2010 SOUTHEASTERN NATURALIST 9(2):251–258
Emergence-site Selection by the Dragonfly
Epitheca spinosa (Hagen)
Wade B. Worthen*
Abstract - Odonates are vulnerable during emergence, when they shed their larval
skin (exuvia) to take flight as adults. Emergence-site selection should adapt
to the local mortality risks. Here, I characterized emergence-site selection of Epitheca
spinosa (Robust Baskettails) by noting the substrate, height, and distance
from water of exuviae in a 300 m x 5 m plot at Weston Lake, Congaree National
Park, Hopkins, SC. Of the 82 Robust Baskettail exuviae sampled, 52 (63.4%)
were found on trees with corky bark (Nyssa aquatica [Water Tupelo], Nyssa biflora
[Swamp Tupelo], Fraxinus pennsylvanica [Green Ash]), while no exuviae
were found on the peeling, flaky trunks of Taxodium distichum (Bald Cypress)
or the smooth, platy trunks of Acer rubrum (Red Maple). However, 26 (31.7%)
exuviae were on T. distichum pneumatophores. This pattern was significantly
different from the relative abundances of these substrate types (χ2 = 19.8, df = 3,
P < 0.001). Most exuviae (93.9%) were on substrates touching the water, suggesting
that larvae climb directly from the water to their emergence site. The mean
height of exuviae on trees was 3.3 ± 1.37 m, with a range from 1.8–7.7 m. Highclimbing
by Robust Baskettail larvae may be an adaptation to flooding at Weston
Lake; major flood events (>3 m) are common (5 of the last 10 years) during their
March–April emergence period.
The corduliid dragonfly Epitheca (Tetragoneuria) spinosa (Hagen)
(Robust Baskettail) ranges from New Jersey to Georgia and the Florida
panhandle (Needham et al. 2000). The drab, hoary adults hover along the
margin of ponds and lakes. Larvae live in swamps where there is slowmoving
water (Dunkle and Westfall 1982). Although uncommon throughout
the majority of its range (Tennessen 1994), it is locally abundant at Congaree
National Park, 30 km southeast of Columbia, SC. This is not surprising, as
the 10,700+ ha of Congaree National Park encompass the largest old-growth
floodplain forest in the United States, including several ancient oxbow
lakes that are ideal habitat for Robust Baskettails and congeners Epitheca
cynosura (Say) (Common Baskettail) and Epitheca semiaquea (Burmeister)
(Mantled Baskettail). Because Robust Baskettails are uncommon, little is
known of their larval ecology.
In odonate development, emergence is defined as the period when an
individual clambers from the water, attaches to a substrate, climbs from
its larval exoskeleton (exuvia), expands its abdomen and wings, and takes
*Biology Department, Furman University, Greenville, SC 29613; wade.worthen@
252 Southeastern Naturalist Vol. 9, No. 2
flight as an adult (Fig. 1a–c; Corbet 1999:237). Once they are attached to
a substrate, they are immobile, soft-bodied, and vulnerable to damage and
predation. This vulnerability is probably why emergence usually occurs
quickly—in less than an hour under ideal conditions—and at night. Selection
should also favor organisms that select emergence sites that reduce
these risks (Cordero 1995). The presence of new, intact exuviae attached to
a substrate allows easy characterization of odonate emergence sites. Here, I
describe emergence-site selection by Robust Baskettail larvae by describing
the distribution of exuviae at Weston Lake, Congaree National Park.
Weston Lake (33°49'17.33"N, 80°49'05.31"W; 29 m elevation) is approximately
400 x 50 m, and is skirted by a bottomland forest dominated by
Taxodium distichum (L.) L. C. Rich. (Bald Cypress), Nyssa aquatica L. (Water
Tupelo), Fraxinus pennsylvanicus Marsh (Green Ash) and Acer rubrum
L. (Red Maple). There is also an occasional Nyssa biflora Walt. (Swamp
Tupelo), Alnus serrulata (Ait.) Willd. (Alder) Quercus pagoda Raf. (Cherrybark
Oak), and Liquidambar styraciflua L. (Sweetgum). The first six species
grow at the bank, and tree trunks and cypress “knees” (pneumatophores) are
the only vertical surfaces available to climbing larvae in early spring.
