Applied Herpetology 6 (2009) 343353

www.brill.nl/ah

Observations of oviposition in northern clade Bufo fowleri

in kettle lakes at Cape Cod National Seashore, USA:
implications for management
Todd A. Tupper 1 , Megan D. McLean 2 and Scott W. Buchanan 3
1

Northern Virginia Community College, Alexandria Campus, Department of Science and

Applied Technology, 3001 North Beauregard Street, Alexandria, Virginia 22311, USA
Corresponding author; e-mail: ttupper@nvcc.edu
2
Antioch University New England, Department of Environmental Studies, 40 Avon Street, Keene,
New Hampshire 03431-3516, USA
3
United States Department of the Interior, National Park Service, Cape Cod National Seashore,
99 Marconi Site Road, Wellfleet, Massachusetts 02667, USA

Abstract
Kettle lakes are important breeding habitats for Fowlers toad (Bufo fowleri) and are also used for human
recreation. We sampled four kettle lakes used for recreation at Cape Cod National Seashore, Barnstable
County, Massachusetts, USA to examine patterns in B. fowleri oviposition. Most oviposition occurred within
4 m of shorelines in shallow water and on vegetation or debris that positioned embryos close to the waters
surface. Only 1.22% of egg strings were located in recreational-use accessways. Public-use related habitat
disturbance likely inhibits B. fowleri oviposition due to the removal of attachment substrates. Resource managers should consider restricting human activity in kettle lakes so that egg strings and attachment substrates
are not destroyed.
Koninklijke Brill NV, Leiden, 2009
Key words
Anuran conservation, Bufo fowleri, Cape Cod National Seashore, egg strings, oviposition, Southern New
England.

Introduction
Fowlers toad, Bufo fowleri (proposed Anaxyrus fowleri, by Frost et al., 2006; we
have maintained Bufo because of its long established use in the literature), is a
subclimax community habitat specialist that breeds in open-canopy permanent and
non-permanent wetlands (Tupper and Cook, 2008). In much of the northeastern
part of its range, its preferred terrestrial habitats are intensely developed and heavily populated. Northern clade (Masta et al., 2002) B. fowleri are federally protected
in Canada (Green, 2005) and local extirpations have been reported in certain areas
Koninklijke Brill NV, Leiden, 2009

DOI:10.1163/157075309X12470350858514

344

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

of the northeastern United States (Klemens, 1993; Tupper and Cook, 2008). Shallow sections of kettle lakes (permanent bodies of water existing in glacial outwash
plains formed from retreating glaciers that are not connected to streams and are
fed by groundwater and precipitation; Galvin, 2008) are among the most important breeding habitats (Tupper and Cook, 2008) in southern New England, because
they support vocalization, oviposition, larval development and annual recruitment
(Green, 2005; Tupper and Cook, 2008). In certain areas (particularly U.S. Parklands) kettle lakes are also used for recreational activities such as fishing, boating,
wading and swimming. Over time, these activities convert naturally-vegetated sections of the lake littoral zone into patches of bare sand that lack aquatic vegetation,
leaf litter and deadwood. Size and proportion of lake littoral zone occupied by these
public use areas varies, depending on intensity of human activity.
Though much is known about the natural history (Green, 2005) and reproductive ecology of B. fowleri (Tupper et al., 2007; Tupper and Cook, 2008), little
is known about its oviposition patterns within kettle lakes. Such knowledge may
be relevant to managing lakes used for human recreation and a lack of such information could potentially lead to reduced annual recruitment at wetlands that
serve as critical nurseries for amphibians. Here we use a series of descriptive statistics and hypothesis tests to describe oviposition patterns in B. fowleri at four
kettle lakes used for recreation at Cape Cod National Seashore, Barnstable County
Massachusetts. Our goals were to determine if a casual relationship existed between cardinal direction, vegetation distribution and egg string distributions in
kettle lakes. We broadly hypothesized that public accessways (areas repeatedly
disturbed by human activity) would contain significantly fewer egg strings than
non-disturbed areas. Our data raise important considerations for future research and
potential management of breeding habitats critical to the long-term persistence of
this species.
Methods
Study sites
We selected four kettle lakes (Duck [NAD 83: 19T 417059, 4642806], Dyer [NAD
83: 19T 416317, 4643467], Kinnacum [NAD 83: 19T 417249, 4644995] and Spectacle Ponds [NAD 83: 19T 417457, 4644851]) for egg string sampling. Selection
was based on prior documentation of B. fowleri breeding aggregations and recreational use of the lakes (Tupper et al., 2008; NPS unpublished data). All sites
were located within Cape Cod National Seashore (see Tupper et al., 2007 for
a detailed description of Cape Cod National Seashore) in the town of Wellfleet,
Barnstable County, Massachusetts. Selected lakes ranged from 0.65 ha to 4.91 ha
(Mean = 3.24; SE = 0.97). With the exception of shoreline vegetation, lakes were
open-canopy and sparsely vegetated. Of the vegetation within-wetland, leatherleaf (Chamaedaphne calyculata) was most common and restricted to shorelines,
though some graminoids, herbaceous emergents, and floating-leaved macrophytes

