher specimens are deposited
at the College of the Atlantic Herbarium, Bar Harbor, ME (HCOA). Nomenclature is based on Haines (2011). Rand (1900)
did not distinguish between the islands so his study is excluded.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
ADOXACEAE Sambucus racemosa L. N X* X 0 1
Viburnum lantanoides Michx. N O – 0 0
Viburnum nudum L. N X X 0 0
ALLIACEAE Allium schoenoprasum L. N X* – 0 0
Allium sp. N/E O – 0 0
AMARANTHACEAE Atriplex prostrata Boucher ex DC. N – X 0 0
Atriplex subspicata (Nutt.) Rydb. N – X* 0 0
Chenopodium album L. E X X 3 2
Chenopodium berlandieri Moq. N – X* 0 0
AMARYLLIDACEAE Narcissus sp. E X – 0 0
ANACARDIACEAE Rhus glabra L. N – O 0 0
APIACEAE Angelica lucida L. N X X 3 12
Aralia hispida Vent. N X X* 2 1
2016] 71
Negoita et al.—Flora of Duck Islands, Maine

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Aralia nudicaulis L. N X X 12 3
Daucus carota L. E O – 0 0
Ligusticum scoticum L. N X X* 1 0
APOCYNACEAE Apocynum androsaemifolium L. N – X 0 0
AQUIFOLIACEAE Ilex verticillata (L.) Gray N X X* 0 0
Ilex mucronata (L.) M. Powell, Savol. & S.
Andrews
NO – 0 0
ASTERACEAE Achillea millefolium L. N X X 24 14
Anaphalis margaritacea (L.) Benth. & Hook. f. N X* – 1 0
Cirsium arvense (L.) Scop. E X X 5 1
Cirsium vulgare (Savi) Ten. E X X* 0 0
Doellingeria umbellata (P. Mill.) Nees N X X 14 3
Eurybia radula (Aiton) G.L. Nesom N O – 0 0
Euthamia graminifolia (L.) Nutt. N X X* 3 0
Gnaphalium uliginosum L. E O – 0 0
Hieracium aurantiacum L. E X X* 0 0
Hieracium caespitosum Dumort. E X X 0 0
Hieracium kalmia L. N O – 0 0
Hieracium pilosella L. E X O 0 0
Hieracium sp. N/E X* X 14 2
Matricaria discoidea DC. E X X 0 0
Nabalus altissimus (L.) Hook. N X X 0 0
Nabalus serpentarius (Pursh) Hook. N O – 0 0
72 [Vol. 118
Rhodora

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Nabalus trifoliolatus Cass. N – X 2 15
Oclemena acuminata (Michx.) Nesom N X X 26 19
Oclemena nemoralis (Aiton) Greene N X – 0 0
Scorzoneroides autumnalis (L.) Moench E X X 6 0
Senecio viscosus L. E – X* 0 0
Senecio vulgaris L. E X O 1 0
Solidago canadensis L. N X – 2 0
Solidago puberula Nutt. N X – 2 0
Solidago rugosa P. Mill. N X X 11 20
Solidago sempervirens L. N X X 0 4
Sonchus arvensis L. E X* X 1 1
Sonchus asper (L.) Hill E X – 3 0
Sonchus oleraceus L. E – X* 0 1
Symphyotrichum novi-belgii (L.) Nesom N X X 34 31
Tanacetum vulgare L. E X – 0 0
Taraxacum laevigatum (Willd.) DC. E – O
Taraxacum officinale G.H. Weber ex Wiggers E X X 3 7
BALSAMINACEAE Impatiens capensis Meerb. N X X 5 4
Impatiens pallida Nutt. N O – 0 0
BETULACEAE Betula alleghaniensis Britton N X X 0 6
Betula cordifolia Regel N – X* 0 0
Betula papyrifera Marshall N X X 6 11
BORAGINACEAE Symphytum officinale L. E X – 0 0
2016] 73
Negoita et al.—Flora of Duck Islands, Maine

