Richness and Elevation Patterns of a Stonefly (Insecta, Plecoptera) Community of a Southern Appalachian Mountains Watershed, USA

Abstract

Protected areas are relatively free of human disturbance, are especially common in montane regions, and provide opportunities to study native fauna and flora. Stoneflies are model organisms to assess present and future environmental changes. While stoneflies inhabit cold lakes and a wide size range of lotic systems, diversity is greatest in streams draining mountain landscapes. This study addressed seasonal and elevation patterns of a stonefly fauna inhabiting a protected landscape draining the eastern flank of Mount Mitchell, the highest point of the Appalachian Mountains, USA. In total, 58 species were collected with estimated richness values ranging from 63 to 65. Species determinations were assisted with an integrative molecular approach using the mitochondrial barcode gene. Sampling during spring yielded the most species seasonally. Although certain species were only found at lower or higher sites, several were collected from across a broad range of elevations. Roughly 1/3 (21 = 36%) of the fauna present is known only from the southern Appalachian Highlands region, including one species described as new during this study. The assemblage reported here, however, did not closely align with other Appalachian fauna. Overall, well-structured faunal research continues to be important in light of continuing habitat modifications and climate change.

1. Introduction

Established protected areas are relatively free of human-induced disturbances and Libbyprovide opportunities to study native flora and fauna [1]. Protected areas are widespread throughout the United States of America (USA), typically in remote areas and especially within montane regions. Mountain regions represent approximately 25% of the earth’s land and contain diverse ecosystems [2]. Mountains are important components of climatic and hydrologic systems as they are a barrier to atmospheric flow and are the main sources of several major river systems globally.

Protected freshwater sites provide opportunities to address impacts throughout a catchment area [3]. Freshwater ecosystems are subject to several stressors, including pollutants [4,5], deforestation [6], invasion of non-native species [7], urbanization [8], and climate change [9,10]. Aquatic organisms are influenced by tolerance adaptations to natural thermal regimes as the majority can only endure a specific range of temperatures [11,12]. Freshwater systems are particularly vulnerable to climate change due to the variable thermal properties of water [13] and anticipated precipitation stochasticity [14].

Populations are responding to climate change by lessening population connectivity and changing geographic distributions [10,15]. Species ranges have shifted both in altitude and latitude [16,17]. North American fishes and aquatic macroinvertebrates have responded with a latitudinal shift in response to a 3–4 °C increase [18,19]. Extinction can only be avoided if organisms can move to a more favorable habitat [20]. Some mountain environments, including north–south ranges, have limited possibilities for species to migrate [21] and are therefore sensitive to alterations in climatic conditions [22]. Many species are mountaintop endemics to which climate warming would pose a large threat [23]. Low-elevation mountain ranges are especially susceptible to summit traps in which species will run out of space to occupy [11,24].

Taxonomic surveys can assess the influence of multiple stressors on freshwater species [25,26]. Macroinvertebrates are used in nearly 80% of all biomonitoring studies of lotic systems [27]. Stoneflies are used in water-quality biomonitoring studies due to their high sensitivity to environmental contamination and specified environmental tolerances [28]. Using Natural Heritage methodologies, stoneflies were assessed as the third most imperiled freshwater biotic group in the United States of America (USA), behind only mussels and crayfish [29]. Stoneflies are among the first aquatic insects to disappear from freshwater systems due to altered environmental conditions [30,31]. Stoneflies inhabit a broad range of stream sizes, elevations, streamflow permanence conditions, and thermal regimes [32], but species richness is highest in cold-water streams of montane regions. The most vulnerable are cold-stenotherm species at higher altitudes that are confined to upper reaches of catchments [33].

