Conservation, Biology, and Natural History of Crayfishes from the Southern US
2010 Southeastern Naturalist 9(Special Issue 3):175–184
Status and Distribution of Orconectes williamsi (Williams’
Crayfish) in Arkansas, with New Records from the
Arkansas River Drainage
Brian K. Wagner1,*, Christopher A. Taylor2, and Mark D. Kottmyer1
Abstract - Orconectes williamsi (Williams’ Crayfish) is a rare stream-dwelling
crayfish that is endemic to the upper White River basin of Arkansas and Missouri.
It was described in the 1960s from a type locality in Madison County, AR, and the
three localities named in the description comprise the published range of the species
in Arkansas. In contrast, 27 locations have been reported in Missouri. This study
surveyed a semi-random selection of stream sites in the Arkansas portion of this
range in order to characterize the crayfish communities and evaluate the status of O.
williamsi in Arkansas. A total of 2372 crayfishes were collected at 68 sites, including
197 Williams’ Crayfish from 23 sites. Orconectes meeki meeki (Meek’s Crayfish)
was the crayfish species most commonly associated with O. williamsi, occurring
sympatrically at 87% of sites occupied by O. williamsi. Orconectes williamsi was
found in small streams, with coarse substrates and no aquatic vegetation. Orconectes
williamsi showed a strong preference for riffle habitats. Due to its limited range and
habitat requirements, O. williamsi is moderately imperiled in Arkansas and should be
considered rare and vulnerable range-wide.
Introduction
Orconectes williamsi Fitzpatrick (Williams’ Crayfish) is a small, streamdwelling
crayfish, typically growing to a total size of 30.5–50.8 mm (1.2–2
inches; Pflieger 1996). It is most similar to O. meeki meeki (Faxon) (Meek’s
Crayfish), with which it commonly occurs (Pflieger 1996). Orconectes williamsi
differs from O. m. meeki in having gonopods with nearly straight
processes (vs. strongly curved in O. m. meeki), lacking a dark saddle on the
carapace near the abdominal junction, and often having a pale, vase-shaped
zone near the midline of the carapace (Fig. 1; Pflieger 1996).
Orconectes williamsi has a limited distribution in headwater streams
of the upper White River basin in Arkansas and Missouri. In Arkansas, it
is known from three locations in the White River headwaters in Madison
County, including the type locality 2.8 miles east of Pettigrew (Fitzpatrick
1966). It also occurs in at least four counties in Missouri (Pflieger 1996),
where its status was recently reviewed, increasing its published distribution
in that state to 27 locations (Westhoff et al. 2005, 2006).
This crayfish is a tertiary burrower occupying cavities excavated under
rocks seated in gravel in upland streams (Pflieger 1996). Fitzpatrick (1966)
1Arkansas Game and Fish Commission, 915 East Sevier Street, Benton, AR 7201. 2Illinois
Natural History Survey, Center for Biodiversity and Ecological Entomology, 1816
South Oak, Champaign, IL 61820. *Corresponding author - bkwagner@agfc.state.ar.us.
176 Southeastern Naturalist Vol. 9, Special Issue 3
found it occurred in pools, but was replaced by O. m. meeki in riffles.
Contrastingly, Westhoff et al. (2006) found it associated with high current
velocities, shallow depths, and cobble and small boulder substrates. Pflieger
(1996) found reproductive form males in fall, winter, and spring collections,
and recently molted individuals in April. Details about the reproduction and
life history of this species are unknown.
Figure 1. Orconectes williamsi with the characteristic vase-shaped marking on carapace.
This coloration was not always evident in specimens collected during the study.
2010 B.K. Wagner, C.A. Taylor, and M.D. Kottmyer 177
Westhoff et al. (2006) found O. williamsi at 27 sites, but failed to collect
it at 2 historic sites. They noted that many stream segments harboring O.
williamsi were isolated from one another by reservoirs. This finding raises
some concern as to the genetic isolation of populations, and the range-wide
significance of this concern cannot be assessed due to the lack of Arkansas
distributional data beyond the locations reported in the species description
(Fitzpatrick 1966).
