Northeast Natural History Conference 2011: Selected Papers
2012 Northeastern Naturalist 19(Special Issue 6):67–76
The Distribution of Cordulegaster
(Odonata: Cordulegastridae) Nymphs in Seeps and
Springs of Nelson Swamp (Madison County, NY)
Barbara J. Hager1,*, Nina J. Kalantari1, and Van A. Scholten1
Abstract - Given the presence of foraging and reproducing adult Cordulegaster (spiketail)
dragonflies in Nelson Swamp (Madison County, NY), we examined nymph distribution
and abundance in the seeps and springs found within the swamp. From 9 September–4
November 2010, we surveyed 8 sites along Chittenango Creek in order to determine:
(1) the species present and their distribution/occurrence among sites, (2) factors influencing
species presence and abundances, and (3) patterns in size and age distribution among
and within sites. For sites, we delineated habitat zones (inlet, middle, outlet), determined
the benthic substrate, and measured shoreline perimeters. For nymphs, we measured head
width, body length, and wing pad length and identified some to species. The majority of
spiketails we identified were Cordulegaster diastatops (Delta-spotted Spiketail); Cordulegaster
maculata (Twin-spotted Spiketail) was also present. Most nymphs occurred
in inlets with muck and cobble bottoms and in water depths less than 10 cm. Spiketail
densities ranged from 0.13–8.13 individuals/m of shoreline. Smaller individuals occurred
in cobble substrate, while muck substrates had individuals of larger size and greater
abundance. We demarcated at least 2 age cohorts of nymphs based on their body measurements
in relation to growth patterns observed in other spiketail species.
Introduction
The 8 North American species of Cordulegaster (spiketail) dragonflies are
geographically widespread yet only locally common inhabitants of streams,
springs, and seeps (Nikula et al. 2007). In a review of Odonata literature, Corbet
(1999) concluded that spiketail nymphs require microhabitats with particular
sediment particle sizes—to permit the shallow burrowing by nymphs of different
sizes—and with water flows of not more than 10–15 cm/sec. Recent studies of
Cordulegaster erronea Hagen (Tiger Spiketail), Cordulegaster dorsalis Hagen
(Pacific Spiketail), Cordulegaster sayi Selys (Say’s Spiketail), and Cordulegaster
maculata Selys (Twin-spotted Spiketail) confirm that current, sediment
size, and sediment composition influence presence and abundance (Burcher and
Smock 2002, Glotzhober 2006, Marczak et. al. 2006, Stevenson et al. 2009).
Thus, seeps and springs running into lentic waters as well as occurring alongside
streams and rivers may support populations of spiketails.
From 2006–2010, one of us (B.J. Hager) observed numerous adult Cordulegaster
diastatops Selys (Delta-spotted Spiketail) and a few Twin-spotted
Spiketails foraging along field edges at our study area (B.J. Hager, unpubl. data).
Likewise, many Delta-spotted Spiketails were observed engaging in reproductive
1Environmental Studies Program, Cazenovia College, 22 Sullivan Street, Cazenovia, NY
13035. *Corresponding author - bhager@cazenovia.edu.
68 Northeastern Naturalist Vol. 19, Special Issue 6
activities in both the seeps and springs beside the branch of Chittenango Creek
running through Nelson Swamp, and also in the creek itself. At the same time,
only one Twin-spotted male was seen patrolling along Chittenango Creek itself
and at the spring that was nearest the creek bed (site 6; Table 1). During the fall
of 2010, we set out to examine spiketail nymph distribution within and among
the seeps and springs at this locale. In particular, we focused on the following
questions: (1) How are spiketail nymphs distributed among the individual seeps
and springs? (2) Do the nymphs appear to prefer different microhabitats or substrates?
(3) What are the sizes and abundances of the spiketail nymphs in the
various seeps and springs?
Methods
Study area description
Nelson Swamp Unique Area is a 607-ha (1500-acre) cedar/hemlock swamp
located in west central Madison County, NY (roughly 42°53'N, 75°47'W). Two
branches join to form the main channel of Chittenango Creek, which flows through
the westernmost portion of the swamp (Fig. 1; NYSDEC Region 7 Management
Team 2000). Classified as a fen by the New York State Natural Heritage Program,
the area adjacent to the creek is underlain by Wayland silt-loam soil, and, upslope,
the adjoining Carlisle-Palms muck soils (NYSDEC Region 7 Management
Figure 1. Location of Madison County in New York State (shaded county on state map) and
the seeps and springs (dashed line encircling creek) sampled along Chittenango Creek (CC)
in Nelson Swamp Unique Area. A band of trees, ranging in width from 10–30 m, stretched
north away from the creekbed. The dot indicates an open field where many teneral and
adult Cordulegaster diastatops and a few Cordulegaster maculata occurred. LT indicates
the Link Trail of Nelson Swamp and CBR stands for Constine Bridge Road.