The southeastern USA in general has served as a biodiversity hotspot and has provided protection for regional biota. The overarching objective of this study was to perform a comprehensive study of the stonefly fauna of a southern Appalachian Highlands watershed in western North Carolina, USA, by collecting over different seasons of multiple years. More specifically, we addressed if (a) there were distinct elevation patterns present for individual families and species and (b) this watershed acted as a protective landscape for species that are considered to be rare, regionally restricted, or currently known only from a geographic small range.

2. Materials and Methods

2.1. Blue Ridge Region

This research was conducted along the eastern flank of Mount Mitchell, the southern extent of the north–south Black Mountains that are part of the ancient Blue Ridge region of the Appalachian Mountains that formed approximately 480 million years ago and spans portions of eight USA states from north Georgia to southern Pennsylvania (Figure 1). More specifically, Mount Mitchell is located within the U.S. Environmental Protection Agency (USEPA) Level III Blue Ridge Ecoregion (66). In North Carolina, this ecoregion is equivalent to the Mountains Ecoregion in Beaty [34]. Mount Mitchell is also the highest point of eastern North America at 2037 m (=6684 ft). Igneous and metamorphic rocks allowed Mount Mitchell to erode at a slower pace compared to other subranges in the southern Appalachian Highland region [35].

After the decline in Fraser fir (Abies fraseri (Pursh) Poir) in the 1960s and 1970s, this species has returned as a co-dominant canopy species at higher elevations [36], together with red spruce (Picea rubens Sarg.) and eastern hemlock (Tsuga canadensis (L.) Carrière). Several hardwoods are also present, including mountain ash (Sorbus americana Marshall), yellow birch (Betula allegheniensis Britt.), and fire cherry (Prunus pensylvanica L.f.) [37].

Mean regional temperatures range from −3.0 °C (January) to 19.3 °C (July) and precipitation occurs across all seasons, totaling nearly 110 cm of rainfall annually [38]. Streams draining the Black Mountains are part of the upper Tennessee River Basin and are nested between the South Toe River and French Broad River subbasins [39]. High-elevation springs, seeps, and streams have a primarily boulder-dominated substrate, whereas the lower-elevation higher-order streams have a gravel–pebble substrate intermixed with boulders. Regional tributaries are generally cold and remain this way even as they enter larger rivers at valley floors [40].

2.2. Field Methods

Sampling occurred within the Mount Mitchell State Park (MMSP) and the adjacent Appalachian Ranger District of Pisgah National Forest (PNF) in 2014–2017 and 2019. Sites sampled in 2014–2017 were assessed for gaps in elevation and drainage area. New sites in 2019 were added to address these gaps. Field collections occurred mainly between spring and autumn based on weather conditions at the MMSP (https://www.ncparks.gov/mount-mitchell-state-park/home, accessed on 30 March 2019). The Blue Ridge Parkway leading into the access road to the MMSP is closed during harsher winter weather. Sites were accessed mainly along trails, forest service roads, and at campgrounds along an elevation gradient of ca. 1150 m (831–1983 m ASL;

Table 1. Locality information for 43 unique sites at Mount Mitchell State Park and adjacent Pisgah National Forest, North Carolina, U.S.A. Sites are listed in order of most to fewest species collected.