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
BRASSICACEAE Barbarea vulgaris Aiton f. E – O 0 0
Cakile edentula (Bigelow) Hook. N X X 1 0
Capsella bursa-pastoris (L.) Medik. E O O 0 0
Cardamine parviflora L. N O – 0 0
Cardamine pensylvanica Muhl. ex Willd. N – O 0 0
Rorippa palustris (L.) Bess. N – O 0 0
Sisymbrium officinale (L.) Scop. E X* X 1 0
CAMPANULACEAE Campanula rotundifolia L. N – X* 0 0
CAPRIFOLIACEAE Diervilla lonicera P. Mill. N X X 0 3
Linnaea borealis L. N X X 1 8
Lonicera canadensis Bartram ex Marshall N – X 0 11
Valeriana officinalis L. E – X* 0 2
CARYOPHYLLACEAE Cerastium arvense L. E X X 16 8
Cerastium fontanum Baumg. E X X 4 12
Cerastium glomeratum Thuill. E O – 0 0
Moehringia lateriflora (L.) Fenzl N X X 18 24
Sagina procumbens L. N X X* 2 0
Spergularia marina (L.) Griseb. N X* X 1 0
Spergularia rubra (L.) J. & K. Presl E X – 2 0
Stellaria graminea L. E X X 1 0
Stellaria media (L.) Vill. E X X 1 3
CONVOLVULACEAE Calystegia sepium (L.) R.Br. N X* X 9 8
Cuscuta gronovii Willd. ex J.A. Schultes N X* X 3 6
74 [Vol. 118
Rhodora

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Ipomoea purpurea (L.) Roth E – O 0 0
CORNACEAE Chamaepericlymenum canadense (L.) Aschers. &
Graebn.
NX X 27 4
CUPRESSACEAE Juniperus communis L. N X – 0 0
CYPERACEAE Bolboschoenus maritimus (L.) Palla N X X* 1 0
Carex atlantica Bailey N X* – 1 0
Carex brunnescens (Pers.) Poir. N X X* 1 1
Carex canescens L. N O O 0 0
Carex debilis Michx. N O – 0 0
Carex folliculata L. N – X* 0 0
Carex gynandra Schwein. N – X 0 0
Carex hormathodes Fernald N X* X* 3 0
Carex mackenziei Krecz. N X – 0 0
Carex magellanica Lam. N O – 0 0
Carex nigra (L.) Reichard N X – 7 0
Carex paleacea Schreb. ex Wahlenb. N O – 0 0
Carex pallescens L. N O – 0 0
Carex pensylvanica Lam. N X – 1 0
Carex stipata Muhl. ex Willd. N – X* 0 0
Carex stricta Lam. N O – 0 0
Carex trisperma Dewey N X – 18 0
Eleocharis sp. N/E O X* 0 0
Eleocharis tenuis (Willd.) J.A. Schultes N X* – 0 0
2016] 75
Negoita et al.—Flora of Duck Islands, Maine

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Eleocharis uniglumis (Link) J.A. Schultes N O – 0 0
Eriophorum angustifolium Honckeny N X – 0 0
DENNSTAEDTIACEAE Dennstaedtia punctilobula (Michx.) T. Moore N X – 2 0
Pteridium aquilinum (L.) Kuhn N – X 0 0
DROSERACEAE Drosera rotundifolia L. N X – 3 0
DRYOPTERIDACEAE Dryopteris campyloptera (Kunze) Clarkson N – X 0 0
Dryopteris carthusiana (Vill.) H.P. Fuchs N X O 0 0
Dryopteris sp. N X* X 24 32
ELAEAGNACEAE Elaeagnus angustifolia L. E O – 0 0
EQUISETACEAE Equisetum arvense L. N X X 1 1
ERICACEAE Empetrum nigrum L. N X – 0 0
Gaultheria hispidula (L.) Muhl. ex Bigelow N O O 0 0
Kalmia angustifolia L. N X – 1 0
Moneses uniflora (L.) Gray N X X 2 0
Monotropa uniflora L. N X X 5 4
Rhododendron canadense (L.) Torr. N X* – 0 0
Vaccinium angustifolium Aiton N X X 13 1
Vaccinium macrocarpon Aiton N X O 3 0
Vaccinium oxycoccos L. N X – 0 0
Vaccinium vitis-idaea L. N X X* 9 0
FABACEAE Lathyrus japonicus Willd. N X X 2 9
Lathyrus palustris L. N X X 1 2
Trifolium arvense L. E – X* 0 0
76 [Vol. 118
Rhodora