2012 B.J. Hager, N.J. Kalantari, and V.A. Scholten 69
Team 2000). These soil types are well saturated year-round, and it’s likely that
groundwater percolating from both of them contributes to the seeps and springs
lining the creek. In the area we studied (northern side of the main branch of the
creek near Constine Bridge Road, Town of Cazenovia; Fig. 1), Tsuga canadensis
(L.) Carr (Eastern Hemlock) and Thuja occidentalis L. (Northern White Cedar)
form the prominent canopy cover, while mats of Callitriche sp. (water starwort)
dominate the aquatic vegetation.
Seep/spring data
We sampled 8 seeps and springs located along the northern edge of the main
stem of Chittenango Creek. We distinguished seeps from springs by assessing
whether the groundwater flowed from a single source (a spring) or appeared to
have no source or multiple, very small trickle sources spread over a large area
(a seep) (Dodson 2005). For each site, we measured the shoreline perimeter, delineated
the inlet, middle, and outlet zones, and measured water depths near the
shore (within 15 cm) in the inlets and outlets (when applicable) and at the deepest
point. For each zone of a site, we classified the benthic substrate as large (>10
cm) or small (<10 cm) cobble, sand, muck (fine, black organic matter), clay, and
vegetation (Table 1). The most common substrates—cobble and muck/sand—
were later classified into cobble, muck, and mixed (cobble/muck) for analysis.
Lastly, we recorded the presence of frogs (Lithobates clamitans Latreille [Green
Frog]), fish (mostly Rhinichthyes spp. [daces]), and salamanders (Desmognathus
spp.) encountered during our surveys.
Spiketail sampling methods
From 9 September to 4 November 2010, we surveyed each spring and seep for
nymphs by systematically sifting the substrate by hand. All but two of the sites
(sites 4 and 5) were sampled exhaustively for nymphs. The substrate of these latter
two sites was predominantly clay and large cobble with a water depth greater
than 0.5 m; therefore, we could not remove much material for sifting, although
we did feel among the cobble for the nymphs.
We placed the nymphs into separate shallow, water-filled pans corresponding
to each zone. Using calipers, we measured head width, body length, and wing
pad length of each spiketail to the nearest 0.05 mm. After processing, we released
most of the nymphs where they were collected, except for those brought back to
the laboratory for species identification. We brought 75 late instar (with headwidths
greater than 5.5 mm) and 21 early instar spiketails back to the laboratory
for species identification using the nymphal keys from Bright and O’Brien (1999)
and Needham et al. (2000). These keys rely on 2 major characteristics of late-stadium
nymphs to distinguish C. maculata from C. diastatops—frontal shelf shape
and setae and epaulet shape; thus, we examined the early instar individuals to see
if we could use the keys for individuals this young. Based on our observations,
the differences in frontal shelf and epaulet shape of younger-stadium individuals
were not as obvious as in older specimens, especially when trying to observe living
individuals. Due to this level of uncertainty in identification, we decided not
70 Northeastern Naturalist Vol. 19, Special Issue 6
to include these individuals in our counts of determined species identification.
All specimens were released the following week.
Results
We found a total of 296 nymphs, ranging in body length from 6.30 to 35.35
mm (Fig. 2), at 7 of the 8 sites (Table 1). The two sites with 0 or 1 nymph captured
had clay-based bottoms, shallow or limited fine sediment, and extensive areas
covered with large cobble. Of the 75 larger (head width > 5.5 mm and body length
> 20 mm) individuals identified to species, 73 (97%) were C. diastatops, while
the other two were C. maculata. The smaller individuals could not be identified
to species with a high degree of certainty with the keys, but most appeared to be
C. diastatops. Given that the keys are based on characteristics found in the last
nymphal stadium, and that the relative appearance of these characteristics can
change with nymphal age, it is possible that there could be more C. maculata at
the sites.
Spiketail density varied among the sites, but the overall average density of
spiketails among sites did not differ (F1,7 = 0.324, P > 0.75). Studies on C. dorsalis
showed that females may preferentially choose inlet areas for oviposition
and as larval habitat (Marczak et al. 2006), so we examined nymph distribution
by zone. More individuals occurred in the inlets (Fig. 3); spiketails in the inlets
had smaller average head widths (hw) than those in other zones (inlet hw in mm:
Figure 2. Frequency distribution of body-length size classes for Cordulegaster nymphs
recovered from the eight sites within Nelson Swamp, NY. All sites with more than 6
spiketails had a mixture of size classes within them.