Site #	Water Body	Longitude	Latitude	Altitude (m)	Drainage Area (km2)
1	South Toe River	35.75176	−82.22033	911	29.27
2	Lower Creek	35.75564	−82.26803	1719	0.54
3	Neals Creek	35.74176	−82.21492	954	2.93
4	Lower Creek	35.75803	−82.26772	1764	0.39
5	Left Prong South Toe River	35.71130	−82.25058	1184	1.79
6	Right Prong South Toe River	35.72900	−82.28330	1648	0.22
7	Big Lost Cove Creek	35.74298	−82.21127	950	0.21
8	Hemphill Creek	35.71024	−82.25057	1192	1.35
9	tributary Lower Creek	35.75630	−82.26460	1747	0.14
10	Right Prong South Toe River	35.73015	−82.28414	1684	0.12
11	South Fork Upper Creek	35.73634	−82.27946	1705	0.22
12	South Toe River	35.74417	−82.22833	934	25.80
13	Little Mountain Creek	35.75202	−82.22534	937	0.12
14	unnamed seep	35.74279	−82.27468	1703	0.09
15	Lower Creek	35.75953	−82.26683	1800	0.13
16	Setrock Creek	35.75561	−82.24325	1471	0.28
17	Setrock Creek	35.75369	−82.23779	1373	0.47
18	tributary Setrock Creek	35.75393	−82.23773	1373	0.09
19	tributary Right Prong South Toe River	35.72718	−82.28379	1636	0.54
20	Thee Creek	35.78222	−82.25435	1741	0.12
21	Setrock Creek	35.75047	−82.23009	1016	0.85
22	tributary Hemphill Creek	35.71005	−82.24959	1206	0.06
23	Thee Creek	35.78013	−82.25093	1600	0.62
24	Upper Creek	35.73135	−82.23858	984	6.60
25	Lower Creek	35.73407	−82.23466	969	4.14
26	unnamed seep	35.75298	−82.26823	1704	0.08
27	Balsam Spring	35.76647	−82.26406	1983	0.01
28	tributary Little Mountain Creek	35.75627	−82.23494	1044	0.12
29	(left) tributary South Toe River	35.71342	−82.24881	1154	3.55
30	rock faces	35.75592	−82.23534	1320	0.14
31	Right Prong South Toe River	35.73095	−82.28487	1723	0.12
32	Middle Fork Rock Creek	35.76408	−82.25726	1781	<0.01
33	South Toe River	35.74021	−82.23237	960	23.36
34	Right Prong South Toe River	35.72194	−82.24888	1074	4.12
35	unnamed seep	35.75655	−82.23523	1320	<0.00
36	Lower Creek	35.75694	−82.26825	1746	0.47
37	North Fork Rock Creek	35.77470	−82.25595	1775	0.09
38	South Toe River	35.74000	−82.23166	950	23.36
39	unnamed seep	35.75369	−82.23779	1373	0.47
40	South Toe River	35.80400	−82.20570	831	84.95
41	Rock Creek	35.77315	−82.21896	1017	<0.00
42	Spring	35.74395	−82.28172	1847	0.01
43	Hemphill Creek	35.71030	−82.25784	1332	1.48

). Location data in decimal degrees for each locality were recorded directly on site with a portable GPS unit and were georeferenced for accuracy using Acme Mapper 2.2 (https://mapper.acme.com, accessed on 20 August 2022). A dot distribution map for all collection localities was prepared with ArcMap 10.8 (Figure 2). The drainage area of each site was determined using the CONTDA function in StreamStats 4.0 (https://streamstats.usgs.gov/ss/, accessed on 20 August 2022). The drainage area was used as a proxy of stream size. The elevation of each site was determined using GPS Visualizer (https://www.gpsvisualizer.com/elevation, accessed on 20 August 2022).

During each site visit, adult stoneflies were collected using an aerial net and a beating sheet to dislodge individuals from riparian vegetation plus by hand-picking from emergent leaf packs, rocks, and trees. An ultraviolet light (UV) was deployed on four warm summer evenings to collect adults of Acroneuria, Beloneuria, Perlesta, and Isoperla. The UV light was used at only one site (site 1—see Table 1). Males of each of the four aforementioned genera have an aedeagus that needs to be fully extruded to ensure positive identification. Therefore, males were kept alive in 50 mL polypropylene conical tubes and the aedeagus was extruded under a Leica S6E dissecting microscope. Larval samples were collected using a standard dip net and by hand-picking from rocks and debris in October 2019 to ensure representation of larger-bodied individuals of Perlidae and Perlodidae. We used a structured, sampling design of exactly 15 min to provide equal effort across all 43 sites. All specimens were preserved in 95% ethanol. Sites were assigned a unique field number identifier and specimens were stored in the Western Kentucky University Collection, Bowling Green (WKUC).