I visited Weston Lake on March 20–22, 2009, and recorded the location
of Robust Baskettail exuviae in a 300- x 5-m plot along the northwestern
shore. Species emerging under cold conditions often favor this aspect,
perhaps to be warmed by direct rays of early morning sunshine (Beynon
1995). The height, distance from the bank, and substrate type were noted.
Substrates were characterized as trees with: 1) corky bark (Water Tupelo,
Swamp Tupelo, or Green Ash), 2) peeling bark (Bald Cypress), 3) smooth,
platy bark (Red Maple), and 4) pneumatophores (Bald Cypress).
Results and Discussion
I observed 82 Robust Baskettail exuviae in the plot, including 4 with
individuals in the process of emergence (as in Fig. 1a–c). Robust Baskettail
larvae and exuviae are easily distinguished from co-occurring congeners because
the lateral spines on abdominal segment 9 do not extend beyond the tip
of the abdomen, as they do in the other species (Fig. 2; Needham et al. 2000).
I found only two exuviae of other species, both belonging to Basiaeschna
janata Say (Springtime Darner).
Dragonfly larvae can walk remarkable distances overland to find a
suitable emergence site. Several species walk more than 10 m from water
(Corbet 1999:630–631), and larvae of Plathemis lydia (Drury) (Common
2010 W.B. Worthen 253
Figure 1. (a–c) Emergence of Epitheca
spinosa (stages 2–4, respectively).
Whitetail) have walked more than
45 m from water (Jacobs 1955).
At Weston Lake, however, Robust
Baskettails did not travel far from
water to emerge; most (93.9%)
were found on supports in direct
contact with water, and the rest
were found on supports within 2 m
of water. I returned to the plot on
July 24–25, 2009, to determine
the relative abundances of the substrate
types within 2 m of water and
in contact with the water. No account
was taken of the bark texture
below the water surface or the size
of the trees.
Most Robust Baskettail exuviae
were on trees with corky bark
(63.4%, Table 1). No exuviae were
found on trunks of Bald Cypress or
Red Maple, but 26 (31.7%) were
found on Bald Cypress pneumatophores
and 4 (4.9%) were found in
one cluster of three Alder saplings
(Table 1). To determine whether this
distribution represented a selective
substrate preference, I compared
the observed distribution of exuviae
across substrate types with an
expected distribution based on the
Figure 2. Exuvia showing diagnostic
short lateral spines on segment 9.
254 Southeastern Naturalist Vol. 9, No. 2
relative abundances of these substrates. I limited this analysis to exuviae
and trees found at the waterline, because this was where the vast majority of
larvae (93.9%) selected a substrate to climb. Alder was both included and excluded
from analyses because, with exuviae in the single cluster of saplings,
the “replicated” selection of Alder by larvae is debatable.
The distribution of exuviae across substrate types was significantly
different from the relative abundances of substrates at the water line; regardless
of whether pneumatophores are excluded from the comparison
(χ2 = 20.4, df = 2, P < 0.001), are included with Bald Cypress (χ2 = 14.9,
df = 2, P < 0.001), or are included as their own category (χ2 = 19.8, df = 3,
P < 0.001). Including Alder as a separate category only made each of the
above comparisons more significant. Trees with corky bark would seem
to provide an ideal surface for grasping tarsi. The gnarled, flaky pneumatophores
of Bald Cypress might also provide an adequate surface for
climbing and attachment. In contrast, Bald Cypress trunks have shaggy,
peeling, outer bark that could trap or impede climbing larvae. Where the
outer bark has shed, the inner bark may be too smooth for larvae to grip.
The smooth (sometimes plating) bark of Red Maple may account for the
absence of exuviae on this species, as well.
The exuviae found on mature trees (not Alder saplings or pneumatophores),
where climbing height was not limited by the height of the substrate,
were found at an average height of 3.3 ± 1.37 m (Fig. 3), with a range from
1.8–7.7 m. Exuviae could be high in trees if larvae float to the surface of
flood waters and then cling to trees they encounter, or mean climbing height
could be skewed if a flood removed low exuviae from supports. However,
1) the only flood before my sampling period (on March 4; 3.2 m) receded
well before any emergences occurred during this cold year, 2) there were
26 exuviae on low pneumatophores, and 3) I observed a Robust Baskettail
complete emergence on a tree trunk at a height of 5.0 m on March 22. So, I
believe the exuviae sampled on March 21–22 are representative indicators
of emergence-site selection by actively climbing Robust Baskettail larvae at
Table 1. The distribution of Robust Baskettail (Epitheca spinosa) exuviae on different substrates
at Weston Lake, Congaree National Park, Richland County, SC, including all exuviae within the
5- x 300-m plot, and the subset (93.9%) found on substrates in direct contact with water.