2012 B.J. Hager, N.J. Kalantari, and V.A. Scholten 71
4.33 ± 1.42 vs. 4.88 ± 1.33 in middle and 4.45 ± 1.65 in outlets; ANOVA F 2,294 =
3.86, P < 0.02). Because substrates and zones are confounded and sample sizes
Table 1. Physical descriptions and select biological findings for the 8 springs and seeps along Chittenango
Creek, Nelson Swamp, NY, sampled for the presence of Cordulegaster (spiketail) nymphs.
The perimeter is the sum of the length of the shoreline of the spring or seep. For benthic substrate,
m = muck, s = sand, lc = large cobble (>10 cm), sc = small cobble (<10 cm), cl = clay and v =
vegetation. The density of Cordulegaster is the number of individuals per meter of shoreline perimeter.
NA= Not applicable
Water Cordulegaster
Benthic substrate depths (cm) nymphs Presence of
Perimeter Inlet/ Larval Fish or
Site (m) Inlet Middle Outlet outlet Middle # Density salamanders frogs
Springs
1 21.6 s/m m/sc m/v 50 56 30 1.38 No Yes (frogs)
3 15.5 m/v m/lc m/sc 10 90 124 8.00 Yes Yes (both)
4 7.7 m/cl cl/lc cl/sc 13 85 1 0.13 No Yes (fish)
5 8.1 cl/sc cl/lc cl/lc 11 88 0 0.00 No Yes (both)
6 11.1 m/sc m m/sc 7 7 66 5.94 Yes Yes (frogs)
Seeps
2 15.2 NA m sc 10 43 5 0.33 Yes Yes (frogs)
7 7.5 sc m m 5 44 61 8.13 Yes Yes (frogs)
8 12.1 NA m NA 2.5 5 9 0.74 No No
Figure 3. Abundances of Cordulegaster nymphs in the three habitat zones combined for
all sites.
72 Northeastern Naturalist Vol. 19, Special Issue 6
are small, we could not determine the relative importance of each in influencing
nymph distribution. However, we performed 2 preliminary analyses of substrate
and nymphal age (as indicated by head width) with the data at hand.
When we combined all sites and zones, more nymphs occurred in the muck
substrates than in the cobble and mixed muck/cobble substrates (χ2 = 125.4, 2 df,
P < 0.0001; Fig 4). Smaller individuals were found more often in the cobble substrate,
while larger individuals occurred in muck, whether we combined all of the
zones (inlet, outlet, and middle) (ANOVA: F 2,293 = 9.66, P < 0.009) or conducted
the analysis for inlets only (ANOVA: F 2,174 = 8.48, P < 0.0003) (Fig. 5).
Extrapolating from work done with other spiketail species (Glotzhober 2006),
it appears that the Nelson Swamp sites supported at least 2 cohorts of nymphs
(Fig. 6). Forty-eight individuals most likely belonged to the F-1 or F-0 stadium
(F-1 is the penultimate and F-0 is the ultimate nymphal stadium before molting
to adulthood) and were separated clearly from the younger nymphs by having a
body length > 26 mm and a head width > 6 mm.
Discussion
Given the small number of sites (Table 1), we cannot draw conclusions about
seep or spring preference by the spiketail species found in Nelson Swamp. However,
we can make some initial assessments about habitat associations. Consistent
Figure 4. Frequency of individuals found in the three major types of substrate. Expected
numbers calculated by assuming equal abundances across the three substrates. More individuals
were found in muck than expected (χ2 = 125.4, 2 df, P < 0.0001).
2012 B.J. Hager, N.J. Kalantari, and V.A. Scholten 73
Figure 5. Average head widths of Cordulegaster nymphs in relation to the substrate type
of the inlets (ANOVA: F2,174 = 8.48, P < 0.0003).
Figure 6. Correlation of body lengths and head widths for Cordulegaster nymphs in seeps
and springs of Nelson Swamp, NY, all sites combined. The circle indicates nymphs of a
separate age cohort.
74 Northeastern Naturalist Vol. 19, Special Issue 6
with other studies (Burcher and Smock 2002, Corbet 1999, Glotzhober 2006, Stevenson
et al. 2009), spiketails in our sites use habitats with slow water currents
and muck or sand substrates, both of which permit burrowing by the nymphs.