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

345

were also common (NPS unpublished data). Lake shorelines (perimeters) were predominantly undisturbed, but each site contained at least one well defined public
accessway. Public accessways accounted for 12% (94/802 m), 8% (71/880 m)
10% (13/455 m) and 3% (8/304 m) of total shoreline lengths at Duck, Dyer,
Kinnacum and Spectacle Ponds (these sites are colloquially termed, ponds but are
actually lakes; Galvin, 2008), respectively. Water was tannin-lignin poor and mildly
acidic. Based on a 25 year data set, mean pH values were: 4.83 (95% CI = 0.04)
at Duck Pond, 4.85 (95% CI = 0.05) at Dyer Pond, 4.65 (95% CI = 0.05) at Kinnacum Pond and 5.05 (95% CI = 0.04) at Spectacle Pond. Each kettle lake has a
maximum depth > 2.5 m (Tupper and Cook, 2008; NPS unpublished data).
Data collection
We used methods described by Cook and Boland (2005) to locate egg strings within
10 m of the lakes shoreline during known peaks in breeding season (Tupper
et al., 2007) in 2008. We completed one round of counts in each lake toward the
end of the ovipositon period (10 June-25 June) to help ensure that the majority of
ovipositon occurred. Recently hatched strings were swollen and opaque. Consequently, they were conspicuous and also counted. To quantify trends in oviposition,
we recorded attachment substrate, attachment depth (furthest vertical distance of
egg string from lake basin in cm), maximum water column depth (cm), distance
from shore (cm), cardinal direction section, accessway status (string in a public
accessway or not) and canopy cover (binary data [open/closed], via a periscope
densitometer) upon locating egg strings. Cardinal direction sections (or slices of
pie) were approximately equal-sized, wedge-shaped areas that originated at the
center of the lake and radiated outward toward each of the 8 cardinal directions
(N, NE, E, SE, S, SW, W, NW). These sections were calculated with ArcGIS version 9. Attachment substrates were identified as one of the following types: bare
substrate, herbaceous emergent vegetation, woody emergent vegetation, floatingleaved macrophyte, submerged aquatic vegetation, graminoid and litter (Tupper and
Cook, 2008). We visually estimated percent (using Braun-Blanquet cover classes;
Wilkum and Shanholtzer, 1978) of each cover class within 1 m 1 m plots around
egg string loci and in plots that were selected via a stratified-random scheme. This
scheme allowed us to randomly sample 5% of the total shoreline within each directional section. The locations of the random plots were chosen based on randomly
generated meter intervals along the perimeter of each directional section and within
5 m (toward the center of the lake) of the shore. We used 5 m as a guideline for
selection to avoid oversampling in areas where there is little to no oviposition.
We truncated cover classes into four categories (woody, non-woody, bare, litter)
when examining differences in mean Braun-Blanquet cover classes between random plots and egg string plots. To be certain that annual variation (Ficetola et al.,
2006) was not causing spurious results, we conducted pilot counts at Duck pond
the previous summer (2007) and compared number of eggs strings in each direc-