Taylor et al. (1996) considered O. williamsi to be of “special concern,”
while Taylor et al. (2007) updated the status to “currently stable” based
on the results of this study and Westhoff et al. (2005, 2006). The absence
of good distributional and ecological information for Arkansas populations
of O. williamsi is an impediment to accurate status assessment and
effective conservation planning for this species. Thus, a survey is needed
in close temporal proximity to the Missouri study in order to provide a
complete picture of range-wide conservation status.
Methods
Study site selection
This study focused on portions of the Upper White River and Bull Shoals
Reservoir basins in northern Arkansas, encompassing the extent of the presumed
range of the species in Arkansas, predicted from the three historic
records from the state and the species’ adjoining range in Missouri. Based
on the National Hydrography Dataset (NHD), these two hydrologic units
comprise 7291 identified stream segments totaling 15,170 km. The Arkansas
portion of these units includes parts of Benton, Boone, Carroll, Franklin,
Madison, Marion, Newton, Searcy, and Washington counties. Since these
areas are largely in private ownership, road access to sampling sites was
particularly important. United States Census Bureau data on roads in these
counties were used to identify stream segments that intersect roads using
ArcMap™ Geographic Information System (GIS) software, resulting in the
identification of 1924 accessible segments. A semi-random subset of these
segments was selected for sampling by generating a random number between
1 and 20 as a start point, and then every 20th segment sequentially after this
start point in the pooled list of accessible stream segments was chosen. Since
the NHD segments were generally adjacent to one another in order, this approach
reduced the selection of clustered sampling sites and provided a fairly
uniform distribution of potential sampling areas. This procedure resulted in
the selection of 96 stream segments as potential sample sites. The selected
stream segments were then mapped on USGS 7.5-minute topographic maps
using ArcMap™ to allow for field navigation to appropriate road accesses
to those stream segments. Because headwater streams are more numerous
and more easily bridged than larger streams, it was acknowledged that site
selection was biased toward headwater streams, the habitat for the target
species (Pflieger 1996). Some of these headwater streams were intermittent
and did not hold water or crayfish when visited for sampling. When this was
178 Southeastern Naturalist Vol. 9, Special Issue 3
the case, the site was replaced with a nearby site on a larger stream that was
not initially selected for sampling. Distribution data were supplemented with
collections including O. williamsi from other areas encountered during field
collections for unrelated studies, in order to provide as complete a distributional
picture as possible.
Sampling methods
All available habitats at sites were sampled using minnow seines (1.2-m
x 1.2-m, or 1.8-m x 6.2-m, both with 6-mm mesh) or dip nets (0.3 m with
3-mm mesh) appropriately sized to the area being sampled. These net-based
sampling efforts were supplemented at sites with large slab rocks (those
bigger than 0.3 m in any dimension) by spending approximately 30 personminutes
(number of searchers x number of minutes searched) turning over
these rocks, visually searching, and capturing by hand crayfish found under
the overturned rocks. This method was implemented at approximately half
of the sites visited.
At each sample site, geographic coordinates (North American Datum
1927), water temperature, predominant substrate sizes, and qualitative
notes regarding depth and width of pool and riffle habitats, aquatic vegetation,
riparian vegetation, turbidity, and flow were recorded. For this
study, macrohabitats were classified as pools if they had slower flow,
undisturbed surface, and were the deeper habitats in the sample area;
conversely, riffles were habitats with rapid flow, surface disturbance, and
relatively shallow water.
Crayfish data were recorded separately for distinct macrohabitats (i.e.,
pools vs. riffles) at each site. Crayfish were sorted by perceived species,
their gender determined, and their carapace length measured to the nearest
1 mm. A series of voucher specimens including males and females of each
species were preserved in 70% ethanol and deposited in the collection of the
Illinois Natural History Survey or the Arkansas Game and Fish Commission
Nongame Aquatics Program reference collection after their identification to
species was verified by the second author.