One of us (B.J. Hager) has observed C. diastatops patrolling and ovipositing at 5
of the 7 sites over a 6-year period; in contrast, C. maculata was not seen foraging
in the fields or at sites occupied by nymphs until the summers of 2010 and 2011
(B.J. Hager, unpubl. data). Whether or not these 2 species occupy different microhabitats
within Nelson Swamp remains to be determined. Burcher and Smock
(2002) and Corbet (1999) noted that C. maculata is more of a habitat generalist
compared to many species within the genus Cordulegaster because, while
also recorded as nymphs or ovipositing adults in small streams (Donnelly 1992,
Dunkle, 2000, Glotzhober and McShaffery 2002), they are commonly found in
the slower currents and backwaters of larger streams and along lake shorelines
near inlets. Likewise, Santos and Stevenson (2011) reported in their study of
perennial and non-perennial stream fauna in Massachusetts that C. maculata occurred
only in perennial streams. We found both of the definitive specimens of
C. maculata in site 6, a 1.5-m-wide spring that parallels Chittenango Creek and is
separated from the creek at its nearest point by roughly 40 cm of land. In contrast,
nymphs of C. diastatops are generally limited to smaller streams and are common
in “puny streams” in “upland spring bogs” (Needham 1901) and small streams of
forests and marshes (Donnelly 1992, Dunkle 2000).
While many nymphs occupied the inlets (Fig. 3), we could not determine
what influenced this distribution pattern. An initial analysis of the influence of
substrate, using inlet data only, shows that nymph abundances and average sizes
were greatest in muck (Figs. 4, 5). In contrast, the smallest individuals occurred
in cobble (Fig. 5), which could be a consequence of one or more factors. First,
female C. maculata and C. diastatops tend to oviposit in muck and vegetation
in the shallowest portions of the water (Corbet 1999, Dunkle 2000, Nikula et al.
2007), where small and large cobble were common, and the nymphs might not
have dispersed after egg hatch. Second, the narrower spaces among cobbles as
well as the thinner sediment layer may prohibit larger individuals from burrowing
here; Marczak et al. (2006) found that habitat particle size and body size were
positively correlated for nymphs of C. dorsalis. Third, smaller nymphs may seek
refuge from predators, including conspecifics, in the spaces among the cobble.
Subsequent studies at the Nelson Swamp sites need to more carefully examine
the distribution and density of nymphs in relation to muck depth, size of cobble,
and size of sediments for each of the habitat zones (inlet, outlet, middle).
Three of the 4 springs and seeps with the highest spiketail nymph densities
also contained larval salamanders (Desmognathus spp.; Table 1). Cordulegaster
erronea co-occurred with larval Pseudotriton ruber Latreille (Red Salamanders)
in spring-fed stream habitats in Ohio (Glotzhober 2006), while larval
Eurycea cirrigerra Green (Southern Two-lined Salamanders), Pseudotriton
montanus Baird (Mud Salamanders), Red Salamanders, and larval and adult
Desmognathus fuscus conanti Rossman (Spotted Dusky Salamander) inhabited
86% of C. sayi sites in Georgia (Stevenson et al. 2009). These investigators
2012 B.J. Hager, N.J. Kalantari, and V.A. Scholten 75
both suggest that salamanders might be important predators, prey, and/or competitors
with the spiketail nymphs. We observed a spiketail nymph feeding on
one of the larval salamanders.
Most spiketail species are semivoltine and take from 2–5 years to reach
maturity (Burcher and Smock 2002, Corbet et al. 1999, Ferreras-Romero and
Corbet 1999, Glotzhober 2006). Assuming that growth and molting follow similar
patterns among Cordulegaster species, we conclude that there are at least two
age cohorts at our sites. Life-history studies of C. erronea (Glotzhober 2006),
Cordulegaster boltonii Donovan (Golden-ringed Spiketail; Ferreras-Romero and
Corbet 1999), and C. maculata (Burcher and Smock 2002) demonstrated clear,
correlated changes in head width, body length, and wing pad length associated
with the F-1 and F-0 stadia. These authors show that, when plotting the measurements
in size classes, a clear separation from the younger individuals generally
occurs for those near or entering the F-1 and F-0 stadia. Based on their findings,
roughly 16% of our nymphs were probably at the F-1 or F-0 stadium (Fig. 6).
Conversely, we had 14% with head widths of less than 3 mm. These latter individuals
are likely to be from eggs laid and hatched in 2010, if we assume that
all Cordulegaster eggs develop directly without overwintering, as is the case for
the Golden-ringed Spiketail (Ferreras-Romero and Corbet 1999, Schutte 1997).