346

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

tional sections between years. No differences were found (Wilcoxon signed-rank

test; Z = 0.700; P > 0.05; Spearmans rho correlation; r = 0.884; P < 0.05).
Data analyses
Differences between data collected in random plots and egg string plots were assessed with Mann-Whitney U-tests, independent sample t-tests and Pearson Chisquare statistics (Zar, 1999). We also used a Pearson Chi-Square statistic to determine if anuran oviposition was evenly distributed between directional sections. We
used a Spearmans correlation coefficient (Zar, 1999) to examine the correlation between directional section perimeter length and number of egg strings detected and
the correlation between relative amount of vegetative substrate available in each
lake and the amount of egg strings attached to a given substrate.
Results
Descriptives
We documented 899 egg strings; 532 (59%), 164 (18%), 136 (15%) and 67
(7%), at Dyer, Duck, Kinnacum and Spectacle Ponds, respectively. Egg strings
were located from 0.08-10 m from the shore, but were predominantly within a few
meters of shore (fig. 1; Mean = 2.889; standard error (SE) = 7.657), though this
was likely in part influenced by detectability. Some egg strings were attached to substrate in deeper water (up to 1.9 m deep), but attachment sites were usually near the
waters surface (distance below surface: Mean = 0.076 m; SE = 0.006 m) in shallow water (Mean water depth = 0.414 m; SE = 0.008 m). Egg strings were attached
to woody and herbaceous emergent vegetation, submerged aquatics, floating-leaved
macrophytes and leaf litter, but were rarely found on bare substrates (fig. 2). Five

Figure 1. Percent of egg strings found in distance (from shoreline) categories.

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

347

Figure 2. Percent of egg strings found in attachment substrate categories. Black bars represent
Duck Pond, grey bars represent Dyer Pond, stippled bars represent Kinnacum Pond and white bars
represent Spectacle Pond. BARE = non-vegetated substrates, FLM = floating-leaved macrophytes,
GRAM = graminiods, HERB = herbaceous emergent vegetation, SAV = submerged aquatic vegetation, WOODY = woody emergent vegetation.

hundred eighteen of 899 (58%) of egg strings were under an open canopy. With
the exception of Duck Pond, relationships between distribution of attachment substrate and direction were not apparent. One hundred percent of public use areas
were bare and contained only 11 of 899 (1.22%) of the total egg strings detected.
Hypothesis tests
When we pooled data from all lakes and truncated attachment classes, we found
significantly fewer egg strings ( 2 = 3, 858; df = 3; P < 0.001) attached to bare
surfaces (0.11%) than non-woody vegetation (41.9%), woody vegetation (32.5%)
and substrate litter (25.3%). We found significantly less woody vegetation (Z =
6.413; P < 0.001), non-woody vegetation (Z = 4.998; P < 0.001) and substrate litter (Z = 6.281; P < 0.001) in random plots than in plots with egg strings.
Egg string plots occurred significantly closer to the shore (t = 3.141; P < 0.05)
than random plots. As a result, random plots contained significantly fewer positive
(closed) densiometer readings than egg string plots ( 2 = 9.718; df = 1; P < 0.05).
We found significantly more bare substrate (Z = 5.916; P < 0.001) in random
plots, but no apparent differences in water depth (t = 1.332; P > 0.05) (fig. 3).
We found a significant, positive relationship between number of egg strings detected per vegetation attachment substrate in each lake and mean Braun-Blanquet
index value recorded for each vegetative (non-bare categories) cover type in each
lake (r = 0.628; P < 0.05), indicating that oviposition occurred on the more abundant attachment substrates.
Egg strings were not equally distributed among directional sections within each
lake (fig. 3). Duck ( 2 = 36.096; df = 7; P < 0.05) and Spectacle Ponds ( 2 =

348

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

Figure 3. Comparison of habitat variables between random plots and those that contained egg strings
at all study sites. Black bars represent random plots and gray bars represent egg string plots. Error
bars represent the 95% confidence interval. Habitat variables (with the exception of depth and distance
from shore) are mean Braun-Blanquet index values. Bare = non-vegetated substrates, Dist = distance
in meters from the shore, Herb = herbaceous emergent vegetation, submerged aquatic vegetation and
graminoids, Litter = deadwood, leaves/needles and other nonliving debris, Woody = woody emergent
vegetation.