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Trifolium aureum Pollich E X X* 0 0
Trifolium pratense L. E X – 0 0
Trifolium repens L. E X X 13 5
Vicia cracca L. E X X 0 0
GERANIACEAE Geranium robertianum L. N X X 1 5
GROSSULARIACEAE Ribes glandulosum Grauer N X X 0 27
Ribes hirtellum Michx. N O X* 0 1
Ribes lacustre (Pers.) Poir. N X* – 0 0
Ribes triste Pallas N – O 0 0
HEMEROCALLIDACEAE Hemerocallis fulva (L.) L. E X – 0 0
HUPERZIACEAE Huperzia lucidula (Michx.) Trevisan N – X 0 0
HYPERICACEAE Hypericum canadense L. N X – 0 0
Hypericum perforatum L. E – X* 0 0
Triadenum virginicum (L.) Raf. N X – 0 0
IRIDACEAE Iris hookeri Penny ex D. Don N O – 0 0
Iris sp. (cultivated) N/E O – 0 0
Iris versicolor L. N X X 5 0
Sisyrinchium angustifolium Mill. N O – 0 0
Sisyrinchium atlanticum Bicknell N O – 0 0
Sisyrinchium montanum Greene N X – 1 0
JUNCACEAE Juncus balticus Willd. N X – 2 0
Juncus brevicaudatus (Engelm.) Fernald N X* – 1 0
Juncus bufonius L. N O X* 0 1
2016] 77
Negoita et al.—Flora of Duck Islands, Maine

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Juncus effusus L. N/E – X 0 0
Juncus gerardii Loisel. N X X 2 0
Juncus pelocarpus E. Mey. N O – 0 0
Juncus tenuis Willd. N X – 2 0
Luzula multiflora (Ehrh.) Lej. N X X* 12 0
JUNCAGINACEAE Triglochin maritima L. N X – 2 0
LAMIACEAE Galeopsis bifida Boenn. E X* X* 2 4
Galeopsis tetrahit L. E X O 2 0
Lycopus americanus Muhl. ex W. Bart. N X – 4 0
Lycopus virginicus L. N X – 1 0
Mentha arvensis L. E X – 0 0
Scutellaria galericulata L. N X X 4 1
LILIACEAE Clintonia borealis (Aiton) Raf. N O X 0 27
Streptopus lanceolatus (Aiton) Reveal N X X 0 24
LINACEAE Linum radiola L. E O – 0 0
LYCOPODIACEAE Lycopodiella inundata (L.) Holub N X* – 0 0
Lycopodium clavatum L. N X O 1 0
LYTHRACEAE Lythrum salicaria L. E O – 0 0
MENYANTHACEAE Menyanthes trifoliata L. N O – 0 0
MYRICACEAE Comptonia peregrina (L.) Coult. N O – 0 0
Morella caroliniensis (P. Mill.) Small N X – 0 0
Myrica gale L. N X – 1 0
78 [Vol. 118
Rhodora

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
MYRSINACEAE Lysimachia borealis (Raf.) U. Manns & A.
Anderb.
NX X 23 22
Lysimachia maritima (L.) Galasso, Banfi &
Soldano
NX – 0 0
Lysimachia terrestris (L.) Britton, Sterns &
Poggenb.
NX – 3 0
ONAGRACEAE Chamerion angustifolium (L.) Scop. N X X 3 17
Circaea alpina L. N X X 5 1
Epilobium ciliatum Raf. N X X 2 2
Epilobium coloratum Biehler N O – 0 0
Epilobium leptophyllum Raf. N X – 0 0
Epilobium palustre L. N O – 0 0
Oenothera biennis L. N X – 0 0
Oenothera perennis L. N X – 1 0
ONOCLEACEAE Onoclea sensibilis L. N X – 0 0
OPHIOGLOSSACEAE Botrychium sp. N X* – 0 0
ORCHIDACEAE Cypripedium acaule Aiton N X – 1 0
Malaxis unifolia Michx. N X – 1 0
Platanthera clavellata (Michx.) Luer N X – 4 0
Platanthera lacera (Michx.) G. Don N X – 2 0
Pogonia ophioglossoides (L.) Ker-Gawl. N X – 0 0
Spiranthes cernua (L.) L.C. Rich. N X – 0 0
OROBANCHACEAE Euphrasia nemorosa (Pers.) Wallr. N X – 3 0
2016] 79
Negoita et al.—Flora of Duck Islands, Maine

