2012. Proceedings of the Indiana Academy of Science 121(1):54–61

A SURVEY OF OF THE PATOKA RIVER NATIONAL WILDLIFE REFUGE AND MANAGEMENT AREA

Donald L. Batema* and Amanda Bellian: Department of Chemistry, Environmental Studies Program, University of Evansville, 1800 Lincoln Avenue, Evansville, IN 47722 USA Lindsey Landowski: Mingo National Wildlife Refuge, Puxico, MO. 63960 USA

ABSTRACT. The Patoka River National Wildlife Refuge and Management Area (hereafter Patoka River Refuge or the Refuge) represents one of the largest intact bottomland hardwood forests in southern Indiana, with meandering oxbows, marshes, ponds, managed moist-soil units, and constructed wetlands that provide diverse and suitable habitat for wildlife. Refuge personnel strive to protect, restore, and manage this bottomland hardwood ecosystem and associated habitats for a variety of wildlife. The Patoka River National Wildlife Refuge Comprehensive Conservation Plan (CCP) lists many of management priority (McCoy 2008), but Odonata are not included, even though they are known to occur on the Refuge. The absence of Odonata from the CCP is the result of lack of information about this ecologically important group of organisms. Therefore, we conducted a survey, from May to October 2009, to document their presence, with special attention being paid to rare, threatened, and endangered species. A total of 43 and species were collected and identified. No threatened or endangered species were found on the Refuge, but three species were found that are considered imperiled in Indiana based on Nature Serve Ranks (Stein 2002). Additionally, 19 new odonate records were documented for Pike County, Indiana. The results of this survey will be used by Refuge personnel to assist in management decisions and to help establish priorities for the Patoka River Refuge activities and land acquisition goals. Keywords: Odonata, , , management, wetlands

Odonata play vital roles as predators and terrestrial habitats (Bried 2005). Odonates are prey, contribute significantly to ecosystem also dominant or co-dominant organisms with biomass, are important bioindicators of eco- respect to biomass in permanently flooded system health, and are of conservation concern wetlands (Batzer & Wissinger 1996), beaver in wetland and aquatic habitats. Odonate impoundments (Benke et al. 1999), constructed nymphs and adults have been extensively or restored wetlands (Fairchild et al. 1999), and examined for their roles in predator-prey and many other wetland types. trophic relations (Kennedy 1950, Sih 1987, The presence of odonates in aquatic and Corbett 1999, Peyman 2000, Silsby 2001). Both terrestrial habitats has been linked to ecosystem damselfly and dragonfly nymphs feed on small health. Odonata are used as indicators of aquatic organisms, fish, and tadpoles (Corbet ecosystem health because they respond readily 1999, Colburn 2004), while adults feed on to changes in hydrology, are sensitive to mosquitoes, butterflies, and other winged degradation of water quality, and are directly associated with vegetation of aquatic linked to the composition and structure of habitats. Odonata nymphs and adults also fall vegetation (Burton et al. 1999, Chovanec & prey to other organisms, especially fish and Rabb 1997, Chovanec & Waringer 2001). Clark birds (Kennedy 1950, Orians & Horn 1969, & Samways (1996) noted that Odonata are King & Wrubleski 1998). highly rated (top 20%) for their suitability as Odonates associated with the vegetation of indicators of ecosystem health because they lakes, ponds, and streams contribute a signif- occupy a wide spectrum of aquatic habitats. icant amount of biomass to aquatic and Recently, odonates, in conjunction with plants, have been evaluated for use as possible ‘umbrella’ species for bioassessment purposes * Corresponding author: Donald L. Batema, 812-488- (Bried et al. 2007). The umbrella species 2795, e-mail: [email protected] concept focuses on a few core species that can