21.518; df = 7; P < 0.05) had significant differences in number of egg strings

found in each directional section, whereas Dyer ( 2 = 5.968; df = 7; P > 0.05) and
Kinnacum Ponds ( 2 = 10.313; df = 7; P > 0.05) did not. The trends within lakes
did not appear consistent between lakes (fig. 4). Directional section perimeter and
number of egg strings detected were positively correlated (r = 0.545; P = 0.05).
We found positive correlations between the amount of shoreline vegetation at Duck
(r = 0.951; P < 0.001) and Dyer Ponds (r = 0.786; P > 0.05) and the number
of egg strings in each directional section, but we were unable to demonstrate this
relationship at Kinnacum (r = 0.071; P > 0.05) and Spectacle Ponds (r = 0.395;
P > 0.05) (fig. 3).
Discussion
Attachment substrates, canopy cover and cardinal direction
Unlike amphibian species that show a preference for particular attachment substrates (Jakob et al., 1998), B. fowleri appeared to use the most frequently occurring
shoreline leaf litter and vegetation as attachment substrates. Lake areas that were
devoid of vegetation typically did not contain egg strings. Egg strings were located
predominately within 4 m of kettle lake shorelines in shallow areas just beneath
the waters surface. The few egg strings that were found in deeper areas of the lake
were still attached to substrates and positioned near the waters surface. Egg strings
are more difficult to locate in deeper water, so detectability biases may have influenced our results. However, our sites had clear water and we sampled on sunny days
with polarized glasses, so detectability biases are likely minimal. Additionally, our

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

349

(a)

(b)

Figure 4. Distribution of egg strings and vegetation in directional sections in sampled kettle lakes at
Cape Cod National Seashore: (a) Duck Pond, (b) Dyer Pond, (c) Kinnacum Pond, (d) Spectacle Pond.
Black bars represent total numbers of eggs strings and gray bars represent mean % vegetation per
directional section. Error bars represent the 95% confidence interval.

results are supported by literature indicating a biological necessity for oviposition

attachment substrates. Attachment substrates elevate eggs higher in the water column. This elevation restricts predation by benthic predators (Howard, 1978; Egan
and Paton, 2004), prevents drifting into thermally unsuitable water (Licht, 1971;
Ficetola et al., 2006), and provides a proximal food source for larvae (Hartel et
al., 2008). Thus, attachment substrates may be actively selected during chorusing
(Sexton and Ortleb, 1996).
Canopy cover did not appear to be a major factor influencing ovipositon site selection in our study because all sites were predominantly open-canopy. We know
that dense canopy cover can limit the reproductive success of various amphibian

350

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

(c)

(d)

Figure 4. (Continued).

species (Skelly et al., 2005), including B. fowleri (Tupper and Cook, 2008) and
B. americanus (Klemens, 1993). However, most of the vegetation that contributed
to positive densiometer readings did not extend far enough into the lakes to create
permanent shade over egg strings. Thus, the canopy was likely not dense enough
to negatively affect embryonic development (see Tupper and Cook, 2008 regarding breeding effort of B. fowleri and canopy cover). We suggest that B. fowleri
chose oviposition sites where ample attachment substrate existed without regard
to canopy; as we found a near 50:50 ratio of eggs strings under canopy to those
that were not. Though we suggest canopy did not play a large role in oviposition site selection in this study, it likely plays a greater role in more structurally
heterogeneous wetlands with more extensive canopy cover and varying hydroperiods.