All exuviae Exuviae, waterline
Species (n) Substrate type (n) (n = 82) (n = 75)
Water Tupelo (108)
Green Ash (31) Corky bark (153) 52 (63.4%) 47 (62.7%)
Swamp Blackgum (14)
Red Maple (41) Platy bark (41) 0 (0%) 0 (0%)
Bald Cypress (25) Peeling bark (25) 0 (0%) 0 (0%)
Pneumatophore (103) 26 (31.7%) 24 (32.0%)
Alder (3) 4 (4.9%) 4 (5.3%)
2010 W.B. Worthen 255
Although it is remarkable to see
dragonfly exuviae more than 5 m up
tree trunks, it is not unprecedented.
Larvae of other Epitheca species
(either E. cynosura or E. costalis)
were found at a median height of
5.5 m at a lake in Alabama, and
an average distance of 4.73 m
from shore (Tennessen 1979). Several
other species have been found
above 5 m (Corbet 1999:630–631),
and Brachythemis contaminata
(Fabricius) and Pantala flavescens
(Fabricius) exuviae have been
found at heights of 12.5 m during
the monsoon season in India
(Mathavan and Pandian 1977).
Such an extreme behavior begs
an adaptive explanation, and several
have been proposed. High climbing
may be a response to high larval
densities—i.e., it spaces individuals
out and reduces the damage they
could cause one another during the
soft-bodied molting process (Bennett
and Mill 1993, Corbet 1957). It
may also reduce density-dependent
predation by birds or other dragonflies (Coppa 1991, Miller 1964),
and may reduce interspecific competition
for emergence sites (Cordero
1995). But climbing height also
correlates with physical factors like
temperature and humidity. Anax
junius (Drury) (Common Green
Darner), for example, climb higher
when the water and air are warm
and humidity is low; so climbing
may be a response to desiccation
stress (Trottier 1973).
At Weston Lake, high climbing
may be an adaptation to the
frequent and dramatic flooding
that occurs at Congaree National
Figure 3. Location of exuviae (circled)
at the top of a 3.4-m snag, with close-up
256 Southeastern Naturalist Vol. 9, No. 2
Park. The Congaree River inundates the Park an average of 10 times a
year (National Park Service 2006). One of the two peak flooding periods
is March–April, when Robust Baskettails are emerging. There is a USGS
gauging station 0.87 km from Weston Lake, on a tributary of the Congaree
River called Cedar Creek (USGS gauging station #02169672; 33°49'0.47"N,
80°49'39.06"W, 29 m elevation). In 5 of the last 10 years, there has been
at least one flood event over 3 m on Cedar Creek during the March–April
emergence period (Table 2; US Geological Survey 2009). The exact relationship
between gauge height on Cedar Creek and flood levels at Weston Lake
has not been described explicitly, but anecdotal evidence (photos of Weston
Lake at flood stage compared with contemporaneous gauge height readings
at Cedar Creek) suggests that a 3-m gauge height at the Cedar Creek station
corresponds to a 2-m flood at Weston Lake—roughly the minimum climbing
height of Robust Baskettails on trees. There are also direct observations
of 3-m floods at Weston Lake (F. Rametta, NPS Ranger, Congaree National
Park, SC, pers. comm.).
Floods at Congaree National Park are not only frequent, they can be severe
and prolonged. Flood waters crested at the Cedar Creek station at 4.6 m
and 4.1 m in 2003 and 2007, respectively, and gauge height exceeded 3 m for
26 of 62 days in March–April 2003 (Table 2). As such, it seems that flooding
could be a primary factor selecting for high climbing by Robust Baskettail
nymphs. This explanation may also apply to high climbing by tropical species
during monsoon season (Mathavan and Pandian 1977).