Thus, with the information at hand, we cannot determine if there are more than 2
age cohorts in our sites. Repeated sampling and replacement of nymphs at these
habitats over a more extended period of time, as done by Glotzhober (2006), is
necessary to determine the specifics of the age cohorts in Nelson Swamp. Additionally,
rearing of specimens should be conducted to determine life-history
parameters and to obtain more precise information about the relative numbers
and distributions of the two spiketail species using Nelson Swamp.
Acknowledgments
Partial funding for this project came from Cazenovia College’s Faculty Development
Fund. The authors thank Erin White (NY Natural Heritage Program, NYS DEC)
and Matthew Schlessinger (NY Natural Heritage Program, Chief Zoologist, DEC) for
placement of B. Hager on the collection permit (for odonates) during 2010–2011. We
gratefully acknowledge the statistical advice and editorial expertise of Dr. William M.
Shields, manuscript editor Maria Aliberti Lubertazzi and two anonymous reviewers,
who provided outstanding insights to improve this manuscript. Special thanks goes to
the manuscript editor, Maria Aliberti Lubertazzi for her patience and excellent advice in
revising this manuscript. We thank T. Yorks for confirming the fish identification.
Literature Cited
Bright, E., and M.F. O’Brien. 1999. Key to mature larvae of Michigan Cordulegaster. In
Odonate Larvae of Michigan. Available online at: http://insects.ummz.lsa.umich.edu/
michodo/MOL/Home.htm. Accessed 10 September 2010.
Burcher, C.L., and L.A. Smock. 2002. Habitat distribution, dietary composition, and
life-history characteristics of Odonate nymphs in a blackwater coastal plain stream.
American Midland Naturalist 148:75–89.
76 Northeastern Naturalist Vol. 19, Special Issue 6
Corbet, P.S. 1999. Dragonflies: Behavior and Ecology of Odonata. Cornell University
Press, Ithaca, NY. 829 pp.
Dodson, S. 2005. Introduction to Limnology. McGraw-Hill Publishers, New York, NY.
416 pp.
Donnelly, T.L. 1992. The Odonata of New York. Bulletin of American Odonatology
1(1):1–27.
Dunkle, S.W. 2000. Dragonflies through Binoculars. Oxford University Press, New York,
NY. 266 pp.
Ferreras-Romero, M., and P.S. Corbet. 1999. The life cycle of Cordulegaster boltonii
(Donovan, 1807) (Odonata:Cordulegastridae) in the Sierra Morena Mountains (southern
Spain). Hydrobiologia 405:39–48.
Glotzhober, R.C. 2006. Life-history studies of Cordulegaster erronea Hagen (Odonata:
Cordulgastridae) in the laboratory and field. Bulletin of American Odonatology
10(1):1–18.
Glozthober, R.C., and D. McShaffrey. 2002. The Dragonflies and Damselflies of Ohio.
Ohio Biological Survey, Columbus, OH. 364 pp.
Marczak, L.B., J.S. Richardson, and M.C. Classen. 2006. Life-history phenology and
sediment size association of the dragonfly Cordulegaster dorsalis (Odonata: Cordulegastridae)
in an ephemeral habitat in southwestern British Columbia. The Canadian
Field Naturalist 120(3):347–350.
Needham, J.G. 1901. Aquatic insects of the Adirondacks. Bulletin of the NY State Museum
47:381–612.
Needham, J.G., M.J. Westfall, Jr., and M.L. May. 2000. Dragonflies of North America.
Scientific Publishers, Gainesville, FL. 939 pp.
Nikula, B., J.L. Loose, and M.R. Burne. 2007. A Field Guide to Dragonflies and Damselfl
ies of Massachusetts, 2nd Edition. Massachusetts Division of Fisheries and Wildlife,
Natural Heritage and Endangered Species Program, Westborough, MA. 196 pp.
NY State Department of Conservation (NYSDEC) Region 7 Management Team. 2000.
Nelson Swamp Unique Area stewardship management plan. Available online at:
www.dec.ny.gov/docs/lands_forests_pdf/nelswp.pdf. Accessed 5 January 2011.
Santos, A.N., and R.D. Stevenson. 2011. Comparison of macroinvertebrate diversity and
community structure among perennial and non-perennial headwater streams. Northeastern
Naturalist 18:7–26.
Schutte, C. 1997. Early development and early instars in Cordulegaster boltonii immaculifrons
Selys: A field study (Anisoptera: Cordulegastridae). Odonatologica
26:83–87.
Stevenson, D.J., G. Beaton, and M.J. Elliott. 2009. Distribution, status, and ecology of
Cordulegaster sayi Selys in Georgia, USA (Odonata: Cordulegastridae). Bulletin of
American Odonatology 11:20–25.