Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 501 2016 NORTHEASTERN NATURALIST 23(4):501–524 Distribution and Abundance of Odonata Species Across Massachusetts: Results of a Long-term Monitoring Program Robert Buchsbaum1,*, Christopher W. Leahy1, and Taber Allison1, 2 Abstract - Surveys of Odonata were carried out at Mass Audubon wildlife sanctuaries in all regions of the state and in multiple habitats. Our goals were to provide a comprehensive look at patterns of species distribution and relative species richness across Massachusetts and compare surveys where effort was and was not controlled. Observers encountered a total of 146 species, 11 of which were very widespread, having been recorded at more than 40 of the 54 properties examined. Thirty-five species were relatively rare, occurring at only 1 or 2 sanctuaries. A few sanctuaries were particularly notable for supporting somewhat uncommon species. These sites were not located in any particular ecoregion, but reflected local conditions. In surveys where effort was not controlled, a regression analysis indicated that about two thirds of the variation in species richness among sanctuaries could be explained by the amount of observer effort, the size of the sanctuary, and the extent of wetland habitat. Quantitative surveys that used transects or point counts to control for sampling effort resulted in observation of fewer species, including state-listed taxa, compared to the non-quantitative surveys. Despite producing fewer species, data from these quantitative surveys can be used to make statistical comparisons with data from future studies and detect changes over time in species richness, abundance, and frequency of occurrence. Introduction The insect order Odonata is a diverse and attractive group for ecological study. The order includes 2 suborders, Zygoptera (damselflies) and Anisoptera (dragonflies), both of which are characterized by aquatic nymphal stages and terrestrial adults. The aquatic nymphs occupy a variety of habitats depending upon the species, such as streams, rivers, ponds, lakes, bogs, and vernal pools, and they may be major predators or serve as prey for fish in some of these habitats (McPeek 2008, Paulson 2011). Odonate nymphs are often central to biological monitoring programs that assess stream- and river-habitat quality (Chovanec and Waringer 2001, Oertli 2008, USEPA 2002). Adults are highly visible, aerial predators on insects over aquatic communities, but can also be found far from water in fields, within forest openings, and on ridgetops. Some odonates are generalists and occur in a variety of habitats, and others are specialists that are limited to particular habitats, such as cold, rocky streams or coastal sandplain ponds. The diversity of odonate species found in the Northeast and the association of at least some species with particular ecological communities makes the group a suitable subject for comparing diversity and distribution across communities 1Mass Audubon, 346 Grapevine Road, Wenham, MA. 2Current address - American Wind Wildlife Institute, 1110 Vermont Avenue NW, Suite 950, Washington, DC 20005-3544. *Corresponding author - rbuchsbaum@massaudubon.org. Manuscript Editor: Pamela Hunt Northeastern Naturalist 502 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 and geographic ranges. New England is home to 188 species of odonates, of which 168 occur in Massachusetts (OdonataCentral 2016). With the aid of new field guides (e.g., Lam 2007, Nikula et al. 2007, Paulson 2011), it is possible to characterize the odonate fauna much more completely than it is for other groups of organisms that contain a much higher number of species and present greater identification challenges (e.g., beetles, moths). The recent popularity of “ode watching” also makes possible studies of odonate distribution based on species lists created by volunteer naturalists. In this study, we present the results of observations of adult odonates on 54 Mass Audubon wildlife sanctuaries located throughout the Commonwealth of Massachusetts. Our research examined (1) patterns of odonate species richness in particular regions and sanctuaries of Massachusetts, (2) the distribution of odonate species across Massachusetts to determine which ones were the most or the least widespread, (3) the influence of some physical characteristics (i.e., sanctuary size and extent of wetlands) on odonate species richness, and (4) the effect of observer effort on apparent richness. Alteration in the distribution of odonates has been suggested as an indicator of climate change (Brooks et al. 2007, Hassall and Thompson 2008, Parmesan 2006); thus, another goal of our inventory was to provide baseline data that could be used to assess future odonate distribution in response to a changing climate. Species lists are usually non-quantitative with only presence/absence recorded and no information given about sampling effort (temporal and spatial extent). We evaluated whether quantitative surveys could be incorporated into a monitoring program, and compared the results of quantitative to non-quantitative surveys. Methods Field site description Odonate surveys were conducted at 54 wildlife sanctuaries owned and managed by Mass Audubon, including at least 1 sanctuary in 12 of the 13 ecoregions defined by the US Environmental Protection Agency (USEPA) (MassGIS 2015; Fig. 1). These properties contain a variety of landscapes (e.g., forests, agricultural, suburban, and urban) and habitats (e.g., fresh and saline wetlands, ponds, rivers, streams, bogs and fens, fields, and ridgetops) that, in combination with regional geological and climatic differences, would be expected to support different odonate species. In total, the Mass Audubon sanctuary system is highly representative of the entire state, as well as much of southern New England. We used ArcGIS v.10 (ESRI 2012) in conjuction with a property-boundary data layer maintained by Mass Audubon to determine the area of the sanctuaries. Sanctuaries varied in size from 0.7 ha to 917 ha, with an average size of 197 ha (± 218 SD). For the wetlands calculations, we used the 2009 Massachusetts wetlands data layer from the Massachusetts Department of Environmental Protection (MassGIS 2011) after “clipping” it with the property boundary data layer. Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 503 Surveying odonates Our study focused on adult odonates. Odonates were sampled in 2 ways: general surveys to determine whether an odonate species was present at a particular location, and quantitative surveys within fixed areas. General surveys. General surveys were carried out by experienced observers, many of whom were volunteers, at different times of the field season with the goal of determining overall species richness of a sanctuary. The focus was on presence/ absence rather than on quantifying abundance or sampling effort. An observer walked through likely odonate habitat with a sampling net and recorded all species identified. In most cases, no information was collected about the sampling effort (e.g., amount of time spent in the field, number of habitats visited). Field notes occasionally provided a general sense of species abundances, but this information was not collected in a rigorous manner. Individual observers generally focused their efforts on one or a few sanctuaries, thus reducing the complication of having different observers with different abilities sampling the same site. These types of surveys were similar to odonate surveys carried out in many reserves and are analogous to the lists of observed species that birders keep. Beginning in 2004, we began to track sampling effort for these general surveys. We asked observers to note the amount of time they spent searching and to record basic weather information (temperature, cloud cover, wind speed). We divided Figure 1. The ecoregions of Massachusetts as defined by USEPA (MassGIS 2015). Locations of Mass Audubon sanctuaries included in this study are indicated by dots. Northeastern Naturalist 504 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 sanctuaries into sampling zones that contained specific habitats, and encouraged observers to relate their observations to one or more of those zones. We developed a scale to record abundance classes. We also conducted sampling at different times throughout the field season to account for seasonal changes. As a quality-control measure, we asked observers to photograph species that were challenging to identify so that experienced observers could confirm the identificati ons. We defined our index of effort for the presence/absence surveys as the number of days a sanctuary was sampled for odonates. This is not a precise measurement because it does not take into account the number of hours spent searching, the number of observers, or the area and habitats covered. Observers are often motivated by adding new species to a sanctuary list; thus, entries that were from shorter visits might often have been reports of a new species observed incidental to other activities on the sanctuary, thus complicating the relationship between the number of hours spent searching and the number of species recorded. However, most sampling involved a visit of several hours, and we think that our large sample-size minimizes the effect of this variation on our results. Effort ranged from 1 day to 433 days per sanctuary, with most sanctuaries subject to 2–20 days of sampling. Quantitative surveys. In 2005, we developed 2 types of quantitative surveys to more rigorously control for sampling effort and to provide information on relative abundances of odonates: transects and point counts. Their use depended upon the landscape, as described below. Transects were used when the terrain allowed an observer to sample a relatively uniform habitat, such as along an open pond shore or a field. Our method resembled the Pollard Walk used for censusing butterflies (Pollard 1977). An observer walked 50–100 m, depending on the landscape, and recorded all odonate species and the number of individuals of each that passed within 5 m of either side of the line. Thus, the area sampled a rectangle 10 m wide and 50–100 m long. Observers spent a minimum of 15 min on each transect. No specimens were collected; odonates for which a specimen was necessary for identification to species (e.g., Enallagma spp.) were only identified to genus. Observers sometimes used an insect net to briefly catch an odonate to aid in identification, but the time it took to examine a netted specimen was not considered when estimating the time spent along the transect in a sample run. Surveyors used point counts at sites where they could not freely walk a transect, e.g., a small opening on the shoreline of a densely vegetated pond or a bridge over a stream. The observer stood at an observation point for 10 min and recorded the genus, species, and number of individuals of all odonates that passed within 5 m of the sample point. We considered the area sampled to be a circle or a semi-circle with a radius of 5 m, depending upon whether the observation point was at the edge or in the middle of odonate habitat. We reduced the chances of recounting the same individual as it flew back and forth by estimating the maximum number of individuals of each species we could observe at any one point in time. As with transects, a net was used to capture specimens that could not be identified as they flew by or perched, but only after the 10-min sample period ended. Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 505 Point counts and transect surveys were carried out between 9 AM and 4 PM (EDT) in sunny weather when odonates tend to be most active. We calculated abundance as the number of individuals divided by the area sampled. All samples presented included ≥5 transects or point counts that encompassed different portions of the sampling season. Potential sources of error One source of error was in not detecting odonates that were present (MacKenzie et al. 2008). Failure to detect a species that was present could have occurred if the species was not active when the observer was present or was simply missed. We sampled mostly between 9AM and 4PM; thus, we could have missed some crepuscular species such as Neurocordula obsoleta (Umber Shadowdragon). Also, observers differed in their ability to detect odonates. Surveyors were more likely to miss smaller damselflies than the larger dragonflies; therefore, the error is probably not uniform across species. In general, we made observations during different times of the field season to take into account the seasonality of odonate species. Eight sanctuaries were sampled only once; therefore, the data from those sites reflect a bias toward the time of year of that sample. Another source of error is misidentification. The primary observers in this study were experienced with odonates. To enhance their identification skills, less-experienced observers went out in the field with more experienced observers. We asked observers to record their identifications only to the lowest taxonomic level for which they were certain. To avoid a potential source of error, we did not distinguish between Sympetrum rubicundulum rubicundulum (Say) (Ruby Meadowhawk) and S. internum Montgomery (Cherry-faced Meadowhawk), which are challenging to discern. We also asked observers to provide digital images or specimens to support their observations, particularly for less-common species. In almost all cases, the same observers sampled the same sanctuary or set of sanctuaries throughout the season, resulting in consistency in skill levels within sites. Data from different observers could have contributed to some of the variation in the number of species recorded among sites. Much of our focus was on patterns of species accumulation; thus, we do not believe that having different observers collecting data from the same site affected our results to any great extent. One possible source of quantitative error for both transects and point counts was counting the same individual more than once. Adult dragonflies often fly back and forth or “patrol” a fairly long stretch of shoreline, and thus, may encounter the sample area several times during the counting period. Observers