54 BATEMA ET AL.—ODONATA OF PATOKA RIVER REFUGE 55 be used as proxies for protecting and managing authorized acquisition of approximately a much larger community of species. 23,000 acres along a 20-mile stretch of the A number of Odonata are beginning to lower Patoka River in Pike and Gibson appear on lists of rare, threatened, or endan- Counties in southwest Indiana. The Patoka gered organisms as a result of declines in River Refuge is one of the last remaining intact odonate populations caused by environmental bottomland hardwood forests in southern stresses such as habitat loss, introduction of Indiana, providing a variety of habitats for a exotic invasive species, or grazing activities number of wildlife species and other organisms, (Foote & Hornung 2005, Stewart & Samways including odonates. 1998). Clausnitzer et al. (2009) examined the We sampled forested wetlands, scrub-shrub current status of a subset (1,500 species) of the wetlands, emergent marshes, moist-soil units, 5,680 known species of Odonata using the constructed wetlands, other aquatic habitats International Union for Conservation of Na- (e.g., ponds, streams, and river), and grass- ture (IUCN) Red List criteria. Their analysis lands. We also sampled similar habitats on the revealed that 10% of the world’s Odonata nearby White River Wildlife Management Area species are threatened. However, because 35% (WMA) and Cane Ridge WMA, both managed of the 1,500 species they examined are poorly by Refuge personnel. known biologically, the number of threatened Terrestrial (or aerial) sampling.—Adult drag- species may actually be higher. onflies and damselflies were captured at Odonata are important in ecosystem struc- various lentic and lotic habitats on the Refuge ture and function, as indicators of ecosystem with an net (diameter 15 inches). Odo- health, and increasing as organisms of conser- nates were captured from behind with a vation concern. Therefore, odonates warrant swinging motion and ending with a twist to study in an area, such as the Patoka River secure the individual in the net. Captured National Wildlife Refuge and Management individuals were identified in the field and Area (hereafter Patoka River Refuge or the released, or if new to our collection placed in a Refuge), where they are poorly understood. ZiplocH bag, put on ice to retard movement, The Patoka River Refuge Comprehensive and returned to the laboratory for identifica- Conservation Plan (CCP) (McCoy 2008) out- tion to species. We used the nomenclature of lines a vision and states goals that will direct Curry (2001) for adult dragonflies and either future activities at the Refuge. An overall goal DuBois (2005) or Westfall & May (2006) for of Refuge personnel is to restore, protect, and adult damselflies. Data recorded included date manage the bottomland hardwood ecosystem of capture, collector, location on the Refuge, and associated habitats for migratory birds, GPS location, and type of habitat, as well as threatened and endangered species, and indig- specific distinguishing features to aid in enous fish and wildlife, while striving to identification. After identification, adult odo- develop citizen understanding and support for nates new to our collection were scanned (both the protection of natural resources by provid- top and side view) at 300 dpi using a Hewlett- ing wildlife-related education and recreation Packard flatbed scanner (Fig. 1). Once opportunities (McCoy 2008). With this man- scanned, specimens were placed in acetone to agement goal in mind, our objectives were to: remove excess fats and air dried. Dried (1) Conduct a survey to document what species specimens were stored individually in 3 5/ of odonates are present on the refuge, (2) 16 inch by 5 1/8 inch clear, plastic envelopes examine species-habitat associations, with an (www.cistamps.com) along with unlined index emphasis on locating rare, threatened, and cards containing pertinent collection data. A endangered species, and, (3) provide a voucher voucher collection for adults, housed at the collection (and other materials) as tools to University of Evansville, was prepared and will educate the public about the habitats odonates be maintained for educational and research occupy and the role they play in the ecological purposes. processes on the Refuge. Aquatic sampling.—Nymphs of dragonflies and damselflies were sampled from lentic and METHODS lotic habitats within the Refuge with a standard Study area.—The Patoka River Refuge was D-frame dipnet (0.5 mm mesh). We sampled established in 1994. The federal government each habitat qualitatively by dragging the 56 PROCEEDINGS OF THE INDIANA ACADEMY OF SCIENCE