2.3. Microscopy and Molecular Methods

Most individuals were identified in terms of species using Olympus SZ61 and SZX12 stereo microscopes (

Table 2. Taxonomic references used to identify adult and larval stoneflies in terms of family, genus, and species. References are organized by family and genus (if applicable).

Family	Genus (If Applicable)	References
All families		[43]
Capniidae	Allocapnia	[44,45]
	Paracapnia	[46]
Chloroperlidae	all genera	[47]
	Alloperla	[47,48]
	Sweltsa	[47,49]
Leuctridae	all genera	[50]
	Leuctra	[51,52,53,54,55]
	Megaleuctra	[56]
	Paraleuctra	[57]
Nemouridae	all genera	[58,59]
	Amphinemura	Grubbs and Baumann (unpublished key)
	Soyedina	[60]
Peltoperlidae	all genera and species	[61]
Perlidae	all genera (larvae only) and species	[62,63]
	Acroneuria (larvae only)	[50]
	Paragnetina (larvae only)	[64]
Perlodidae	to Perlodinae genera (larvae only) and species)	[62,65]
	Isoperla	[66,67,68,69]
	Remenus	[70]
Pteronarcyidae	Pteronarcys (larvae only)	[71]
Taeniopterygidae	all genera/species	[72]

). Females of Leuctra, Isoperla, and Remenus, however, are difficult to identify in terms of species. An integrated molecular approach was necessary to tease out richness patterns more accurately for these genera. The mitochondrial cytochrome c oxidase subunit 1 (mtCO1) gene for 41 males and 282 females of these 3 genera was extracted, amplified, sequenced, and analyzed with phylogenetic methods as outlined in [41,42].

Total DNA was extracted from two adult legs and associated thoracic tissue per individual using a DNeasy Blood and Tissue Kit according to the manufacturer’s instructions. The mtCOI gene was amplified through a polymerase chain reaction (PCR) using the primers HCO2190 (5′-TAAACTTCAGGGTGACCCAAAAAATCA-3′) and LCO1490 (5′-GGTCAACAAATCATAAAGATATTGG-3′ [73]. The PCR products were sequenced at the North Carolina State University Genomics Lab (Raleigh, NC, USA).

Molecular data were analyzed with each of tree- and genetic-distance phylogenetic methods. A maximum-likelihood tree was generated using RAxML with the autoMRE setting. A Bayesian-inference tree was constructed using MrBayes MPI with the following settings: GTR_I_G model; default Markov chain Monte Carlo commands (four chains, three heated); default number of runs (n = 2); stationary nucleotide frequency; burn-in of 25%; random starting tree; a sampling frequency of 1000 generations; and a stop value between 0.01 and 0.05. A consensus tree was generated following the stop rule.

Pairwise divergence distances were calculated in MEGA X [74] using the Kimura two-parameter model of nucleotide substitution with the complete codon deletion option. We focused on the maximum intraspecific divergence between males and putative females of the same species and the minimum interspecific divergence between all species. In addition, Automatic Barcode Gap Discovery (ABGD [75]) was employed to complement the pairwise divergence approach. Relative gap width was set at 1.5 with the Kimura two-parameter model and all other settings at default.

2.4. Data Analyses

Species data were organized in matrix format by incidence (1 = collected, 0 = not collected) by sampling trip and site. The freeware program EstimateS [76] was used to calculate two estimators of species richness (Chao2 estimator and ICE mean) and the number of singletons and doubletons. Both richness estimators are non-parametric models that are appropriate for use with incidence data [77]. While the Chao 2 estimator focuses on rare species (i.e., singletons and doubletons [78]), ICE takes into account both rare and common (i.e., in 10 or more samples [79]). We compared study-scale proportionality to state-level proportionality, the latter derived from the species list for North Carolina available as a faunal list search in Plecoptera Species File [80]. Proportionality was calculated at the family level as the number of species within a family divided by the total number of species present in the study area or state.