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Euphrasia randii B.L. Robins. N X – 0 0
Euphrasia stricta D. Wolff ex J.F. Lehm. E O – 0 0
OSMUNDACEAE Osmundastrum cinnamomeum (L.) C. Presl N X X 5 1
OXALIDACEAE Oxalis montana Raf. N X X 17 25
Oxalis stricta L. N O O 0 0
PINACEAE Abies balsamea (L.) P. Mill. N X X 1 29
Picea glauca (Moench) Voss N X X 8 5
Picea rubens Sarg. N X X 18 16
PLANTAGINACEAE Hippuris vulgaris L. N O – 0 0
Plantago major L. E X X 0 0
Plantago maritima L. N X X 5 3
Veronica arvensis L. E – O 0 0
Veronica peregrina L. N O – 0 0
POACEAE Agrostis scabra Willd. N X – 5 0
Agrostis sp. N/E X X 22 2
Agrostis stolonifera L. E X* X 0 0
Alopecurus pratensis L. E X* – 1 0
Anthoxanthum nitens (Weber) Y. Schouten &
Veldkamp
NX – 1 0
Anthoxanthum odoratum L. E X – 2 0
Calamagrostis canadensis (Michx.) Beauv. N X X* 7 0
Cinna latifolia (Trev. ex Goepp.) Griseb. N – X* 0 2
Dactylis glomerata L. E O – 0 0
80 [Vol. 118
Rhodora

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Danthonia spicata (L.) Beauv. ex Roemer & J.A.
Schultes
NO – 0 0
Deschampsia flexuosa (L.) Trin. N X – 10 0
Dichanthelium boreale (Nash) Freckmann N X* – 0 0
Elymus repens (L.) Gould E X* X 3 21
Elymus trachycaulus (Link) Gould ex Shinners N – O 0 0
Elymus virginicus L. N X X* 1 4
Festuca rubra L. N/E X X 29 9
Hordeum jubatum L. N O – 0 0
Panicum sp. N/E O – 0 0
Phleum pratense L. E X X* 5 0
Poa alsodes Gray N – X* 0 0
Poa compressa L. E O X 0 0
Poa palustris L. N – X 0 4
Poa pratensis L. N/E X 17 29
Poa trivialis L. E – X* 0 0
Schedonorus pratensis (Huds.) Beauv. E O – 0 0
Spartina pectinata Link N O – 0 0
POLYGONACEAE Fallopia cilinodis (Michx.) Holub N – X 0 0
Fallopia convolvulus (L.) A. L ¨
ove E – X* 0 0
Fallopia scandens (L.) Holub N – O 0 0
Polygonum aviculare L. E O O 0 0
Polygonum buxiforme Small N X* X* 1 0
2016] 81
Negoita et al.—Flora of Duck Islands, Maine