Climbing high may be doubly important for Robust Baskettails; low
temperatures in early spring can prolong emergence and further increase
the risk of damage from rising flood waters. For example, March 21, 2009,
was sunny but cool (16 °C). It took one individual three hours (22:30–1:30
local solar time) to complete emergence from “hanging” stage 2 to its virgin
flight (end of stage 4) in the “heat” of mid-day (Fig. 1a–c). This was twice
as long as it took two dragonflies to complete emergence from stage 2–4
on March 22, at 21 °C at local solar noon. Most Robust Baskettails emerge
Table 2. Flooding frequency and severity at Cedar Creek, Congaree National Park, during
the months of March and April, from 2000–2009. Data compiled from USGS gauging station
Year Days >3 m (events) Maximum (m)
2000 2 (1) 3.1
2001 0 2.8
2002 0 1.1
2003 26 (4) 4.6
2004 0 1.8
2005 5 (1) 3.6
2006 0 1.4
2007 6 (1) 4.1
2008 0 1.9
2009 4 (2) 3.2
2010 W.B. Worthen 257
before dawn when the temperatures are even cooler. When climbing time
(stage 1) is included, the entire emergence process probably takes many
hours to complete. High climbing would reduce the chance that the emerging
dragonfly would be caught by rising water during this protracted process.
It seems likely that emergence-site selection by these Robust Baskettails is
an adaptation to avoid the frequent, predictable, and occasionally extreme
spring flooding at Congaree National Park. However, comparisons of climbing
heights at habitats with different flooding patterns are necessary to test
I thank the staff at Congaree National Park, particularly Dr. Theresa Thom and
Ranger Fran Rametta, for their pictures and recollections of flooding at Weston Lake.
This research was conducted under NPS permit # CONG-2008-SCI-0008. I also
thank Dennis Paulson and an anonymous reviewer for their helpful comments.
Bennett, S., and P.J. Mill. 1993. Larval development and emergence in Pyrrhosoma
nymphula (Sulzer) (Zygoptera: Coenagrionidae). Odonatologica 22:133–145.
Beynon, T.G. 1995. Leucorrhinia dubia (Vander Linden) at Shooters Pool, Chartley
Moss, Staffordshire, in 1994. Journal of the British Dragonfly Society 11:1–9.
Corbet, P.S. 1957. The life history of the Emperor Dragonfly Anax imperator Leach
(Odonata: Aeshnidae). Journal of Animal Ecology 26:1–69.
Corbet, P.S. 1999. Dragonflies: Behavior and Ecology of Odonata. Comstock Publishing
Associates, Ithaca, NY. 829 pp.
Coppa, G. 1991. Notes sur l’émergence d’Epitheca bimaculata (Charpentier) (Odonata:
Corduliidae). Martinia 7:53–57.
Cordero, A. 1995. Vertical stratification during emergence in odonates. Notulae Odontologicae
Dunkle, S.W., and M.J. Westfall, Jr. 1982. Order Odonata. Dragonflies and damselflies. Pp. 32–45, In R. Franz (Ed.). Rare and Endangered Biota of Florida. Vol. 6:
Invertebrates. University of Florida Press, Gainesville, fl. 131 pp.
Jacobs. M.E. 1955. Studies in territorialism and sexual selection in dragonflies. Ecology
Mathavan, S., and T.J. Pandian. 1977. Patterns of emergence, import of egg energy,
and energy export via emerging dragonfly populations in a tropical pond. Hydrobiologia
Miller, P.L. 1964. Notes on Ictinogomphus ferox Rambur (Odonata: Gomphidae).
National Park Service. 2006. Congaree National Park: Boardwalk Loop. Available
online at http://www.nps.gov/cong/planyourvisit/boardwalk.htm. Accessed 14
Needham, J.G., M.J. Westfall, and M.L. May. 2000. Dragonflies of North America.
Scientific Publishers, Gainesville, fl. 940 pp.
Tennessen, K.J. 1979. Distance traveled by transforming nymphs of Tetragoneuria at
Marion County Lake, Alabama, United States (Anisoptera: Corduliidae). Notulae
258 Southeastern Naturalist Vol. 9, No. 2
Tennessen, K.J. 1994. Description of the nymph of Epitheca (Tetragoneuria) spinosa
(Hagen) (Odonata: Corduliidae). Bulletin of American Odonatology 2:15–19.
Trottier, R. 1973. Influence of temperature and humidity on the emergence behavior of
Anax junius (Odonata: Aeshnidae). The Canadian Entomologist 105:975–984.
United States Geological Survey. 2009. USGS real-time water data for the nation.
USGS 02169672 Cedar Creek at Congaree NP near Gadsden, SC. Available online
at http://waterdata.usgs.gov/nwis/uv?02169672. Accessed 15 April 2009.