We assess the relationship of the Mount Mitchell fauna to 11 eastern Nearctic assemblages. We amplified the presence–absence CSV data file from DeWalt and Snyder [81] (p. 73) with [82] (Great Smoky Mountains National Park, North Carolina and Tennessee) [83], (Talladega Mountains, Alabama) [84], (Mammoth Cave National Park, Kentucky), and this study (Supplementary Table S1). The data list by Parker et al. [82] was amended with Soyedina parkeri [85] in lieu of Soyedina, i.e., “new species” [82], and Zapada fumosa [86]. A Jaccard distance matrix between faunal assemblages was constructed and then analyzed on average linkage and percent distance in SYSTAT 11. Faunal relationships were displayed as a simple dendrogram.

3. Results and Discussion

3.1. Species Richness and Community Composition

This study addressed species richness patterns of stoneflies occupying streams draining the highest-elevation mountain peak in eastern North America. Intensive and thoughtful faunistic studies, especially within protected landscapes, have the capacity to address specific ecological questions. There are several recent examples that have addressed questions regarding Nearctic stonefly communities, including altitudinal zonation patterns [83], dispersal capacity [87], habitat suitability [88], and streamflow duration [84,89].

In total, the 2014–2019 sampling resulted in 726 records (=vials) and 3128 adults and larvae in total from 43 unique locations (Supplementary Table S2). Over 80% of the sampling trips added unique species (Figure 3). Although sampling efforts were amplified during 2019, relatively few additional species were collected across the final five months of collecting. Between 1 and 31 species were collected from each unique location; this broad range was obtained because some sites were sampled more often than others (Figure 4).

Morphological and molecular methods resulted in 58 verified species, including an undetermined Leuctra OTU (

Table 3. Species list for Mount Mitchell State Park and adjacent Pisgah National Forest, North Carolina, USA. Families are presented in alphabetical order. * = new state records, ** = species with known distributions only in the southern Appalachian Highlands region.

Species	Total No. Collections	No. Unique Sites	Species	Total No Collections	No. Unique Sites
Family Capniidae			Family Perlidae
Allocapnia harperi *	1	1	Acroneuria abnormis	9	3
Paracapnia angulata	19	14	Acroneuria carolinensis	5	3
			Beloneuria georgiae **	1	1
Family Chloroperlidae			Beloneuria stewarti **	2	2
Alloperla chloris	3	2	Eccoptura xanthenes	5	5
Alloperla nanina **	4	2	Paragnetina immarginata	5	3
Alloperla neglecta **	3	2	Perlesta frisoni *	6	4
Alloperla petasata	3	3	Perlesta nelsoni	4	3
Alloperla usa	10	7
Suwallia marginata	9	3	Family Perlodidae
Sweltsa lateralis	33	23	Isoperla arcana/holochlora	6	6
Sweltsa onkos *	10	7	Isoperla cherokee **	5	5
Sweltsa urticae **	24	16	Isoperla dewalti **	7	6
			Isoperla kirchneri	1	1
Family Leuctridae			Isoperla orata	2	1
Leuctra carolinensis	6	5	Isoperla pauli **	5	4
Leuctra ferruginea	64	24	Isoperla powhatan	3	3
Leuctra grandis	54	31	Isoperla pseudosimilis **	2	2
Leuctra mitchellensis **	23	20	Isoperla starki **	3	3
Leuctra sibleyi	19	13	Isoperla stewarti **	3	3
Leuctra tenella	8	8	Isoperla WKUC 1	2	2
Leuctra tenuis	11	7	Isoperla WKUC 2	1	1
Leuctra triloba	35	27	Cultus decisus	1	1
Leuctra truncate *	19	19	Remenus bilobatus	1	1
Leuctra variabilis	3	3	Remenus WKUC 1	1	1
Megaleuctra williamsae **	2	2	Malirekus hastatus **	8	8
Paraleuctra sara	34	25	Yugus arinus **	2	2
			Yugus bulbosus **	2	2
Family Nemouridae
Amphinemura appalachia	6	4	Family Pteronarcyidae
Amphinemura nigritta	3	1	Pteronarcys scotti **	1	1
Amphinemura wui	50	26
Soyedina sheldoni **	16	14	Family Taeniopterygidae
			Oemopteryx contorta	6	6
Family Peltoperlidae			Strophopteryx limata	6	6
Tallaperla anna **	18	14
Tallaperla maria	15	14