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Rheum rhabarbarum L. E X* – 0 0
Rumex acetosella L. E X X 20 5
Rumex crispus L. E X X 6 5
Rumex longifolius DC. E – X* 0 4
Rumex obtusifolius L. E – X 0 0
Rumex pallidus Bigelow N – O 0 0
POLYPODIACEAE Polypodium virginianum L. N – X 0 11
PORTULACACEAE Montia fontana L. N O – 0 0
RANUNCULACEAE Coptis trifolia (L.) Salisb. N X X 9 0
Ranunculus acris L. E X X 9 1
Ranunculus bulbosus L. E – O 0 0
Ranunculus cymbalaria Pursh N O – 0 0
Ranunculus repens L. E O – 0 0
Thalictrum pubescens Pursh N X X 3 4
ROSACEAE Amelanchier laevis Wiegand N O X* 0 0
Amelanchier sp. N X* X 0 4
Argentina anserina (L.) Rydb. N O O 0 0
Argentina egedii (Wormsk.) Rydb. N X* X* 3 0
Aronia floribunda (Lindl.) Spach N X – 0 0
Aronia melanocarpa (Michx.) Elliott N X – 1 0
Fragaria virginiana Duchesne N X X 25 20
Malus pumila P. Mill. E X X 0 0
Potentilla argentea L. E O – 0 0
82 [Vol. 118
Rhodora

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Potentilla argentea L. E O – 0 0
Potentilla canadensis L. N O O 0 0
Potentilla norvegica L. N X X 0 3
Potentilla simplex Michx. N X – 8 0
Prunus pensylvanica L. f. N – X 0 0
Rosa nitida Willd. N X – 3 0
Rosa rugosa Thunb. E X X* 3 0
Rosa sp. (cultivated) N/E O – 0 0
Rosa virginiana P. Mill. N X* – 1 0
Rubus hispidus L. N X – 17 0
Rubus idaeus L. N/E X X 25 33
Rubus pubescens Raf. N X X 7 0
Sibbaldiopsis tridentata (Aiton) Rydb. N X – 0 0
Sorbus americana Marshall N X X 0 0
Sorbus aucuparia L. E – O 0 0
Sorbus decora (Sargent) Schneid. N – X* 0 0
Sorbus sp. N/E X* X 20 26
Spiraea alba Du Roi N X – 7 0
Spiraea tomentosa L. N X* – 0 0
RUBIACEAE Galium aparine L. N X X 1 3
Galium trifidum L. N O – 0 0
Galium triflorum Michx. N X X* 1 1
RUPPIACEAE Ruppia maritima L. N O – 0 0
2016] 83
Negoita et al.—Flora of Duck Islands, Maine

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
RUSCACEAE Maianthemum bifolium (L.) F.W. Schmidt N O – 0 0
Maianthemum canadense Desf. N X X 37 32
Maianthemum trifolium (L.) Sloboda N X – 2 0
SALICACEAE Populus grandidentata Michx. N – O 0 0
Populus tremuloides Michx. N – X 0 1
Salix bebbiana Sargent N – O 0 0
Salix discolor Muhl. N – X 0 0
Salix sp. N/E X – 0 0
SAPINDACEAE Acer pensylvanicum L. N – X 0 25
Acer platanoides L. E – X* 0 0
Acer rubrum L. N X X 2 0
Acer spicatum Lam. N – X 0 26
SELAGINELLACEAE Selaginella selaginoides (L.) P. Beauv. ex Mart.
& Schrank
N– O 0 0
SOLANACEAE Solanum dulcamara L. E X X 2 8
Solanum nigrum L. E O – 0 0
THELYPTERIDACEAE Thelypteris palustris Schott N O – 0 0
TYPHACEAE Typha latifolia L. N X – 0 0
URTICACEAE Urtica dioica L. N/E X* X 2 8
Urtica sp. N/E O – 0 0
VIOLACEAE Viola blanda Willd. N – X 0 0
Viola cucullata Aiton N – O 0 0
Viola lanceolata L. N X – 3 0
84 [Vol. 118
Rhodora

APPENDIX.Continued.
Family Scientific Name Nativ. GDI 2011 LDI 2010
Frequency
GDI Plot LDI Plot
Viola pallens (Banks ex DC.) Brainerd N X – 0 0
Viola sororia Willd. N X – 0 0
Viola sp. N/E X X 17 5
WOODSIACEAE Athyrium angustum (Willd.) C. Presl. N – X 0 0
Deparia acrostichoides (Sw.) M. Kato N – X 0 0
Gymnocarpium dryopteris (L.) Newman N – X* 0 0
2016] 85
Negoita et al.—Flora of Duck Islands, Maine