used their best judgment to avoid counting the same individual more than once. As described above, we reduced the chances of double counting during our point counts by estimating the maximum number of individuals we could observe at any one point of time. It was also possible for the observer to miss individuals that were out of the circle or transect rectangle during the time period that the observer was present. We believe that the time periods were long enough to reduce that possibility. Northeastern Naturalist 506 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 Data analysis For our general surveys, we used regression analysis and analysis of variance to relate factors such as sampling effort, sanctuary area, the extent of freshwater wetlands, and ecoregion to species richness of those sanctuaries. We examined the regression residuals to determine which sanctuaries had actual odonate species richness values that differed the most from the predicted value based on effort and area, thereby determining the sanctuaries that were substantially more or less rich in species than expected. We used EstimateS (Colwell 2013) to compare species richness among transects and points, and to account for differences in the number of samples among sites. We employed the Chao2 estimate of sample-based species richness to project the number of species in a location by combining what was observed with a projection of how many species were likely to be missed based on the number of singleton and doubleton occurrences of repeated examinations of sample units. Results Sampling effort Our database contained over 13,900 records of odonates from our sanctuaries. About 2500 of these were from quantitative surveys (transects or point counts); the remainder were from general surveys. We defined as an “event” a sample from a delineated area on a sanctuary on a particular day; about 1300 separate sampling events were carried out. Eighty-seven percent of the observations were recorded from 2000 through 2013 (the latest year included in this paper). We also included 3 records from the literature: Erythemis simplicicollis (Eastern Pondhawk) from a Nantucket shoreline that is now part of a Mass Audubon sanctuary (Johnson 1930), Somatochlora linearis (Mocha Emerald) from Broadmoor Sanctuary in Natick (White et al. 1974), and Enallagma vernale (Vernal Bluet) from Cheshire Pond in Ashburnham (White et al. 1974). Patterns of species richness in general surveys Overall, 146 species of odonates were recorded on 54 Mass Audubon sanctuaries during general (non-quantitative) surveys, which represents 87% of the species known to occur in Massachusetts. Species richness at sanctuaries varied between 1 and 83 (Fig. 2). Most sanctuaries contained 20–50 species. Eleven species were particularly abundant and were recorded on more than 40 sanctuaries (Appendix 1). In contrast, 18 species were recorded at only 1 sanctuary and 17 species at only 2 sanctuaries (Appendix 1). These less-widespread species included 19 that are classified as rare or endangered by the Massachusetts Natural Heritage and Endangered Species Program (2016). Twenty-one species of odonates known to occur in Massachusetts were not recorded during this study (Appendix 1). Riverine odonates, in particular, were underrepresented. Mass Audubon sanctuaries harbored only 65% of species associated with rivers, as characterized in Nikula et al. (2007). Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 507 Factors influencing species richness at sanctuaries. Multiple regression analysis indicated that sampling effort, sanctuary area, and the percent of freshwater wetlands on a sanctuary explained about two thirds of the observed species richness on a sanctuary (R2 = 0.67, F = 37.3, P < 0.001). These 3 variables were each statistically significant (P < 0.01). Sampling effort, defined as the number of sample days at a sanctuary, was the most important factor associated with the variation in species richness in our general surveys. The log of sampling effort alone explained about half the variation in species richness in these general surveys (R2 = 0.54; Fig. 3). Adding sanctuary acreage to the multiple regression model increased the R2 value to 0.64, and a further increase in R2 to 0.67 was achieved by incorporating the percentage of the sanctuary area occupied by freshwater wetlands, including open water. The best model based on Akaike information criteria (AIC) included these 3 variables but did not include the interactions between them. We looked at other factors that might be associated with differences in species richness. We compared the average species richness of sanctuaries within different ecoregions and found no statistically significant difference associated with ecoregion. This finding suggests that there was a fair amount of variation in richness among sanctuaries within an ecoregion. We also compared urban to non-urban areas. Of the 3 sanctuaries in close proximity to urban land-uses (Boston, Worcester, and Attleboro), 1 contained only 12 species, while the other 2 cont ained 38 and 45 species, respectively—values close to the median richness value for all sanctuaries. We noted from an analysis of the residuals that many of the sanctuaries that were lower Figure 2. Location of Mass Audubon sanctuaries sampled in this study and their range of species richness values. We used the totals from both the general and quantitative surveys to calculate the richness values. There was 1 sanctuary with 2 species whose “dot” is hidden behind a much larger circle on Cape Cod. Northeastern Naturalist 508 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 in species richness than would be predicted by the model are in the eastern part of the state, and a number of these sanctuaries are in close proximity to the sea coast. Five sanctuaries were particularly noteworthy for harboring the greatest number of odonate species found only at those sites. These areas are of obvious importance from a conservation perspective. One site that contained 4 species not recorded elsewhere was in the Connecticut River valley and contained uncommon riverine species. Two sites containing 3 species not recorded elsewhere were in the Northern Worcester Plateau, an ecoregion harboring northern fauna and flora not common in much of the rest of Massachusetts. These 2 sanctuaries were both relatively large in area and contained a diversity of wetland types, including bogs, ponds, shrub swamps, sedge-dominated wetlands, and a cold stream. Another sanctuary that also contained 3 species not recorded at other sites was in the southern part of the state and was noteworthy for its diversity of wetland types—stream, lake, beaver pond, and marsh. One sanctuary on Cape Cod contained 2 odonate species not recorded at other sanctuaries. This site had species associated with coastal-plain ponds and the southeastern Massachusetts coastal plain. Species-accumulation curves at individual sanctuaries in general surveys. Sanctuaries had different-shaped curves representing the relationship between the number of sample days spent at that sanctuary and cumulative odonate species richness. We present 5 examples that represent typical patterns. In 2 of these, an asymptote was reached, but the number of samples it took to reach that asymptote differed (Oak Knoll Wildlife Sanctuary [OK] and Wellfleet Bay [WB]; Fig. 4]). In 2 others, the curve reached a “temporary” asymptote, but then increased with continued sampling (Endicott Wildlife Sanctuary [END]; Fig. 4 and Wachusett Meadow [WM]; Fig. 5). At END, the secondary increase occurred after 31 visits to Figure 3. Relationship between sampling effort and species richness in general surveys for adults. Effort is defined as the number of days a sanctuary was visited. Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 509 the sanctuary, and at WM, the increase occurred after 391 visits. At Skunknett River (SK), the curve started to flatten out, then had a secondary sharp increase in species richness before ultimately flattening out again (Fig. 4). The sanctuaries represented in Figure 4 were generally sampled by the same observer or team of observers over several years and covered all seasons and areas of the sanctuary, so the skills of the observers and the area and time of years sampled were similar throughout. Quantitative studies with points and transects Ninety-nine species were detected in our transect- and point-count surveys, including 3 that are on the state list of rare and endangered species (Massachusetts Natural Heritage and Endangered Species Program 2016). Species richness values derived from point counts and transects tended to be lower than those derived from Figure 4. Species-accumulation curves based on non-quantitative surveys. OK = Oak Knoll Wildlife Sanctuary, WB = Wellfleet Bay Wildlife Sanctuary, END = Endicott Wildlife Sanctuary, and SK = Skunknett River. Figure 5. Species-accumulation curve at Wachusett Meadow Wildlife Sanctuary. Note that the data are based on many more sample days than those shown in Figure 4. Northeastern Naturalist 510 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 our general, non-quantitative surveys at the same sanctuaries (Table 1). Part of that difference could be the greater amount of sampling for the general surveys in most of the sanctuaries where both types of sampling were carried out, but most of sanctuaries where sample sizes were similar still showed lower richness in the quantitative surveys. Species richness estimates measured by transects and point counts were not directly comparable because the exact habitat characteristics they sampled differed. Comparisons of the 2 methods are useful in determining the methods to be used in a monitoring program. When we pooled all transects and all points, transect surveys of odonates resulted in higher species richness, diversity (Fisher’s alpha), and abundance of odonates than did point counts (t–test: P < 0.01 for all parameters; Table 2). When we compared transects and point counts within individual sanctuaries, transects still yielded measurably higher species richness and abundance (paired t-test: n = 6, t = 2.74, P < 0.05; T = 4.01, P < 0.01 for abundances). Because of the seasonality of most odonate species, the average number of individuals per sampling event over the field season was highly variable. We chose to present maximum numbers of each species to represent peak abundance (Tables 3, 4). Using either transects or point counts, the most abundant species varied by region and habitat. Ischnura spp., Enallagma spp., Pachydiplax longipennis (Blue Dasher), Plathemis lydia (Common Whitetail), and Libellula incesta (Slaty Skimmer) were Table 1. Comparison of species richness from general (non quantitative) surveys and from pointcounts and transects. Data includes sanctuaries that had at least 5 sample days in both categories. # species # sample days Sanctuary General surveys Transects/points General surveys Transects + points Allens Pond 20 7 6 6 Arcadia 64 23 8 9 Ashumet 45 18 15 4 Cedar Pond 14 27 5 7 Daniel Webster 38 18 12 27 Endicott 23 32 9 39 Lake Wampanoag 37 26 5 7 Moose Hill 55 13 23 6 Oak Knoll 41 22 7 18 Sesachacha Heathlands 13 23 5 9 Skunknett River 49 31 25 20 Stony Brook 51 20 40 13 Wellfleet Bay 35 24 44 20 Whetstone Woods 52 31 15 7 Table 2. Species richness (Chao2), diversity (Fisher’s alpha), and abundance (# of individuals recorded per sample) in point-counts and transects. Number of stations Richness Diversity Abundance Points 21 17.9 3.6 8.2 Transects 12 31.2 5.7 24.0 Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 511 particularly abundant at several sample stations. The relatively high abundance of Enallagma recurvatum (Pine Barrens Bluet)—a Massachusetts state-listed species— at a site on Cape Cod is noteworthy from a conservation perspective. In addition to abundances, we examined the frequency of occurrence of species in transects representing 2 communities: freshwater ponds in the coastal zone (within 10 km of coast) and those farther inland (Tables 5, 6). The ponds were similar in size (less than 2 ha) and depth and had been sampled at least 5 times. Although there was some overlap in the species that are common in the 2 habitats, there were also clear differences. Species common in both communities included Ischnura Table 3. Most-abundant species recorded in transect surveys. Values are the 14 highest totals of individuals per species recorded at a transect station during any 1 sample-event. Transect lengths are all normalized to 100 m for comparison (i.e., the numbers are based on a survey area of 1000 m2). Genus/species Station habitat High count Ischnura verticalis Meadow 70 Ischnura verticalis Pond shore 68 Ladona julia Beaver pond shore 62 Ischnura posita Meadow 45 Pachydiplax longipennis Pond shore 40 Enallagma sp. Pond shore 38 Plathemis lydia Wetland edge 38 Lestes sp. Pond shore 30 Sympetrum rubicundulum/internum Pond shore 30 Chromagrion conditum Forest trail 30 Libellula incesta Boggy pond shore 29 Lestes vigilax Boggy pondshore 28 Enallagma hageni Beaver-pond shore 28 Enallagma recurvatum Coastal plain 26 Table 4. Fifteen most-abundant species recorded in individual point-count surveys. Values are the number of individuals per species recorded at a point-count station during any 1 sample-event. Area of count circle was 78.5 m 2. Genus Station habitat High count Plathemis lydia Wetland edge 50 Pachydiplax longipennis Pond shore 50 Plathemis lydia Pond shore 45 Enallagma ebrium Lake shore 30 Enallagma geminatum Lake shore 30 Enallagma aspersum Pond shore 25 Enallagma doubledayi Pond shore 25 Ischnura verticalis Pond shore 20 Enallagma aspersum Pond shore 20 Erythrodiplax berenice Salt marsh 16 Argia fumipennis River 15 Enallagma geminatum River 15 Pachydiplax longipennis Pond shore 15 Libellula incesta Pond shore 14 Calopteryx maculata River 14 Northeastern Naturalist 512 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 verticalis (Eastern Forktail), Anax junius (Common Green Darter), Plathemis lydia, and Libellula cyanea (Spangled Skimmer). However, Ladona julia (Chalk-fronted Corporal), the most widespread species in the inland pond transects, was not encountered at