Species occurrence was reported as the percentage of sample sites where
the species under consideration was encountered. Species associations with
O. williamsi were also examined using the metrics of dominance, constancy,
and fidelity, as described by Pflieger (1978). Dominance of species X was
defined as the proportion of all crayfish collected at sites occupied by O.
williamsi that are the species X. Constancy of species X was defined as the
proportion of O. williamsi sites also having the species X. Fidelity of species
X was defined as the proportion of all sites having species X that also have
O. williamsi.
Results
Of the 96 sites targeted for sampling, several were not accessible to sample,
resulting in samples being collected at 68 sites (Fig. 2). Crayfish species
2010 B.K. Wagner, C.A. Taylor, and M.D. Kottmyer 179
collected and location information for sites occupied by O. williamsi are listed
in Table 1. Collections from 23 sites yielded 197 specimens of O. williamsi,
averaging 16.5 mm carapace length (SD = 3.1 mm). A length-frequency graph
is provided in Figure 3. Males comprised 57% of the specimens collected. The
presence of O. williamsi was also documented by 12 additional specimens
from two sites in the Elk River basin and one site in the Mulberry River basin.
These specimens were not included in this study’s analyses, but these important
range extensions are included in Figure 2 and Table 1.
Nine crayfish species represented by a total of 2372 specimens were
collected within the study area. The most commonly encountered species
was O. neglectus neglectus (Faxon) (Ringed Crayfish) occurring at 63%
of sites sampled, followed by O. m. meeki, (53%), O. williamsi (34%), and
O. ozarkae Williams (Ozark Crayfish; 32%) (Table 2). Cambarus hubbsi
Creaser (Hubbs’ Crayfish), Orconectes longidigitis (Faxon) (Longpincered
Crayfish), Orconectes nana Williams (Midget Crayfish), Orconectes palmeri
longimanus (Faxon) (Western Painted Crayfish), Orconectes virilis (Hagen)
(Virile Crayfish), Procambarus acutus (Girard) (White River Crawfish), and
an unidentified Procambarus sp. were found at 4 or fewer sites and did not
co-occur with O. williamsi.
Figure 2. Map of northwest Arkansas depicting sampling locations included in this
report. Open circles indicate sites sampled as part of this study. Circled solid triangles
indicate study sites where O. williamsi was encountered. Solid triangles alone indicate
sites where O. williamsi was encountered outside study area through sampling
efforts by the principal author during other research projects.
180 Southeastern Naturalist Vol. 9, Special Issue 3
Crayfish species associations
Three species were found to occur at sites with O. williamsi, namely
O. n. neglectus, O. m. meeki, and O. ozarkae. The most common associate
of O. williamsi was O. m. meeki, occurring at 36 total sites during the
study and 87% of sites occupied by O. williamsi. Orconectes n. neglectus,
found at 43 total sites, was present at 26% of sites where O. williamsi was
found. Orconectes ozarkae, found at 22 sites, was only present at 4% of
sites occupied by O. williamsi. Species associations with O. williamsi were
also examined using the metrics of dominance, constancy, and fidelity
(Table 2). Among O. williamsi sites, O. williamsi was of equal dominance
to O. m. meeki, each comprising 39% of the community. The least dominant
species was O. ozarkae (9%), being collected with O. williamsi at only one
site. Constancy results indicated that O. m. meeki (87%) and O. ozarkae
(4%) were found most and least often at sites having O. williamsi, respectively.
Fidelity estimates, incorporating all sites sampled during this study
regardless of O. williamsi presence, were greatest for O. m. meeki (56%)
and lowest for O. ozarkae (5%).
Table 1. Crayfish species and numbers collected by site. The 2005 records of range extensions
are from outside the study area. Species are (1) Orconectes m. meeki (2) O. n. neglectus, (3) O.
ozarkae, and (4) O. williamsi.