Figure 1.—Top and side view scans of a Spangled Skimmer ( cyanea). dipnet through vegetation and searched for Table 1.—Dragonfly species of the Patoka River nymphs. Nymphs were collected with forceps, National Wildlife Refuge and Management Area. placed in vials with 70% ethyl alcohol, and Families, species, and life history stage are listed. returned to the laboratory for sorting and identification. Nymphs were identified to spe- Family/Species Adult Nymph cies following the nomenclature of Needham and Westfall (1954), Merritt & Cummins Anax junius xx (1996), and Westfall & May (2006). Data Basiaeschna janata x recorded included date of capture, collector, heros xx location on the Refuge, GPS location, and type Nasciaeschna pentacantha x of habitat. Voucher specimens were stored in vials with 70% ethyl alcohol for educational and research purposes, and will be housed at villosipes x spinosus x the University of Evansville. Hagenius brevistylus x RESULTS plagiatus x Odonate species.—We identified 43 species of dragonflies and damselflies on Patoka River obliqua x Refuge. These 43 species are distributed among 19 genera in 9 families. cynosura x Dragonflies: Thirty species of dragonflies Epitheca princeps xx were collected and identified (Table 1). The linearis xx dominate dragonfly family was , LIBELLULIDAE which accounted for 18 of the 30 total species, elisa xx followed by Aeshnidae (4 spp.), Gomphidae (4 Celithemis eponina x spp.), Corduliidae (3 spp.), and Cordulegastri- Celithemis fasciata xx dae (1 sp.). Libellulidae accounted for 57% of simplicicollis xx total species richness for dragonflies. No Libellula cyanea x Macromiidae (cruisers) were collected in this Libellula incesta survey. Adults and nymphs were collected for Libellula luctuosa xx 13 species, adults only were collected for Libellula lydia xx Libellula pulchella xx another 13 species, and nymphs only were Libellula vibrans x collected for 4 species (Table 1). Pachydiplax longipennis xx Damselflies: A total of 13 species of damsel- Pantala flavescens x flies were collected during the survey (Table 2). Perithemis tenera xx These 13 species were distributed among 5 ambiguum x genera in 3 families. The dominate damselfly Sympetrum corruptum x family was (9 spp.), followed x by (3 spp.), and Calopterygidae (1 carolina xx sp.). Coenagrionidae accounted for 69% of Tramea lacerata x BATEMA ET AL.—ODONATA OF PATOKA RIVER REFUGE 57

Table 2.—Damselfly species of the Patoka River Table 3.—Species richness for different wetland National Wildlife Refuge and Management Area. and aquatic habitats on Patoka River National Families, species, and life history stage are listed. Wildlife Refuge and Management Area.

Family/Species Adult Nymph Habitat type Species richness CALOPTERYGIDAE NATURAL WETLANDS Calopteryx maculata x Marsh 16 LESTIDAE Oxbow 16 Bottomland 8 rectangularis xx Hardwoods Lestes unguiculatus x Lestes disjunctus xxMANAGED & CONSTRUCTED COENAGRIONIDAE WETLANDS apicalis x Moist-Soil Units 22 Argia tibialis x Seasonal Wetlands 17 Argia sedula xx Permanent Wetlands 8 Enallagma signatum xx Enallagma civile xxAQUATIC HABITATS Enallagma basidens xx Ponds and Lakes 24 posita xx Rivers and Creeks 17 Ischnura hastata xx Ischnura verticalis x oxbows and sloughs were damselflies. The dragonfly species composition in marshes total species richness for damselflies. Adults primarily consisted of Libellulidae (skimmers), and nymphs were collected for 8 species, adults while Libellulidae, Aeshnidae (darners), and only were collected for 4 species, and nymphs Gomphidae (clubtails) were common in ox- only were collected for 1 species (Table 2). bows and sloughs. Also striking was the low Species-habitat associations.