) and additional information for adult Leuctra female presence. For example, L. grandis females were identified that had extended presence through late July, speaking to the importance of accurate identifications, even if with integrated molecular methods.

The most common species were Leuctra grandis (n = 31 sites), L. triloba (n = 27), Amphinemura wui (n = 26), Paraleuctra sara (n = 25), L. ferruginea (n = 24), and Sweltsa lateralis (n = 23) (Table 3). Nine of the ten families known from the Nearctic were represented (Table 3. The most speciose were Perlodidae (n = 18), Leuctridae (n = 12), Chloroperlidae (n = 9), and Perlidae (n = 8). The least speciose family was Pteronarcyidae (n = 1), represented by one Pteronarcys scotti exuvium collected along the South Toe River (site 1, Table 1).

Species richness was both similar to and less than the three richness estimators (Figure 5). The Chao2 estimator and ICE mean predicted 63 and 65 species. Approximately 33% (ten singletons, nine doubletons) of species were collected from only one or two locations (Figure 6), and nearly 90% of all species were collected at less than half of the unique locations.

Two sites with the least number of species appeared to be intermittent and lacked visible surface flow during summer and early autumn. Many species found in intermittent streams are also found in perennial systems [90,91], but only a subset of taxa can exist in the harsh drying conditions of seasonal channels [92]. Species that inhabit intermittent streams typically have univoltine fast life cycles and desiccation-resistant dormant stages in which development and reproduction occur during briefer periods of flow [93]. Leuctra tenella and Soyedina sheldoni were found at the intermittent Railroad Grid spring. Specific life-cycle traits are unknown for these two species. While western USA S. interrupta and S. producta have univoltine fast life cycles [94], eastern USA S. vallicularia and S. calcarea have univoltine slow life cycles [95,96]. Leuctra species are also split between univoltine slow and univoltine fast life cycles [97].

3.2. Proportionality and Comparisons to Other Fauna

Study-specific vs. state-wide proportionality was roughly comparable for Nemouridae, Peltoperlidae, Perlodidae, and Pteronarcyidae (Figure 7). For example, we only collected one species of Pteronarcyidae. However, the ratio of 1/58 study-scale species was proportionally similar to the state (=4/161). The low richness of Nemouridae (four species) was due to having only one Nemourinae (S. sheldoni) collected during this study. Ostrocerca, Paranemoura, Prostoia, three additional species of Soyedina, and Zapada have all been reported in the Blue Ridge Ecoregion in North Carolina [34,60]. Perlodidae make up the richest family in the Blue Ridge region with 39 species [34,66,67,68,69,70] and is likewise the richest in the state (n = 53; [34,43]). Perlodidae were found in relatively low abundance (only 5% of the total stoneflies) in this study, but still made up the richest family in the study area. The integrated molecular approach was not able to confidently place species names on several Isoperla in the absence of males (i.e., I. arcana/holochlora, I. WKUC 1, I. WKUC 2). Molecular methods did, however, yield the separation of two Remenus females into distinct species units [41]. Larval collecting added species records, but six regional genera were not found during this study: monotypic Clioperla clio, Diploperla, Helopicus, Hydoperla, Isogenoides, and monotypic Oconoperla innubila.