Species
Collection Stream, County 1 2 3 4
BKW2004-028 Tributary of White River, Madison 12 3
BKW2004-029 Kilgore Branch, Madison 5 17
BKW2004-034 Tributary of Mill Creek, Franklin 30 6
BKW2004-035 Kings River, Madison 44 3
BKW2004-064 Tributary of W Fork White River, Washington 7 3 9
BKW2004-074 Barren Fork, Boone 8 43 1
BKW2004-075 Mill Hollow, Boone 1 20
BKW2004-078 Hutchins Creek, Washington 16 1
BKW2004-079 Riley Creek, Washington 5 11
BKW2004-080 Tributary of W Fork White River, Washington 6 5
BKW2004-081 Peach Brook, Washington 1 30
BKW2004-082 Tributary of Fritts Creek, Madison 6 12
BKW2004-083 Tributary of War Eagle Creek, Madison 6
BKW2004-084 Jackson Creek, Madison 15 14
BKW2004-085 Tributary of Drakes Creek, Madison 7 15
BKW2004-086 Middle Fork White River, Washington 14 6
BKW2004-087 Leatherwood Creek, Carroll 4 7
BKW2004-088 Keels Creek, Carroll 3 12
BKW2004-089 Pine Creek, Carroll 5 4
BKW2004-090 Warden Creek, Carroll 1 31 4
BKW2004-093 Cedar Creek, Carroll 3 9 6
BKW2004-095 Dog Branch, Carroll 6 4
BKW2004-097 Little Indian Creek, Carroll 4 15 1
BKW2005-027 Washita Creek, Johnson 13 1
BKW2005-069 Tributary of Spanker Brook, Benton 28 10
BKW2005-084 Tributary of Big Sugar Creek, Benton 17 1
2010 B.K. Wagner, C.A. Taylor, and M.D. Kottmyer 181
Table 2. Numbers of crayfish individuals (ind.) collected by species, % of total ind., number of
sites occupied, % of sites occupied, and Orconectes williamsi species association information,
including dominance, constancy, and fidelity, after Pflieger (1978). An asterisk (*) indicates
species not found with O. williamsi.
O. williamsi species associations
Ind. Ind. Sites Sites Dominance Constancy Fidelity
Species (#) (%) (#) (%) (%) (%) (%)
O. hubbsi 7 less than 1 1 2 * * *
O. longidigitus 15 less than 1 4 6 * * *
O. nana 19 less than 1 1 2 * * *
O. m. meeki 523 22 36 53 39 87 56
O. n. neglectus 1004 42 43 63 14 26 14
O. ozarkae 556 23 22 32 9 4 5
O. p. longimanus 1 less than 1 1 2 * * *
O. virilis 1 less than 1 1 2 * * *
O. williamsi 197 8 23 34 39 n/a n/a
P. acutus 2 less than 1 1 2 * * *
Procambarus sp. 47 2 1 2 * * *
Figure 3. Orconectes williamsi length-frequency histogram based on all specimens
collected during this study (n = 197).
Habitat characteristics
Qualitative habitat observations allow only for a general comparison of
sites with O. williamsi against all sites sampled. Orconectes williamsi occupies
smaller streams (87% less than 10 m wide versus 68% over all sites) with coarser
substrates (39% dominated by cobble and boulder versus 15% over all sites)
and no aquatic vegetation (91% lacking vegetation versus 69% over all sites).
Half of the crayfish species collected were found exclusively in pools and O. m.
meeki was found in pools 76% of the time. In contrast, 79% of O. williamsi and
182 Southeastern Naturalist Vol. 9, Special Issue 3
89% of O. nana were found in riffles. Orconectes n. neglectus (36% in pools)
and O. ozarkae (40% in pools) appeared to utilize both habitats.
Discussion
Distribution
Orconectes williamsi was described by Fitzpatrick (1966), who reported
it as restricted to 3 sites in Madison County, AR. Pflieger (1996) reported 9
localities from the upper White River basin in Missouri, expanding the known
range of O. williamsi much further north. While still only reported from 3 published
sites in Arkansas, O. williamsi was also known to occur at additional
locations in the vicinity (H.W. Robison, Southern Arkansas University, Magnolia,
AR, pers. comm.). Recent work by Westhoff et al. (2006) increased O.
williamsi’s documented range in Missouri to 27 sites, which, when combined
with the 23 Arkansas sites documented in this study, provides a much better
picture of the global distribution of this Ozark endemic. Within the White
River drainage, its known range now includes portions of Barry, Christian,
Stone, and Taney counties in Missouri and Benton, Boone, Carroll, Franklin,
Johnson, Madison, and Washington counties in Arkansas. The single specimen
collected from Washita Creek (Mulberry River drainage) in Johnson County,
AR, and the 11 specimens from 2 sites in the Elk River drainage in Benton
County, AR, represent the first known records for the species from outside
the White River basin, specifically in the Arkansas River basin. This overlap
of crayfish distributions into a few headwaters of adjacent basins is common
(C.A. Taylor, pers. observ.).