—We sampled species richness (8 species) found in bottomland adults and nymphs from a number of natural hardwood wetlands relative to that in other (marshes, bottomland forest, oxbows, and naturally occurring wetlands on the Refuge. sloughs), managed (moist-soil units), and con- Managed and constructed wetlands: The structed (seasonal and permanent) wetlands, as managed moist-soil units (e.g., Cane Ridge well as from aquatic habitats (rivers, lakes and WMA and Dillin Bottoms) had the greatest ponds) on the Refuge. The odonates that were species richness among the constructed wetlands most widely distributed, occurring in every on the Refuge, even surpassing the richness of habitat, include Eastern Pondhawk (Erythemis the natural wetland types (Table 3). Seasonal simplicicollis), Blue Dasher (Pachydiplax long- wetlands (e.g., Branum, DuPont), those habitats ipennis), Widow Skimmer (Libellula luctuosa), likely to dry up annually, had species richness Common Whitetail (Libellula lydia), and Frag- similar to naturally occurring marshes. The spe- ile Forktail (Ischnura posita). Other very cies composition on moist-soil units and season- common odonates observed at most sites al wetlands was similar except that the moist- include Common Green Darner (Anax junius), soil units had more Sympetrum (meadowhawk Eastern Amberwing (Perithemis tenera), Hal- dragonfly), Lestes (spreadwing damselfly), and loween Pennant (Celithemis eponina), Familiar Enallagma (bluet damselfly) species present than Bluet (Enallagma civile), and Citrine Forktail did the seasonal wetlands. Permanent wetlands (Ischnura hastata). on the Refuge (e.g., Parke) had the lowest species Natural wetlands: Emergent marshes (e.g., richness among constructed wetlands. Snakey Point, Bucks Marsh) and various Aquatic habitats: The ponds and lakes on the disconnected oxbows and sloughs (e.g., Miller Refuge had the highest species richness (24 Bridge) both had an odonate species richness of spp.) of odonates of any habitat sampled 16 (Table 3). The dragonfly species richness (9 (Table 3). The species richness was also high spp.) and damselfly species richness (7 spp.) in rivers and creeks (17 spp.). A notable was more equitable in marshes than in oxbows difference among aquatic habitats on the and sloughs. Only 2 of the 16 species found in Refuge was a greater number of Aeshnidae 58 PROCEEDINGS OF THE INDIANA ACADEMY OF SCIENCE

Table 4.—The Nature Serve Ranks for Odonata in Indiana (Stein 2002, Faber-Langendoen et al. 2009). The rank, description of rank, and the number of species of odonates of that rank occurring on Patoka River National Wildlife Refuge and Management Area.

Rank Description of rank Species S5 Demonstrably secure in state, and essentially ineradicable under present conditions 15 S4 Apparently secure in state, with many occurrences (greater than 100 occurrences) 15 S3 Presence rare or uncommon in the state, ( 21 to 100 occurrences) 10 S2 Imperiled in state because of rarity (6–20 occurrences) OR because of some factor 3 making it vulnerable to extirpation S1 Critically imperiled in state because of extreme rarity (5 or fewer occurrences) OR 0 because of some factor making it especially vulnerable to extirpation

(darners), Gomphidae (clubtails), Cordulegas- species are considered to be imperiled (S2) by the teridae (spiketails), and Argia (dancers) species ranking system and warrant additional monitor- present in lakes, ponds, and streams compared ing to assess their status on the Refuge. These to numbers found in wetlands. species, all of which are dragonflies, include the Species status.—To evaluate the presence of Mocha Emerald (Somatochlora linearis), Arrow- rare, threatened, and endangered species of head Spiketail (Cordulegaster obliqua), and the odonates on the Refuge, we compared our Dragonhunter (Hagenius brevistylus). No odo- species list to the Nature Serve Rank for nates classified as critically imperiled (S1) in the Indiana (Table 4). Nature Serve Ranks (also state have been found on the Refuge. called Natural Heritage Ranks) is a means of At the national and global level all of the ranking the relative imperilment of species of dragonflies and damselflies occurring on the plants and on a global, national and Refuge are classified as secure. The only state level (Stein 2002 and Faber-Langendoen federally endangered dragonfly, Hine’s Emer- et al. 2009). The ranking addresses the conser- ald (Somatochlora hineana), has not been vation status of an organism by using a 1–5 located on Patoka River Refuge. numerical scale from most vulnerable (e.g., S1) New county records.