all at the coastal transects, nor were Cordulia shurtleffii (American Emerald) or Enallagma hageni (Hagen’s Bluet). Only 1 brackish pond was available on a Mass Audubon sanctuary. This pond contained a mixture of odonates known to occupy saline environments (e.g., Erythrodiplax berenice [Seaside Dragonlet], Enallagma durum [Big Bluet]) and those more typical of freshwater ponds (e.g., Erythemis simplicicollis). The odonate fauna of freshwater ponds in the coastal zone was more similar to this brackish pond than it was to inland freshwater ponds (Bray Curtis similarity index of 0.55 vs 0.43). In our quantitative surveys, curves representing the cumulative number of species per observation showed the expected steep increase during the first few observations followed by a gradual leveling off (Fig. 6). Very typically, however, the curves did not flatten out to an asymptote but showed a continual, gradual rise. Table 6. Occurrence frequency of odonate species at inland ponds >10 km from coast. Only the mostfrequently encountered species are shown. Data are from 2 transects sampled a total of 14 times. Species Percent frequency Ladona julia 71.4% Ischnura verticalis 64.3% Anax junius 50.0% Leucorrhinia frigida 50.0% Plathemis lydia 50.0% Enallagma spp. 42.9% Cordulia shurtleffi 35.7% Enallagma hageni 35.7% Leucorrhinia intacta 35.7% Libellula cyanea 35.7% Sympetrum rubicundulum/internum 28.6% Table 5. Occurrence frequency of odonate species at freshwater ponds near the coast. Only the mostfrequently encountered species are shown. Data are from 5 transects that were sampled a total of 55 times. Species Percent frequency Ischnura verticalis 66.0% Erythemis simplicollis 58.5% Pachydiplax longipennis 58.5% Libellula incesta 52.8% Anax junius 34.0% Ischnura posita 30.2% Plathemis lydia 30.2% Perithemis tenera 28.3% Sympetrum rubicundulum/internum 28.3% Enallagma spp. 22.6% Libellula cyanea 22.6% Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 513 Discussion Patterns of species richness Species richness did not vary geographically (e.g., latitude, longitude, ecoregion) across Massachusetts in our non-quantitative surveys. Many species were common in all parts of the state. Some exceptions are Enallagma hageni, Cordulia shurtleffi, and Ladona julia, which are common inland species with northern affinities and were not found in the southeastern Massachusetts coastal plain. Examples of odonates only recorded in the eastern part of the state include Enallagma durum (Big Bluet), E. doubledayii (Atlantic Bluet), and Libellula needhami (Needham’s Skimmer). These findings are consistent with the known distributions of these species (Nikula et al. 2007, Paulson et al. 2011). The absence of any geographic pattern in richness suggests that the number of regional “specialties” was similar throughout the state even if the species themselves differed. Odonate species richness at a particular sanctuary is likely a function of the diversity of habitats, primarily wetlands, within that sanctuary rather than of regional trends in Massachusetts. In our surveys for adult odonates where sampling effort was not controlled, the amount of sampling effort explained half the variation in species richness. Other contributing variables were sanctuary size and the percent of the sanctuary that is wetland. From the perspective of an overall trend, these are obvious explanatory variables. One would expect sanctuaries sampled more often to have higher species richness. Consistent with ecological theory, larger sanctuaries should generally encompass a greater variety of natural community types that could in turn support greater species richness of organisms. Odonate life cycles are closely associated with wetlands; thus, it is no surprise that sanctuaries with a larger percentage of wetlands tended to have more species, Figure 6. Species/observation curves for selected transects or points. Numbers are derived from estimated species richness using resampling procedures of EstimateS. FNt1 = brackish- pond transect on Martha’s Vineyard; SKt1 = sandy, bog-edged pond transect on Cape Cod; ENDt1 = wet meadow transect north of Boston; DWp1 = point near freshwater pond in grassland. Northeastern Naturalist 514 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 particularly if that larger percentage also indicates a greater diversity of wetland types. Most of the sanctuaries with lower diversity have few freshwater ponds and limited open, freshwater wetlands. Many of these sanctuaries are near the coast and dominated by salt marshes, which, in Massachusetts support only a few species. Only 1 odonate species is known to be limited to salt marshes in Massachusetts: Erythrodiplax berenice (Seaside Dragonlet). The area of lotic and lentic habitats relevant to odonates (e.g., cold- vs. warm-water streams, boggy vs. warmwater ponds, seeps, etc.) would have been a better metric than the simple percent of wetlands, but such a detailed classification has not yet been completed on sanctuaries in this study. Riverine odonate species were underrepresented in both our quantitative and non-quantitative surveys. This result is not surprising, given that many riverine species, such as a number of the Gomphidae, are rare and state-listed. Perhaps more importantly, our finding points to the priority of targeting conservation efforts toward riverine corridors, which are currently underrepresented at Mass Audubon sanctuaries. Volunteers often carried out our general surveys for adult odonates. It is highly likely that larger, more complicated and diverse sanctuaries would naturally be more attractive sample sites to odonate experts. Increased effort by volunteers at certain sanctuaries, therefore, does not only “cause” the greater richness we observed, but is also a reflection of greater observer interest in properties where more species are likely to occur, and the premium that is put on adding species to a sanctuary’s list. How many visits are necessary? One of the questions we were interested in addressing was how many visits to a sanctuary were necessary to characterize its odonate fauna. In carefully controlled sampling of 19 different lentic habitats carried out over 1 entire field season (15–20 weeks), Bried et al. (2011) recommended that biweekly surveys lasting 20–40 minutes each would be an adequate, cost-effective way to compare richness of adult odonates among different sites. All species, particularly rare or cryptic ones, may not be accounted for, so the goals of any study need to be taken into account. Our transect and point-count surveys differed from those of Bried et al. (2011) in that our sampling occurred over multiple field seasons and encompassed lotic and terrestrial habitats as well as lentic habitats. In addition, the number of observations our observers made varied at each site. The accumulation of species in our non-quantitative surveys did not consistently show an expected pattern of initial steep increase as more samples were added followed by a leveling off to an asymptote. We observed this pattern for data from several sanctuaries, but the number of visits required