Species associations
Orconectes williamsi was found with O. m. meeki more commonly than
with any other crayfish, which is consistent with the findings of Fitzpatrick
(1966) and Westhoff et al. (2005, 2006). In Arkansas, O. williamsi was not
as dominant (39%), as Missouri populations (56%). We found O. m. meeki
(39%) to be equally dominant, while it was much less so in Missouri (5%). The
second in dominance in Missouri was O. n. neglectus (28%), which was less
common at Arkansas O. williamsi sites (14%). The dominance of O. ozarkae
was comparable in the two states (AR: 9%, MO: 9%), but the species was only
represented by a high abundance at a single O. williamsi site in our study. Orconectes
virilis was of low dominance in Missouri (2%) and not represented at
Arkansas O. williamsi sites. Constancy results indicated that O. m. meeki was
found most often at Arkansas O. williamsi sites (AR: 87%, MO: 18%), while O.
n. neglectus was most constant in Missouri (AR: 26%, MO: 94%). O. n. neglectus
was only present at one O. williamsi site in the Boston Mountain Ecoregion,
where it appears to be replaced by much higher populations of O. m. meeki
than in the Ozark Highlands Ecoregion, where O. n. neglectus is more abundant.
Fidelity estimates were greatest for O. m. meeki in both states (AR: 56%,
MO: 75%), indicating that it is the best species associate for indicating suitable
habitat within the range of O. williamsi. It should be noted, however, that O. m.
2010 B.K. Wagner, C.A. Taylor, and M.D. Kottmyer 183
meeki has a much broader range in Arkansas, so this fidelity estimate is only applicable
within the range of O. williamsi.
Habitat use
Orconectes williamsi is reported to inhabit “gravelly headwater creeks,
spring branches, and cave streams” (Pflieger 1996). In Fitzpatrick’s (1966)
description, he notes that O. williamsi was associated with “proportionately
very large stones” in pools, while riffles yielded O. m. meeki. This description
is not consistent with our findings, which show that O. williamsi was
strongly associated with riffles. While we encountered large rocks that we
turned in order to capture crayfish by hand at approximately half of the sites
visited, over 80% of sites with O. williamsi did not have such rocks. Our
results are consistent with the findings of Westhoff et al. (2006), who also
reported associations with shallow water and high current velocity. It is
possible that this discrepancy could be due to seasonal shifts in habitat use.
Fitzpatrick (1966) collected in January 1965, Westhoff et al. (2006) sampled
in April through August 2002 and 2003, and the current study included October
through December 2004. All three studies support an association with
cobble or larger substrates. Both the present study and Westhoff et al. (2006)
concur on a negative association with aquatic vegetation.
Overall, O. williamsi habitat can be characterized as small streams that
are well-incised, with coarse substrate, shallow water, fast current, and limited
aquatic vegetation growth, possibly due to riparian forest cover and/or
lack of siltation in the stream channels. However, an outlier to this description
was the site on Leatherwood Creek (collection BKW2004-087), which
was a concrete-lined roadside ditch in downtown Eureka Springs having
limited natural substrate and little riparian forest cover. Crayfish density was
low at this site, but O. williamsi was the dominant species.
Recommendations
One goal of this study was to reassess the conservation status of O. williamsi,
as reflected by the assigned heritage ranks for the species. Orconectes
williamsi is presently ranked as critically imperiled (S1) in Arkansas (NatureServe
2008). Westhoff et al. (2006) recommended a rank in Missouri of
imperiled (S2), after finding the species at 27 sites in Missouri. Since we
found the species at 23 sites in Arkansas, we propose that a rank of imperiled
(S2) would also be appropriate for Arkansas. Based on the total number
of known sites reported in these two studies, the current global rank of rare
(G3) seems appropriate.