—The Patoka River to most secure (e.g., S5). Therefore, an imper- Refuge covers parts of two Indiana counties: iled species (e.g., S2) would be one that is rare Gibson and Pike. Much of what we know or is vulnerable to extirpation because of some about the presence of odonate species in these factor such as habitat loss. two counties comes from Curry (2001) and Most of the dragonflies and damselflies (30 Abbott (2005). The historical records for spp.) found on the Refuge are considered secure Gibson County are more complete than for (S5 or S4) because they are common and occur in Pike County. Thus our survey adds no new a variety of habitats (Table 4). Another 10 species odonate records for Gibson County, whereas are uncommon (S3), but are not considered we added 19 new odonate records for Pike imperiled because they can be found consistently County (Table 5). Two of the new records were in appropriate habitats on the Refuge. Only three damselflies and the rest were dragonflies. Most of these new records represent common species Table 5.—New Odonata records for Pike County except for the Dragonhunter (Hagenius brevis- found in wetlands associated with Patoka River tylus), Russet-tailed Clubtail (Stylurus plagia- National Wildlife Refuge and Management Area. tus), Arrowhead Spiketail (Cordulegaster ob- liqua), Great Blue Skimmer (Libellula vibrans), Calopteryx maculata Celithemis eponina Argia sedula Libellula luctuosa and Blue-faced Meadowhawk (Sympetrum Basiaeschna janata Libellula pulchella ambiguum). Arigomphus villosipes Libellula vibrans Dromogomphus spinosus Somatochlora linearis DISCUSSION Hagenius brevistylus Odonate species.—To gain insights into the Stylurus plagiatus Sympetrum vicinum Odonata fauna at the Patoka River Refuge we Cordulegaster obliqua Tramea carolina made comparisons to other wildlife refuges Celithemis elisa Tramea lacerata with similar habitats and management goals. Celithemis fasciata The Sam D. Hamilton Noxubee National BATEMA ET AL.—ODONATA OF PATOKA RIVER REFUGE 59

Wildlife Refuge (hereafter Noxubee Refuge), at the Patoka River Refuge, while 2 and 3 located in east-central Mississippi, is dominated cruiser species were found at the Noxubee and by bottomland hardwood forests, moist-soil Muscatatuck Refuges, respectively. units, greentree reservoirs, flooded ditches, Species-habitat associations.—Low species ponds, and streams. Muscatatuck National richness in bottomland hardwoods was noted Wildlife Refuge (hereafter Muscatatuck Ref- in this survey (Table 3) relative to other uge), located in southeastern Indiana, is com- naturally occurring wetlands (e.g., marshes prised of various wetland and aquatic habitats and oxbows) and aquatic habitats (e.g., ponds including bottomland hardwood forests, moist- and rivers). Bried & Ervin (2005) made the soil units, marshes, creeks, and springs. In same observation and attributed the lower addition to having similar habitats that are species richness to odonate physiology. Adult suitable for odonates, the Patoka, Noxubee, odonates are ecototherms and require sunny and Muscatatuck Refuges also share manage- areas; they would not be as active in the shade ment strategies. All three refuges provide of a bottomland forest as in more open areas. habitat for wildlife, actively restore a mix of Additionally, forested habitat may obstruct habitats to maintain and increase biodiversity, mobility of adults relative to the more open, and document use by wildlife of refuge herbaceous habitat of marshes, lakes, and resources, especially species that are considered ponds. rare, threatened, or endangered. The moist-soil units at Cane Ridge and Dillin Odonata fauna inhabiting the wetland and Bottoms had greater species richness than that aquatic habitats of the Patoka River Refuge at other constructed wetlands on the Refuge show a remarkable similarity to that occupying (Table 3). Bried & Ervin (2005) also found the habitats at the Noxubee