varied. Other sanctuaries did not reach an asymptote in the time frame of this study despite more than 20 days of observations spaced out over the entire flight season and that covered all areas of potential odonate habitat. Some curves reached an apparent asymptote after 20 or more sample days, but then entered another phase of increasing species accumulation with more observations. These inconsistent patterns could be at least partially Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 515 a function of the limited control of sampling effort in these largely volunteer surveys. However, even in our more-controlled sampling along transects or points, species richness at a number of sanctuaries did not completely flatten out after 20 visits. This finding suggests that some species were missed in earlier visits in both our non-quantitative and quantitative sampling, that the ability of the observer increased with more experience at a site, or that some species may have been missed in some years because their relative abundance fluctuated. Although carried out over a much longer time period than any of our observations, the observations of Shiffer and White (2014) of a Pennsylvania pond show that odonate species can appear, disappear, or decline to very low abundances, and then reappear over a period of decades in a given location. The value of multiple approaches This study incorporated data from surveys for adult odonates carried out by experienced naturalists where no attempt was made to control for time spent surveying or area. These surveys encompassed the variety of habitats present in the sanctuary. We also carried out quantitative surveys in which we sampled for a set distance (transects) or time and distance (point counts). Each transect or point was limited to one particular habitat and a limited area; thus, one would expect a smaller number of species than in the non-quantitative surveys from the same sanctuary. Our comparison of the results of these 2 types of surveys for odonates was similar to a report by Royer et al. (1998) for butterflies. The non-quantitative surveys were useful for developing a checklist of species for each sanctuary while engaging the public, but the results do not lend themselves to statistical comparisons. Their value from a conservation perspective was that they provided a checklist that indicates which species, including rarities, had been found. Of 70 records of odonates on Mass Audubon sanctuaries that are listed as endangered species in Massachussetts, only 9 were from our transect and point-count surveys, yielding a total of 3 listed species; the remainder were from the non-quantitative su rveys. In contrast, our modified Pollard Walk methodology used transects controlled for area searched. Our point-count procedure had a temporal as well as spatial limit. These methods yielded fewer species than the “checklist” approach, but do allow for statistical comparisons. Such quantitative surveys will be more useful for assessing changes in the odonate fauna in a monitoring program that examines the potential effects of water quality, ecological management measures, climate change, and other environmental stressors (Brooks et al. 2007, D’Amico et al. 2004, Hassall and Thompson 2008, Kutcher and Bried 2014, Mabry and Dettman 2010, Oerti 2008). Ultimately, multiple approaches are necessary to characterize the odonate fauna of a region. Our quantitative approaches, like that of Bried et al. (2011), provide a method for characterizing the species richness of odonate faunas that can be used for statistical comparisons, which is a critical factor if we are to use odonates as a barometer of environmental changes. Using these methods, longNortheastern Naturalist 516 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 term changes in richness and abundance should be evaluated on the level of individual sample units because sample sites are inherently variable due to differences among the habitats and area covered by transects and points or the walking routes of Bried et al. (2011). Such quantitative surveys can be supplemented by species searches when the goal is to completely characterize the odonate fauna of an area and include rarer species. Acknowledgments We gratefully acknowledge the field observations of the following individuals: Bob Bowker, Susie Bowman, Alexandra Brown, Martha Gach, Fred Goodwin, Gail Howe Trenholm, Rene Laubach, David Ludlow, David McLain, Paul Miliotis, Blair Nikula, Fred Saintours, Susie Schwoch, Kenneth Shea, Janet Sisterson, Brian Steinberg, Barbara Williamson, Taylor Yeager, and Vin Zollo. We thank Pamela Hunt (manuscript editor), and 2 anonymous reviewers for their helpful comments that greatly improved this manuscript. We also thank Robert Bertin and Jeff Collins who reviewed earlier drafts of the manuscript. Literature Cited Bried, J.T., B.J. Hager, P.D. Hunt, J.N. Fox, H.J. Jensen, and K.M. Vowels. 2011. Bias of reduced-effort community surveys for adult Odonata of lentic waters. Insect Conservation and Diversity. DOI:10.1111/j.1752-4598.2011.00156.x. Brooks, S., A. Parr, and P. Mill. 2007. Dragonflies as climate-change indicators. British Wildlife 19:85–93. Chovanec, A., and J. Waringer. 2001. Ecological integrity of river–floodplain systemsassessment by dragonfly surveys (Insecta: Odonata). Regulated Rivers: Research and Management. Volume 17:493–507. Colwell, R.K. 2013. EstimateS: Statistical estimation of species richness and shared species from samples. Version 9. User’s Guide and application. Available online at http://purl. oclc.org/estimates. Accessed 10 December 2013]. D’Amico, F., S. Darblade, S. Avignon, S. Blanc-Manel, and S.J. Ormerod. 2004 Odonates as indicators of shallow-lake restoration by liming: Comparing adult and larval responses. Restoration Ecology 12:439–446. ESRI. 2012. ArcGIS Desktop Release 10.0. Environmental Systems Research Institute. Redlands, CA. Hassall, C., and D.J. Thompson. 2008. The impacts of environmental warming on Odonata: A review. International Journal of Odonatology 11:131–153. Johnson, C.W. 1930. A list of the insect fauna of Nantucket, Massachusetts. The Nantucket Maria Mitchell Association 3(2):1–174. Kutcher, T.E., and J.T. Bried. 2014. Adult odonata conservatism as an indicator of wetland condition. Ecological Indicators 38:31–39. Lam, E. 2007. Damselflies of the Northeast. Biodiversity Books, Forest Hills, NY. 96 pp. Mabry, C., and C. Dettman. 2010. Odonata richness and abundance in relation to vegetation structure in restored and native wetlands of the prairie pothole region, USA. Ecological Restoration 28:475–484. MacKenzie, D.I., J.D. Nichols, J.A. Royle, K.H. Pollock, L.L. Bailey, and J.E. Hines. 2005. Occupancy Estimation and Modeling: Inferring Patterns and Dynamics of Species Occurrence. Elsevier Publishing, Amsterdam, The Netherlands. 344 pp. Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 517 Massachusetts Geographic Information Systems (MassGIS). 2011. Massachusetts Department of Environmental Protection 1:12000 wetlands data layer. Available online at http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/ office-of-geographic-information-massgis/datalayers/depwetlands112000.html. Accessed 9 June 2011. MassGIS. 