While the range of the O. williamsi is restricted, most locations where the
species occurs have healthy populations. The collections outside the White
River basin are intriguing, and further collections in these areas should be
made to examine the extent of the species’ overlap into adjacent basins.
The biology of O. williamsi is mostly unknown. Future research should
address reproductive timing, fecundity, and growth. Initial efforts to collect
at the type locality found it to be dry, but later visits after fall rains revealed
184 Southeastern Naturalist Vol. 9, Special Issue 3
a refilled channel and the presence of O. williamsi. This discovery is consistent
with the finding of several headwater crayfish species in reaches that dry
seasonally in Missouri (R.J. Distefano, Missouri Department of Conservation,
Columbia, MO, pers. comm.). Life-history advantages and challenges
in such drying habitats could be important in the persistence of this endemic
crayfish and should be examined in future research.
Our qualitative habitat data supports an association of O. williamsi with
small streams with coarse substrates and little silt (as reflected by the absence
of vegetation). Siltation is one of the largest threats to habitat quality
in upland streams and could affect O. williamsi, so efforts should be made to
maintain riparian vegetation in basins occupied by this crayfish.
Acknowledgments
Field assistance was provided by T. Anderson, S. Brown, K. Coffey, F. Leone,
R. Limbird, M. Oliver, K. Shirley, and J. Stinnett of the Arkansas Game and Fish
Commission; E. Inlander and D. Crosswhite of The Nature Conservancy; and S.
Coughlan of Arkansas State University. C. Dillman and J. Ray of Saint Louis University
assisted with one of the reported supplemental collections. Stuart Welsh provided
publication support. The publication of this manuscript was supported, in part, by
the US Geological Survey Cooperative Research Unit Program, including the West
Virginia Cooperative Fish and Wildlife Research Unit.
Literature Cited
Fitzpatrick, J.F., Jr. 1966. A new crawfish of the genus Orconectes from the headwaters
of the White River in Arkansas (Decapoda, Astacidae). Proceedings of the
Biological Society of Washington 79:145–150.
NatureServe. 2008. NatureServe Explorer: An online encyclopedia of life [web application].
Version 7.0. NatureServe, Arlington, VA. Available online at http://
www.natureserve.org/explorer. Accessed 1 April 2008.
Pflieger, W.L. 1978. Distribution, status, and life history of the Niangua Darter,
Etheostoma nianguae. Missouri Department of Conservation Aquatic Series
Report No. 16. Jefferson City, MO. 25 pp.
Pflieger, W.L. 1996. The Crayfishes of Missouri. Missouri Department of Conservation,
Jefferson City, MO. 152 pp.
Taylor, C.A., M.L. Warren, Jr., J. F. Fitzpatrick, Jr., H.H. Hobbs III, R.F. Jezerinac,
W.L. Pflieger, and H.W. Robison. 1996. Conservation and status of crayfishes of
the United States and Canada. Fisheries 21(4):25–38.
Taylor, C.A., G.A. Schuster, J.E. Cooper, R.J. DiStefano, A.G. Eversole, P. Hamr,
H.H. Hobbs III, H.W. Robison, C.E. Skelton, and R.F. Thoma. 2007. Reassessment
of the conservation status of crayfishes of the United States and Canada
after 10+ years of increased awareness. Fisheries 32(8):372–389.
Westhoff, J.T., J.A. Guyot, and R.J. Distefano. 2005. A survey of the distribution of
the imperiled Williams’ Crayfish (Orconectes williamsi) in the upper White River
drainage of Missouri: Associations with multi-scale environmental variables.
Missouri Department of Conservation, Final Report, Columbia, MO. 90 pp.
Westhoff, J.T., J.A. Guyot, and R.J. Distefano. 2006. Distribution of the imperiled
Williams’ Crayfish (Orconectes williamsi) in the upper White River drainage
of Missouri: Associations with multi-scale environmental variables. American
Midland Naturalist 156:273–288.