Refuge. Bried & moist-soil units to have greater species richness Ervin (2005) documented 42 species of odonates than other wetland habitats at the Noxubee at Noxubee Refuge, with 31 dragonfly species Refuge. Both vegetative structure and hydrol- and 11 damselfly species. This compares to the ogy are major mechanisms determining species 43 Odonata species documented in this study, richness and composition at these sites. Moist- with 30 dragonfly species and 13 damselfly soil units have a diversity of plant species and species. The only species occurring in all wetland structure that likely provide the appropriate types in the Noxubee study were the Eastern perch sites, territory boundary markers, and Pondhawk (Erythemis simplicicollis), Blue Dash- oviposition substrates for adults and adequate er (Pachydiplax longipennis), Common Whitetail habitat for feeding and protection from pred- (Libellula lydia), and Fragile Forktail (Ischnura ators for nymphs (Buchwald 1992). posita) (Bried & Ervin 2005). This list is identical Odonata species richness increases with to the most widespread species at Patoka River increasing hydroperiod (Corbett 1980, 1999). Refuge, except for the addition of the Widow However, Colburn (2004) noted that vernal Skimmer in this study. All of these species are pools, which undergo fluctuating water levels, among the most commonly observed and widely are adequate habitats for a number of odonate distributed species in the United States (Don- species. Moist-soil units, which are managed to nelly 2004a, b). simulate fluctuating water levels, are typically Curry (personal communication) surveyed drawn down to promote moist-soil plants that only the dragonflies of the Muscatatuck Refuge are attractive to wildlife (Fredrickson and and identified 33 species. This compares to the Taylor 1982). At moist-soil units we observed 31 species found at the Noxubee Refuge (Bried typical odonate reproductive behaviors such as & Ervin 2005) and 30 species at the Patoka oviposition attempts, tandem flights, copula- Refuge. A total of 17 dragonfly species (37%) tion, and protection of females by males. We were shared by all three refuges. Libellulidae found nymphs for many of the adults using was the dominant family found at all three moist-soil sites, suggesting successful use and refuges; 18 species at the Patoka Refuge, 19 perhaps preference of moist-soil units for species at the Noxubee Refuge, and 17 species reproduction. at the Muscatatuck Refuge. Similar distribu- Species status.—Most of the odonate species tions were noted for the other dragonfly documented for the Patoka River Refuge are families at the three refuges, except for Macro- commonly found in Indiana. Based on the miidae. No Macromiidae (cruisers) were found Nature Serve Ranks, 40 of the 43 species found 60 PROCEEDINGS OF THE INDIANA ACADEMY OF SCIENCE on the Refuge are either secure or, if they are We would also recommend a more quanti- classified as uncommon, can be consistently tative study on the odonate-habitat associa- found in appropriate habitat (Table 4). The tions to better identify what areas of the Refuge three species considered to be imperiled, are the are most beneficial for odonates. We have Arrowhead Spiketail (Cordulegaster obliqua), provided preliminary information on where the Dragonhunter (Hagenius brevistylus), and certain species of odonates occur on the the Mocha Emerald (Somatochlora linearis). Refuge, but we lack sufficient data to determine Curry (2001) lists the Arrowhead Spiketail as which species are transient and which are rare; found in only four counties in Indiana, permanent residents. This distinction is impor- and not in Gibson or Pike County. It is tant if one is to get a clear idea of how odonates typically found in small forest streams. The use the resources of the Refuge. nymph we collected was from Rock Creek, a A final recommendation would be to deter- small, sandy-bottomed stream running through mine if the development of an umbrella index a bottomland hardwood forest on the east side using Odonata would be a benefit for resource of the Refuge in Pike County. The Mocha professionals at the Refuge. The main