2015. US EPA Ecoregions data layer. Available online at http://www.mass.gov/ anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information- massgis/datalayers/eco-reg.html. Accessed 5 April 2016. Massachusetts Natural Heritage and Endangered Species Program. 2016. Commonwealth of Massachusetts. Executive Office of Energy and Environmental Affairs. Department of Fish and Game. Division of Fisheries and Wildlife. Available onlin at http:// www.mass.gov/eea/agencies/dfg/dfw/natural-heritage/species-information-andconservation/. Accessed 6 June 2016 McPeek, M. 2008. Ecological factors limiting the distribution and abundances of Odonata. Pp. 51–62, In A. Cordoba-Aguilar (Ed.). Dragonflies and Damselflies: Model Organisms for Ecological and Evolutionary Research. Oxford University Press, Oxford, UK. 288 pp. Nikula B., J.L. Ryan, and M.R. Burne. 2007. A Field Guide to the Dragonflies and Damselflies of Massachusetts, 2nd Edition. Natural Heritage and Endangered Species Program. MA Division of Fisheries and Wildlife, Westborough, MA. 197 pp. OdonataCentral 2016. University of Alabama Museums. Available online at http://odonatacentral. bfl.utexas.edu/. Accessed 6 June 2016. Oertli, B. 2008. The use of dragonflies in assessing and monitoring aquatic habitats. Pp. 79–96, In A. Cordoba-Aguilar (Ed.). Dragonflies and Damselflies: Model Organisms for Ecological and Evolutionary Research. Oxford University Press, Oxford, UK. 288 pp. Parmesan, C. 2006. Ecological and evolutionary responses to recent climate change. Annual Review of Ecology Evolution and Systematics 37:637–69. Paulson, D. 2011. A Field Guide to Dragonflies and Damselflies of the East. Princeton University Press, Princeton, NJ. 376 pp. Pollard, E. 1977. A method for assessing changes in the abundance of butterflies. Biological Conservation 12:115–134. Royer, R.A., J.E. Austin, and W.E. Newton. 1998. Checklist and “Pollard Walk” butterflysurvey methods on public lands. American Midland Naturalist 140:358–371. Shiffer, C.N., and H.B. White III. 2014. Dragonfly and damselfly colonization of a large, semi-permanent Pennsylvania pond. Northeastern Naturalist 21: 6 30–651. US Environmental Protection Agency (USEPA). 2002. Methods for evaluating wetland condition: Developing metrics and indexes of biological integrity. EPA-822-R-02-016. Office of Water, Washington, DC. White, H.B., III, P. Miliotis, and C. Leahy. 1974. Additions to the odonata of Massachusetts. Entomological News 85:208–210. Northeastern Naturalist 518 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 Appendix 1. List of Massachussets odonates along with the number of sanctuaries at which each species was recorded during surveys. A “0” indicates that the species was not recorded. Rarity status according to the MA Natural Heritage and Endangered Species Program: E= endangered, T= threatened, and SC= special concern. Distribution codes and habitats within Massachusetts are based on Nikula et al. (2007): T= throughout, S = southern, SE = southeast, N= northern, W= central and western. Scientific name Common name # of sanctuaries Rarity Typical habitat Distribution Calopterygidae Calopteryx aequabilis Say River Jewelwing 4 Streams T Calopteryx amata Hagen Superb Jewelwing 1 Streams T Calopteryx dimidiata Burmeister Sparkling Jewelwing 0 Warm streams S Calopteryx maculata (Beauvois) Ebony Jewelwing 29 Streams T Hetaerina americana (Fabricius) American Rubyspot 1 Streams T Lestidae Lestes congener Hagen Spotted Spreadwing 11 General T Lestes disjunctus Selys Common Spreadwing 19 General T Lestes dryas Kirby Emerald Spreadwing 3 General T Lestes eurinus Say Amber-winged Spreadwing 8 General T Lestes forcipatus Rambur Sweetflag Spreadwing 16 General T Lestes inaequalis Walsh Elegant Spreadwing 17 General T Lestes rectangularis Say Slender Spreadwing 45 General T Lestes unguiculatus Hagen Lyre-tipped Spreadwing 3 General T Lestes vigilax Hagen in Selys Swamp Spreadwing 26 General T Coenagrionidae Amphiagrion saucium (Burmeister) Eastern Red Damsel 6 Bogs, boggy ponds T Argia apicalis (Say) Blue-fronted Dancer 3 General T Argia fumipennis (Burmeister) Variable Dancer 24 General T Argia moesta (Hagen) Powdered Dancer 3 General T Argia tibialis (Rambur) Blue-tipped Dancer 1 General S Argia translata Hagen in Selys Dusky Dancer 1 General S Chromagrion conditum (Selys) Aurora Damsel 22 Cold ponds T Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 519 Scientific name Common name # of sanctuaries Rarity Typical habitat Distribution Coenagrion resolutum (Hagen in Selys) Taiga Bluet 2 General N Enallagma annexum (Hagen) Northern Bluet 7 General T Enallagma aspersum (Hagen) Azure Bluet 17 General T Enallagma boreale Selys Boreal Bluet 0 General T Enallagma carunculatum Morse Tule Bluet 1 SC General T Enallagma civile (Hagen) Familiar Bluet 18 General T Enallagma daeckii (Calvert) Attenuated Bluet 0 SC General SE Enallagma divagans Selys Turquoise Bluet 4 General T Enallagma doubledayi (Selys) Atlantic Bluet 5 Coastal plain ponds S Enallagma durum (Hagen) Big Bluet 2 General S Enallagma ebrium (Hagen) Marsh Bluet 17 General T Enallagma exsulans (Hagen) Stream Bluet 3 General T Enallagma geminatum Kellicott, 1895 Skimming Bluet 20 General T Enallagma hageni (Walsh) Hagen's Bluet 15 General T Enallagma laterale Morse New England Bluet 4 Coastal plain ponds T Enallagma minusculum Morse Little Bluet 2 General N Enallagma pictum Morse Scarlet Bluet 2 T Coastal plain ponds S Enallagma recurvatum Davis Pine Barrens Bluet 1 T Coastal plain ponds S Enallagma signatum (Hagen) Orange Bluet 17 General T Enallagma traviatum Selys Slender Bluet 3 General S Enallagma vernale Gloyd Vernal Bluet 1 General N Enallagma vesperum Calvert Vesper Bluet 7 General T Ischnura hastata (Say) Citrine Forktail 6 General S Ischnura kellicotti Williamson Lilypad Forktail 12 General T Ischnura posita (Hagen) Fragile Forktail 46 General T Ischnura prognata (Hagen) Furtive Forktail 0 General S Ischnura ramburii (Selys) Rambur’s Forktail 2 Estuaries S Ischnura verticalis (Say) Eastern Forktail 48 General T Nehalennia gracilis Morse Sphagnum Sprite 19 Bogs, boggy ponds T Nehalennia irene (Hagen) Sedge Sprite 24 General T Northeastern Naturalist 520 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 Scientific name Common name # of sanctuaries Rarity Typical habitat Distribution Aeshnidae Aeshna canadensis Walker Canada Darner 20 General T Aeshna clepsydra Say Mottled Darner 7 General T Aeshna constricta Say Lance-tipped Darner 18 General T Aeshna eremita Scudder Lake Darner 0 Cold ponds N Aeshna interrupta Walker Variable Darner 3 General N Aeshna subarctica Walker Subarctic Darner 1 T Bogs, boggy ponds N Aeshna tuberculifera Walker Black-tipped Darner 22 Bogs, boggy ponds T Aeshna umbrosa Walker Shadow Darner 27 General T Aeshna verticalis Hagen Green-striped Darner 20 General T Anax junius (Drury) Common Green Darner 50 General T Anax longipes Hagen Comet Darner 5 SC Coastal plain ponds S Basiaeschna janata (Say) Springtime Darner 12 General T Boyeria