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

351

Though each lake showed an uneven distribution of egg strings in cardinal direction areas, the patterns were not consistent between lakes. We partially attribute
the uneven distributions of egg strings to the amount of available shoreline vegetation per directional section. At both Duck and Dyer Ponds, the two sites with the
largest bare and contiguous public accessways, the amount of available shoreline
vegetation appeared to play a role in egg string distribution. However, Spectacle
and Kinnacum Ponds have very small, and non-contiguous public accessways, so
shoreline vegetation is more evenly distributed throughout directional sections at
these sites. Because vegetation is not limited at these latter sites, it is likely that
factors driving oviposition site selection were not detectable by methods used in
this study. These factors may include aspects such as intraspecifc interactions and
the composition of upland habitats immediately surrounding breeding sites. While
B. fowleri tend to prefer open canopy and sparsely vegetated wetlands for breeding
(Tupper and Cook, 2008), the integrity of existing within-wetland shoreline vegetation appears to be an important limiting factor in B. fowleri oviposition in kettle
lakes; if there are large areas devoid of vegetation, these areas are also devoid of
eggs strings.
Management implications and limitations
Our results have several important management implications. As suggested by Hartel et al. (2007), and observed in this study, human activity can lead to the loss
of lake vegetation used as attachment substrates. Therefore, human activity may
need to be restricted temporally and spatially to prevent the excessive loss of suitable attachment substrates and disturbance of egg strings. While B. fowleri at Cape
Cod National Seashore has a breeding season that ranges from late April to late
July (Tupper et al., 2007), the beginning of the season (April 19-1 June) is the
most crucial time to restrict human recreational activity because amplectant pairs
of B. fowleri are most frequently observed during the first half of intense calling
(Clarke 1974; pers. obs).
This study is limited in its assessment of between-year variation in egg string
distributions. However, our 2-year examination of Duck Pond yielded no significant differences in oviposition patterns between years. Ficetola et al. (2006) and
Gollmann et al. (2002) suggest that distributions of oviposition sites in ephemeral
wetlands shift between years due to changes in hydroperiod and vegetation. Annual
shifts in kettle lakes are likely to be less pronounced because basins permanently
contain water. Our work was also limited to four kettle lakes. This limitation is due
to sampling a 10 m area around the entire perimeter of each lake. Our method of
searching was time consuming. Had we increased the sample size, it would have
been impossible to complete searches during periods of B. fowleri oviposition. We
recommend that future inquiry consider limiting sampling to a number of transects
within each wetland, while increasing the number of wetlands sampled for more
broadly applicable results. Ideally, future inquiry should include an assessment of

352

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

immediately adjacent upland habitat features because certain upland features may
influence within-wetland oviposition site selection by determining amphibian migration routes and therefore where individuals enter wetlands (Hartel, 2004). Future
work should also determine the specific cause(s) of egg-barren accessways and develop a better understanding of how it affects annual recruitment.
Conclusions
We found that B. fowleri egg strings were most frequently attached to common
shoreline vegetation and were rarely located on bare substrates. All public use areas
were devoid of vegetation and contained only 1.22% of total detected egg strings.
Regardless of the mechanism for human disturbance on oviposition, we recommend that human activity be limited in kettle lakes that support breeding to reduce
interference with oviposition. A more specific management plan is necessary and
should be based on future studies with larger sample sizes that examine attachment
substrate preferences, egg string distributions and public-use related effects on recruitment in greater detail.
Acknowledgements
We would like to thank the following people for contributing to the development of this manuscript: Robert
Aguilar, Safra Altman, Robert P. Cook, Nancy Finley, Rebecca Flaherty, Michelle Galvin, Rebecca Gawley,
Mac Given, Sasha Greenspan, Bill Hamilton, Peter Paton, Carrie Phillips, Fiana Shapiro, Megan Sporndle,
Brett Thelen, Brad Timm and Mike Tyler.

References
Blaustein, L., Garb, J.E., Shebitz, D., Nevo, E. (1999): Microclimate, developmental plasticity and
community structure in artificial temporary pools. Hydrobiology 392: 187-196.
Clark, R.D. (1974): Activity and movement patterns in a population of Fowlers toad, Bufo woodhousei
fowleri. Am. Midl. Nat. 91: 257-274.
Cook, R.P., Boland, K. (2005): A comparison of approaches to counting spotted salamander (Ambystoma maculatum) egg masses in vernal ponds. Herpetol. Rev. 36: 272-274.
Egan, R.S., Paton, P.W.C. (2004): Within-pond parameters affecting oviposition by wood frogs and
spotted salamanders. Wetlands 24: 1-13.
Ficetola, G.F., Valota, M., de Bernardi, F. (2006): Temporal variability of spawning site selection in
the frog Rana dalmatina: consequences for habitat management. Anim. Biol. Cons. 29: 157-163.
Ficetola, G.F., Valota, M., de Bernardi, F. (2006): Within-pond spawning site selection in Rana dalmatina. In: Atti del V Congresso Nazionale Societas Herpetologica Italica, p. 113-116. Zuffi,
M.A.L., Ed., Firenze, Firenze University Press.
Grant, T., Faivovich, J., Bain, R.H., Haas, A., Haddad, C.F.B., De Sa, R.O., Channing, A., Wilkinson, M., Donnellan, S.C., Raxworthy, C.J., Campbell, J.A., Blotto, B.L., Moler, P., Drewes, R.C.,
Nussbaum, R.A., Lynch, J.D., Green, D.M. and Wheeler, W.C. (2006): The amphibian Tree of
Life. Bull. Am. Mus. Nat. Hist. 297: 1-370.
Galvin, M. (2008): Pond and Lake Ecology [presentation]. In: Ponds in Peril Workshop; 2008 Jan 24;
West Yarmouth (MA): Association to Preserve Cape Cod.
Gollmann, G., Gollmann, A., Baumgartner, C., Waringer-Loschenkohl, A. (2002): Spawning site
shifts by Rana temporaria in response to habitat change. Biota 3: 35-41.