advan- Emerald, which Curry (2001) lists as our most tage of an umbrella index is that it provides common emerald, was collected in adult form managers and conservation planners with from Rock Creek as well. The preferred habitat limited resources, personnel, and funds with a of the Mocha Emerald is along dark, tree-lined quick, relatively inexpensive method of assess- streams (Curry 2001), but until now was not ing ecosystem health and to develop priorities recorded in Pike County. (Bried et al. 2007). Curry (2001) describes the Dragonhunter as ACKNOWLEDGMENTS being locally common in a few counties in Indiana (but not found in either Gibson or Pike Funding for this project was provided by the County), with its preferred habitat being University of Evansville UExplore Program medium to large forest streams. We collected and the U.S. Fish and Wildlife Cooperative two specimens along the Patoka River on the Challenge Grant Program, U.S. Fish and east side of the Refuge in Pike County. One was Wildlife Service, Region 3. Thanks to James patrolling the Patoka River, and the other was Curry for help with the identification of some collected on a seasonal wetland site near the of the adult dragonflies; Bill McCoy for his Patoka River (i.e., DuPont wetland) in pursuit advice and for directing us to suitable sampling of another dragonfly. Although most of the sites; Dale Edwards and Cris Hochwender for odonates identified in this survey are secure in use of equipment and references at the Univer- the state and are relatively common, it would sity of Evansville; and to the students of the be worthwhile to investigate those species that summer Youth Conservation Corp program are considered imperiled and to better deter- for assistance in sampling. mine their status on the Refuge. LITERATURE CITED FUTURE RESEARCH Abbott, J.C. 2005. Odonata Central: An online resource for the Odonata of North America. Austin, We would recommend a continuation of the Texas. Available @ http://www.odonatacentral.com odonate survey on the Refuge to better (Accessed October 18, 2012). document the presence (or absence) of dragon- Batzer, D.P. & S.A. Wissinger. 1996. Ecology of flies and damselflies. We suspect that more insect communities in nontidal wetlands. Annual odonate species may be using the Refuge than Review of Entomology 41:75–100. were observed in our survey. One reason for Benke, A.C., G.M. Ward & T.D. Richardson. 1999. this was that high water prevented us from Beaver-impounded wetlands of the southeastern adequately sampling certain aquatic habitats coastal plain: Habitat-specific composition and dynamics of invertebrates. Pp. 217–245. In Inver- such as bottomland hardwoods, oxbows, and tebrates in Freshwater Wetlands of North Amer- rivers where certain taxa (e.g., Macromiidae) ica: Ecology and Management (D.P. Batzer, R.B. are found. A slightly earlier start to the Rader & S.A. Wissinger, eds.). John Wiley & sampling season may also pick up some of the Sons, Inc., New York. early emerging odonate species with short flight Bried, J.T. 2005. Community and conservation periods (e.g., Gomphidae (clubtails)). ecology of dragonfly and damselfly adults in BATEMA ET AL.—ODONATA OF PATOKA RIVER REFUGE 61

Mississippi wetlands. MSc. Thesis. Mississippi Faber-Langendoen, D., L. Masher, J. Nichols, K. State University, Starkville, MS. Snow, A. Tomaino, R. Bittman, G. Hammerson, B. Bried, J.T. & G.N. Ervin. 2005. Distribution of adult Heidel, L. Ramsey & B. Young. 2009. NatureServe Odonata among localized wetlands in east-central Conservation Status Assessments: Methodology for Mississippi. Southeastern Naturalist 4(4):731–744. Assessing Ranks. NatureServe, Arlington, VA. Bried, J.T., B.D. Herman & G.N. Ervin. 2007. Fairchild, G.W., A.M. Faulds & L.L. Saunders. Umbrella potential of plants and dragonflies for 1999. Constructed marshes in southeast Pennsyl- wetland conservation: a quantitative case study vania: Invertebrate food web structure. using the umbrella index. Journal of Applied Pp. 423–445. In Invertebrates in Freshwater Ecology 44:833–842. Wetlands of North America: Ecology and Man- Buchwald, R. 1992. Vegetation and dragonfly fauna: agement (D.P. Batzer, R.B. Rader & S.A. Wis- Characteristics and examples of biocenological singer, eds.). John Wiley & Sons, Inc., New York. field studies. Vegetatio. 