There are two notable examples regarding low proportionality (Figure 7). The first is Capniidae, which emerge in winter and spring in the eastern Nearctic [62,98]. Because sampling sites are much less accessible during winter, only Allocapnia harperi and Paracapnia angulata were collected during this study. Seven other species of Allocapnia Claassen, 1928 have been reported from the Blue Ridge [34]. Two, the only Taeniopterygidae collected, were Oemopteryx contorta and Strophopteryx limata. Like Capniidae, taeniopterygid adults emerge during winter and early spring, and sampling limitations may have led to fewer than anticipated species. Ten species in total have been reported [34], including Bolotoperla rossi, Taenionema atlanticum, and several species of Taeniopteryx.

In contrast, Chloroperlidae and Leuctridae proportionality was notably higher within the study area compared to the state. For example, 12 Leuctridae species were collected during this study. This is nearly similar to the number reported from the Blue Ridge within the state (=14; [34]). Molecular methods helped to positively identify approximately 35% of the total Leuctra females collected, including an undetermined Leuctra OTU [42] and 10 species in total. Paraleuctra sara and Megaleuctra williamsae were the other two species collected during this study.

The clustering of the 11 eastern Nearctic faunal assemblages was partially a function of geographic proximity but imperfect (Figure 8). The middle cluster (i.e., Mammoth Cave NP down through the Crane Hollow Nature Preserve) represents a Midwestern USA grouping. The three southern Appalachian assemblages (Talladega Mountains, Great Smoky Mountains NP, and this study), however, have less faunal overlap. Grubbs and Sheldon (2018) reported a species richness value (=57) from the Talladega Mountain region of eastern Alabama that is nearly identical to that of this study (n = 58). However, as many as seven species may be endemic to the Talladega Mountains and the two study localities/regions are minimally distant from each other by 355 km. Although the high peaks of the MMSP (Mount Mitchell) and the Great Smoky Mountains National Park (Clingmans Dome) are only approximately 115 km apart, 111 species were reported in the latter [82] from a markedly larger landscape, representing a composite of data accumulated over a much longer time period.

3.3. Seasonal Trends

Adults of most species were present for less than three months (Supplementary Table S3), yet there were three collected from late spring through autumn (L. ferruginea, L. truncata, and A. wui). Extended adult presence of L. ferruginea has been documented previously [83,95]. Eighteen species were present in both spring and summer, followed by only four collected in summer and autumn. Paracapnia angulata was the only species collected in both winter and spring, but this is an artifact of data deficiency (Supplementary Table S3). Adult data are missing from December–January. Few data were available for winter (one day in February 2014) and accordingly with the fewest number of species (A. harperi, P. angulata, and O. contorta) (Supplementary Table S3).

Spring sampling yielded the most species (n = 47, Figure 9), including 20 collected from this season only and 9 of the 10 families found in the Nearctic. Perlodidae was the most speciose family (n = 15), followed by Leuctridae (n = 9), Chloroperlidae (n = 7), and Perlidae (n = 6). Summer followed with 31 species collected (Figure 9), with approximately 2/3 represented by the families Leuctridae, Chloroperlidae, and Perlodidae (n = 7 each) plus five Perlidae. Autumn was represented by 10 species (Figure 9), including one undetermined Isoperla (I. WKUC 2) that was collected only as females in early October and five species of Leuctra.

3.4. Elevation Trends

The South Toe River at the Black Mountain Campground yielded the highest number of species (n = 31). This was the second largest drainage site in the study area, yet it was also sampled the most frequently, including four UV light collections. Another 15 sites had at least 10 species collected, but there was no pattern of more species being present at a specific elevation. Median elevations ranged from 889 m (Paragnetina immarginata) to 1747 m (L. variabilis) (Figure 10). Excluding Pteronarcyidae, with too little information to generalize, median elevations at the family-level scale ranged from 950 m (Perlidae) to ≥1600 m for Taeniopterygidae, Chloroperlidae, Nemouridae, Perlodidae, and Capniidae (Figure 10 and Figure 11).