grafiana Williamson Ocellated Darner 2 SC General T Boyeria vinosa (Say) Fawn Darner 14 Streams T Epiaeschna heros (Fabricius) Swamp Darner 13 General T Gomphaeschna antilope (Hagen) Taper-tailed Darner 2 Bogs, boggy ponds S Gomphaeschna furcillata (Say) Harlequin Darner 14 Bogs, boggy ponds T Nasiaeschna pentacantha (Rambur) Cyrano Darner 8 General T Rhionaeschna multicolor (Hagen) Blue-eyed Darner 0 General Vagrant W Rhionaeschna mutata (Hagen) Spatterdock Darner 3 SC General S/W Gomphidae Arigomphus furcifer (Hagen in Selys) Lilypad Clubtail 9 General S/W Arigomphus villosipes (Selys) Unicorn Clubtail 16 General S Dromogomphus spinosus Selys Black-shouldered Spinyleg 10 Rivers T Gomphus abbreviatus Hagen in Selys Spine-crowned Clubtail 2 E Streams T Gomphus adelphus Selys Moustached Clubtail 1 Rivers T Gomphus borealis Needham Beaverpond Clubtail 3 General N Gomphus descriptus Banks Harpoon Clubtail 2 E Streams S Gomphus exilis Selys Lancet Clubtail 19 General T Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 521 Scientific name Common name # of sanctuaries Rarity Typical habitat Distribution Gomphus fraternus (Say) Midland Clubtail 0 E Rivers T Gomphus lividus Selys Ashy Clubtail 3 Streams S/W Gomphus quadricolor Walsh Rapids Clubtail 0 T Rivers T Gomphus spicatus Hagen in Selys Dusky Clubtail 3 General T Gomphus vastus Walsh Cobra Clubtail 0 SC Rivers T Gomphus ventricosus Walsh Skillet Clubtail 1 SC Streams T Hagenius brevistylus Selys Dragonhunter 4 General T Lanthus parvulus (Selys) Northern Pygmy Clubtail 1 Streams T Lanthus vernalis Carle Southern Pygmy Clubtail 3 Streams T Ophiogomphus aspersus Morse Brook Snaketail 2 SC Rivers N Ophiogomphus carolus Needham Riffle Snaketail 1 T Streams N Ophiogomphus howei Bromley Pygmy Snaketail 0 Rivers N/W Ophiogomphus mainensis Packard in Walsh Maine Snaketail 0 Cold streams T Ophiogomphus rupinsulensis (Walsh) Rusty Snaketail 1 Rivers T Progomphus obscurus (Rambur) Common Sanddragon 1 General S Stylogomphus albistylus (Hagen in Selys) Least Clubtail 3 Streams T Stylurus amnicola (Walsh) Riverine Clubtail 0 E Rivers W Stylurus scudderi (Selys) Zebra Clubtail 2 Streams W Stylurus spiniceps (Walsh) Arrow Clubtail 2 Rivers W Cordulegastridae Cordulegaster diastatops (Selys) Delta-spotted Spiketail 9 Streams T Cordulegaster erronea Hagen in Selys Tiger Spiketail 0 Seeps, cold streams SW Cordulegaster maculata Selys Twin-spotted Spiketail 10 Streams T Cordulegaster obliqua (Say) Arrowhead Spiketail 2 Streams T Macromiidae Didymops transversa (Say) Stream Cruiser 12 General T Macromia illinoiensis Walsh Illinois River Cruiser 4 General T Cordulia shurtleffii Scudder American Emerald 8 General T Dorocordulia lepida (Hagen in Selys) Petite Emerald 17 General T Dorocordulia libera (Selys) Racket-tailed Emerald 18 General T Northeastern Naturalist 522 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 Scientific name Common name # of sanctuaries Rarity Typical habitat Distribution Corduliidae Epitheca canis (McLachlan) Beaverpond Baskettail 8 Bogs, boggy ponds T Epitheca cynosura (Say) Common Baskettail 25 General T Epitheca princeps Hagen Prince Baskettail 25 Lakes T Epitheca semiaquea (Burmeister) Mantled Baskettail 0 General S Epitheca spinigera (Selys) Spiny Baskettail 2 General T Helocordulia uhleri (Selys) Uhler’s Sundragon 2 Streams T Neurocordulia obsoleta (Say) Umber Shadowdragon 0 SC Rivers T Neurocordulia yamaskanensis (Provancher) Stygian Shadowdragon 1 SC Rivers T Somatochlora cingulata (Selys) Lake Emerald 0 Cold ponds N Somatochlora elongata (Scudder) Ski-tailed Emerald 4 SC Streams T Somatochlora forcipata (Scudder) Forcipate Emerald 1 SC Streams N Somatochlora georgiana Walker Coppery Emerald 0 E Warm streams SE Somatochlora incurvata Walker Incurvate Emerald 1 E Bogs, boggy ponds N Somatochlora kennedyi Walker Kennedy’s Emerald 0 E Bogs, boggy ponds N Somatochlora linearis (Hagen) Mocha Emerald 4 SC Rivers S Somatochlora minor Calvert in Harvey Ocellated Emerald 0 Cold streams T Somatochlora tenebrosa (Say) Clamp-tipped Emerald 17 General T Somatochlora walshii (Scudder) Brush-tipped Emerald 2 Streams T Somatochlora williamsoni (Walker) Williamson’s Emerald 6 Streams T Williamsonia fletcheri Williamson Ebony Boghaunter 1 E Bogs, boggy ponds N Williamsonia lintneri (Hagen in Selys) Ringed Boghaunter 4 T Bogs, boggy ponds N Libellulidae Celithemis elisa (Hagen) Calico Pennant 25 General T Celithemis eponina (Drury) Halloween Pennant 25 General T Celithemis fasciata Kirby Banded Pennant 3 General T Celithemis martha Williamson Martha’s Pennant 6 Coastal plain ponds S Erythemis simplicicollis (Say) Eastern Pondhawk 44 General T Erythrodiplax berenice (Drury) Seaside Dragonlet 7 Estuaries T Ladona deplanata (Rambur) Blue Corporal 4 Coastal plain ponds T Northeastern Naturalist Vol. 23, No. 4 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 523 Scientific name Common name # of sanctuaries Rarity Typical habitat Distribution Ladona exusta (Say) White Corporal 17 General T Ladona julia (Uhler) Chalk-fronted Corporal 19 General T Leucorrhinia frigida Hagen Frosted Whiteface 16 Bogs, boggy ponds T Leucorrhinia glacialis Hagen Crimson-ringed Whiteface 4 General N Leucorrhinia hudsonica (Selys) Hudsonian Whiteface 7 Bogs, boggy ponds N Leucorrhinia intacta (Hagen) Dot-tailed Whiteface 34 General T Leucorrhinia proxima Calvert Red-waisted Whiteface 7 General N Libellula auripennis Burmeister Golden-winged Skimmer 5 Coastal plain ponds S Libellula axilena Westwood) Bar-winged Skimmer 1 General S Libellula cyanea Fabricius Spangled Skimmer 37 General T Libellula incesta Hagen Slaty Skimmer 41 General T Libellula luctuosa Burmeister Widow Skimmer 35 General T Libellula needhami Westfall Needham’s Skimmer 9 General S Libellula pulchella Drury Twelve-spotted Skimmer 49 General T Libellula quadrimaculata L. Four-spotted Skimmer 24 General T Libellula semifasciata Burmeister Painted Skimmer 23 General T Libellula vibrans Fabricius Great Blue Skimmer 3 General S/W Nannothemis bella (Uhler) Elfin Skimmer 5 Bogs, boggy ponds T Pachydiplax longipennis (Burmeister) Blue Dasher 48 General T Pantala flavescens (Fabricius) Wandering Glider 21 General T Pantala hymenaea (Say) Spot-winged Glider 15 General T Perithemis tenera (Say) Eastern Amberwing 33 General T Plathemis lydia (Drury) Common Whitetail 45 General T Sympetrum corruptum (Hagen) Variegated Meadowhawk 0 General Vagrant W Sympetrum costiferum (Hagen) Saffron-winged Meadowhawk 3 General T Sympetrum obtrusum (Hagen) White-faced Meadowhawk 2 General T Sympetrum rubicundulum/internum Ruby Meadowhawk 52 General T Sympetrum semicinctum (Say) Band-winged Meadowhawk 15 General T Sympetrum vicinum (Hagen) Autumn Meadowhawk 42 General T Tramea abdominalis (Rambur) Vermilion Saddlebags 0 General Vagrant S Northeastern Naturalist 524 R. Buchsbaum, C.W. Leahy, and T. Allison 2016 Vol. 23, No. 4 Scientific name Common name # of sanctuaries Rarity Typical habitat Distribution Tramea calverti Muttkowski Striped Saddlebags 0 General Vagrant S Tramea carolina (L.) Carolina Saddlebags 9 General S Tramea lacerata Hagen Black Saddlebags 23 General S/W