101:99–107. Foote, A.L. & C.L.R. Hornung. 2005. Odonates as Burton, T.M., D.G. Uzarski, J.P. Gathman & J.A. biological indicators of grazing effects on Canadian Genet. 1999. Development of a preliminary prairie wetlands. Ecological Entomology 30:273–283. invertebrate index of biotic integrity for Lake Fredrickson, L.H. & T.S. Taylor. 1982. Management of Huron coastal wetlands. Wetlands 19:869–882. seasonally flooded impoundments for wildlife. U.S. Chovanec, A. & R. Rabb. 1997. Dragonflies (Insecta, Fish & Wildlife Service Resource Publication 148. Odonata) and the ecological status of newly Kennedy, C.H. 1950. The relation of American created wetlands: Examples for long-term bioin- dragonfly-eating birds to their prey. Ecological dication programmes. Limnologica 27:381–392. Monographs 2:103–143. Chovanec, A. & J. Waringer. 2001. Ecological King, R.S. & D.A. Wrubleski. 1998. Spatial and & integrity of river-floodplain systems: Assessment availability of flying insects as potential duckling by dragonfly surveys (Insecta, Odonata). Regu- food in prairie wetlands. Wetlands 18:100–114. lated Rivers Management and Research 17:493–507. McCoy, W. 2008. Patoka River National Wildlife Clark, T.E. & Michael J. Samways. 1996. Dragon- Refuge and Management Area: Comprehensive flies (Odonata) as indicators of biotope quality in Conservation Plan. Patoka River National Wild- Kruger National Park, South Africa. The Journal life Refuge, U. S. Fish and Wildlife Service. 164. of Applied Ecology 33:1001–012. Merritt, R.W. & K.W. Cummins. 1996. An Introduc- Clausnitzer, V., V.J. Kalkman, M. Ram, B. Collen, tion to the Aquatic Insects of North America. Third J.E.M. Baillie, M. Bedjanicˇ, W.R.T. Darwall, edition. Kendall/Hunt Publishers, Dubuque, Iowa. K.B. Dijkstra, R. Dow, J. Hawking, H. Karube, Needham, J.G. & M.J. Westfall, Jr. 1954. Dragon- E. Malikova, D. Paulson, K. Schu¨tte, F. Suhling, flies of North America. University of California R.J. Villanueva, N. von Ellenrieder & K. Wilson. Press, Berkeley & Los Angeles, California. 2009. Odonata enter the biodiversity crisis debate: Orians, G.H. & H.S. Horn. 1969. Overlap in foods The first global assessment of an insect group. and foraging species of blackbirds in the potholes Biological Conservation 142:1864–1869. of central Washington. Ecology 56:930–938. Colburn, E.A. 2004. Vernal Pools: Natural History Peyman, J. 2000. Dragonflies and Damselflies: Intro- and Conservation. McDonald & Woodward duction to the Odonates. Berkeley Press, California. Publishing Company, Blacksburg, Virginia. Sih, A. 1987. Predators and prey life style. Corbett, P.S. 1980. Biology of Odonata. Annual Pp. 203–224. In Predation: Direct and Indirect Review of Entomology. 25:189–217. Impacts on Aquatic Communities (W.C. Kerfoot Corbett, P.S. 1999. Dragonflies: Behavior and & A. Sih, eds.). University Press of New England, Ecology of Odonata. Cornell University Press, Hanover, New Hampshire. Ithaca, New York, USA. Silsby, J. 2001. Dragonflies of the World. Smithso- Curry, J.P. 2001. Dragonflies of Indiana. Indiana nian Institution Press, Washington, D.C, USA. Academy of Science. Stein, B.A. 2002. States of the Union: Ranking Donnelly, T.W. 2004a. Distribution of North Amer- America’s Biodiversity. NatureServe, Arlington, ican odonata. Part II. Macromiidae, Corduliidae VA. and Libellulidae. Bulletin of American Odonatol- Stewart, D.A.B. & M.J. Samways. 1998. Conserving ogy 8:1–32. dragonfly (Odonata) assemblages relative to river Donnelly, T.W. 2004b. Distribution of North Amer- dynamics in an African Savanna Game Reserve. ican odonata. Part III. Calopterygidae, Lestidae, Conservation Biology 12:683–692. Coenagrionidae, Protoneuridae and Platysticidae. Westfall, M.J. & M.L. May. 2006. Damselflies of North Bulletin of American Odonatology 8:33–99. America. Scientific Publishers, Gainesville, Florida. DuBois, B. 2005. Damselflies of the North Woods. North Woods Naturalist Series. Kollath-Stensaas Manuscript received 7 June 2011, revised 18 October Publishing, Duluth, Minnesota. 2012.