For Capniidae, only P. angulata was collected broadly across the study area (Figure 10, median = 1703 m). Most Leuctridae species, except for L. tenuis, were found across the study area, ranging in stream sizes from small seeps to the South Toe River (Figure 10 and Figure 11). With the exception of Amphinemura nigritta (Figure 10), Nemouridae were also found across all elevations (Figure 10 and Figure 11). Although only two Taeniopterygidae, O. contorta and S. limata, were collected, both were found across a wide range of elevations (Figure 10 and Figure 11).

Within Chloroperlidae, Alloperla was mainly found at lower elevations (with the exception of A. usa), whereas Suwallia marginata and two species of Sweltsa were commonly found (Figure 10). Peltoperlidae, Tallaperla anna and T. maria, were present across all elevations (Figure 10 and Figure 11). Perlidae were mainly collected in lower-elevation streams (Figure 11). Eccoptura xanthenes was the only common perlid found at a median elevation >1000 m (Figure 10).

No pattern was observed for Perlodidae, likely because most species were represented by few collections (Table 3). Isoperla pauli was found at a median elevation of 1741 m in contrast to doubleton I. orata (910 m). Undetermined Isoperla WKUC 2 was collected only from one stream at 1711 m in early October. Even in the southern Appalachian Highland Region, females are typically found as adults no later than mid-summer [66]. Several Perlodinae were collected from low-elevation larger streams, ranging from doubleton Yugus arinus (911 m, 934 m) to Malirekus hastatus (median = 1187 m).

3.5. Rare and Regional Species

Approximately 36% (=21/58) of species collected are known only in the southern Appalachian Highlands region (Table 3). Regionally restricted, however, does not necessarily mean “rare”. Isoperla pauli, I. starki, I. stewarti, and S. sheldoni have each been described within the previous eight years and distributional knowledge for all four species is limited at best. Leuctra mitchellensis (Hanson, 1941) may only be present in a small range in western North Carolina (Table 2 and Table S2; Grubbs, S.A. (Western Kentucky University, Bowling Green, KY, USA; Unpublished work)), yet was found commonly during this study. We were mildly surprised that Zapada fumosa was not collected during this study. Although there are only nine published locality records for Z. fumosa from southwestern Virginia, east Tennessee, and western North Carolina [86], the latter harbors six of these sites from habitats close in proximity and similar in elevation and degree of protection. We did collect one (M. williamsae) of the two species listed by North Carolina (NC) as a Species of Greatest Conservation Need in their state’s Wildlife Action Plan (WAP) [99], but the second species listed (Z. chila) is presently known from only one stream in east Tennessee in the Great Smoky Mountains National Park [86]. The full extent of the taxonomic concepts of the two southeastern Zapada species were likely not known to the NC WAP authors at time of publication in 2015.

4. Conclusions

The Mount Mitchell State Park and adjacent Pisgah National Forest provides a protected landscape to address ecological and conservation questions. In a six-year period, fifty-eight species were collected, and most sampling trips added additional species records. Many species were found across elevations and are typical of cold headwater streams in the Appalachian Mountains. Prior to this study, approximately 140 species had been reported from the Blue Ridge of western North Carolina [34]. In addition to one new species description [60], three species, namely A. harperi, L. truncata, and S. onkos, are new additions to the state’s faunal list. More species, especially winter-emergent stoneflies, are likely present in the study area because there were very few sampling efforts between December and March. The integrative molecular approach added additional information on species presence, namely the number of individuals and unique sites for each species. Limitations, however, with female Leuctra identifications remain. Nearly 500 females from 2014–2019 are still unidentified and could potentially provide additional information. This paper lays a foundation for the use of integrative approaches on different taxonomic groups to describe the fauna and assess ecological patterns of a region.