T.A. Tupper et al. / Applied Herpetology 6 (2009) 343353

353

Green, D.M. (2005): The Fowlers toad. In: Amphibian Declines: The Conservation of United States
Species, p. 408-412. Lannoo, M., Ed., University of California Press, Berkeley and Los Angeles,
California.
Hartel, T. (2004): The long term trend and the distribution of amphibian populations in a semi-natural
pond in the middle section of the Tarnava-Mare Valley (Romania). Biota 5: 25-36.
Hartel, T. (2008): Long-term within pond variation of egg deposition sites in the Agile frog, Rana
dalmatina. Biologia 63: 439-443.
Hartel, T., Nemes, S., Cogalniceanu, D., Ollerer, K., Schweiger, O., Moga, C.I., Demeter, L. (2007):
The effect of fish and aquatic habitat complexity on amphibians. Hydrobiology 583: 173-182.
Hartel, T., Nemes, S., Cogalniceanu, D., Ollerer, K., Moga, C.I., Lesbarreres, D., Demeter, L. (2009):
Pond and landscape determinants of Rana dalmatina population sizes in a Romanian rural landscape. Acta. Oecol. 35: 53-59.
Howard, R. (1978): The influence of male-defended oviposition sites on early embryo mortality in
bullfrogs. Ecology 59: 789-798.
Jakob, C., Hoffmann, O., Braun, C., Crivelli, A.J., Veith, M. (1998): Characteristics of intra-pool egg
deposition sites of Triturus marmoratus (Latreille, 1800) and T. helveticus (Razoumowski, 1789).
In: Current Studies in Herpetology, p. 215-219. Miaud, C., Guyetant, G., Eds, Le Bourget du Lac,
France, S.E.H.
Klemens, M.W. (1993): Amphibians and reptiles of Connecticut and adjacent regions. Connecticut
State Geological and Natural History Survey, Hartford, Connecticut.
Licht, L.E. (1971): Breeding habits and embryonic thermal requirements of the frogs, Rana aurora
aurora and Rana pretiosa pretiosa, in the Pacific Northwest. Ecology 52: 116-124.
Seale, D.A. (1982): Physical factors influencing oviposition by the wood frog, Rana sylvatica, in
Pennsylvania. Copeia 1982: 627-635.
Sexton, O.J., Ortleb, E.P. (1966): Some cues used by the leptodactylid frog, Engystomops pustulosus,
in selection of the oviposition site. Copeia 1996: 225-230.
Skelly, D.K., Halverson, M.A., Freidenburg, L.K., Urban, M.C. (2005): Canopy closure and amphibian diversity in forested wetlands. Wetlands Eco. Manage. 13: 261-268.
Tupper, T.A., Cook, R.P., Timm, A.C., Goodstine, A. (2007): Improving call surveys for detecting the
Fowlers toad, Bufo fowleri, in Southern New England, USA. Appl. Herpetol. 4: 245-259.
Tupper, T., Cook, R.P. (2008): Habitat variables influencing breeding effort in northern clade Bufo
fowleri: Implications for conservation. Appl. Herpetol. 5: 101-119.
Winkum, D.A., Shanholtzer, G.F. (1978): Application of the Braun-Blanquet cover-abundance scale
for vegetation analysis in land development studies. Environ. Manage. 2: 323-329.

Accepted: 29 May 2009.