Capturing arthropod diversity in complex cave systems


previous item | next item

Citation
Capturing arthropod diversity in complex cave systems

Material Information

Title:
Capturing arthropod diversity in complex cave systems
Series Title:
Diversity and Distributions
Creator:
Wynne, J. Judson
Sommer, Stefan
Howarth, Francis G.
Dickson, Brett G.
Voyles, Kyle D.
Publisher:
Wiley
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
American Southwest ( local )
Rapa Nui ( local )
Cavernicoles ( local )
Species Accumulation Curves ( local )
Genre:
serial ( sobekcm )

Notes

Abstract:
Aim Identify the optimal combination of sampling techniques to maximize the detection of diversity of cave‐dwelling arthropods. Location Central‐western New Mexico; north‐western Arizona; Rapa Nui, Chile. Methods From 26 caves across three geographically distinct areas in the Western Hemisphere, arthropods were sampled using opportunistic collecting, timed searches, and baited pitfall trapping in all caves, and direct intuitive searches and bait sampling at select caves. To elucidate the techniques or combination of techniques for maximizing sampling completeness and efficiency, we examined our sampling results using nonmetric multidimensional scaling (NMDS), analysis of similarity (ANOSIM), Wilcoxon signed‐rank tests, species richness estimators and species accumulation curves. Results To maximize the detection of cave‐dwelling arthropod species, one must apply multiple sampling techniques and specifically sample unique microhabitats. For example, by sampling cave deep zones and nutrient resource sites, we identified several undescribed cave‐adapted and/or cave‐restricted taxa in the south‐western United States and eight new species of presumed cave‐restricted arthropods on Rapa Nui that would otherwise have been missed. Sampling techniques differed in their detection of both management concern species (e.g., newly discovered cave‐adapted/restricted species, range expansions of cave‐restricted species and newly confirmed alien species) and specific taxonomic groups. Spiders were detected primarily with visual search techniques (direct intuitive searches, opportunistic collecting and timed searches), while most beetles were detected using pitfall traps. Each sampling technique uniquely identified species of management concern further strengthening the importance of a multi‐technique sampling approach. Main conclusions Multiple sampling techniques were required to best characterize cave arthropod diversity. For techniques applied uniformly across all caves, each technique uniquely detected between ~40% and 67% of
Original Version:
Diversity and Distributions, Vol. 24, no. 10 (2018-07-12).

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
This object is protected by copyright, and is made available here for research and educational purposes. Permission to reuse, publish, or reproduce the object beyond the bounds of Fair Use or other exemptions to copyright law must be obtained from the copyright holder.
Resource Identifier:
K26-05199 ( USFLDC: LOCAL DOI )
k26.5199 ( USFLDC: LOCAL Handle )

USFLDC Membership

Aggregations:
University of South Florida
Karst Information Portal

Postcard Information

Format:
serial

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

1478 | wileyonlinelibrary.com/journal/ddi Diversity and Distributions. 2018; 24 : 1478–1491. DOI: 10.1111/ddi.12772 BIODIVERSITY RESEARCHCapturing arthropod diversity in complex cave systemsJ. Judson Wynne1 |1|2| 3,4|51Department of Biological Sciences, Merriam-Powell Center for Environmental Research, Northern Arizona University, Flagstaff, Arizona2Department of Natural Sciences, Bernice P. Bishop Museum, Honolulu, Hawai’i3Conservation Science Partners, Truckee, California4Lab of Landscape Ecology and Conservation Biology, Landscape Conservation Initiative, Northern Arizona University, Flagstaff, Arizona5Saint George Field Office, Bureau of Land Management, St. George, Utah Correspondence J. Judson Wynne, Department of Biological Sciences, Merriam-Powell Center for Environmental Research, Northern Arizona University, Box 5640, Flagstaff, Arizona 86011. Email: jut.wynne@nau.edu National Park Service; American Museum of Natural History; National Speleological Society; The Explorers Club Editor: Jacqueline Beggs AbstractAim : Identify the optimal combination of sampling techniques to maximize the detec tion of diversity of cavedwelling arthropods. Location : Centralwestern New Mexico; northwestern Arizona; Rapa Nui, Chile. Methods : From 26 caves across three geographically distinct areas in the Western Hemisphere, arthropods were sampled using opportunistic collecting, timed searches, and baited pitfall trapping in all caves, and direct intuitive searches and bait sampling at select caves. To elucidate the techniques or combination of techniques for maximizing sampling completeness and efficiency, we examined our sampling re sults using nonmetric multidimensional scaling (NMDS), analysis of similarity (ANOSIM), Wilcoxon signedrank tests, species richness estimators and species ac cumulation curves. Results : To maximize the detection of cavedwelling arthropod species, one must apply multiple sampling techniques and specifically sample unique microhabitats. For example, by sampling cave deep zones and nutrient resource sites, we identified sev eral undescribed caveadapted and/or cave-restricted taxa in the southwestern United States and eight new species of presumed cave-restricted arthropods on Rapa Nui that would otherwise have been missed. Sampling techniques differed in their detection of both management concern species (e.g., newly discovered caveadapted/restricted species, range expansions of cave-restricted species and newly confirmed alien species) and specific taxonomic groups. Spiders were detected pri marily with visual search techniques (direct intuitive searches, opportunistic collect ing and timed searches), while most beetles were detected using pitfall traps. Each sampling technique uniquely identified species of management concern further strengthening the importance of a multi-technique sampling approach. Main conclusions : Multiple sampling techniques were required to best characterize cave arthropod diversity. For techniques applied uniformly across all caves, each technique uniquely detected between ~40% and 67% of the total species observed. Also, sampling cave deep zones and nutrient resource sites was critical for both in creasing the number of species detected and maximizing the likelihood of detecting management concern species.American Southwest, Rapa Nui, cavernicoles, species accumulation curves

PAGE 2

| | Caves are highly sensitive environments often supporting troglo morphic (caveadapted) invertebrate species with narrow geo graphic ranges (i.e., occurring within a single cave, watershed or geologic formation). Although some geographic regions are consid ered hot spots of endemism and subterranean biodiversity (Culver, Master, Christman, & Hobbs, 2000; Culver & Sket, 2000), cave faunas in most areas globally remain largely uncharacterized. Even relatively wellstudied regions with monitoring programs frequently yield previously undescribed taxa. For example, in North America, Culver, Christman, Sket, and Trontelj (2004) estimated more than 20 new troglomorphic species are discovered annually. While numerous studies have sought to identify the most ef ficacious methods for sampling specific taxonomic groups (Cane, Minckley, & Kervin, 2000; Cardoso, 2009; Cardoso, Crespo, Carvalho, Rufino, & Henriques, 2009; Ellison, Record, Arguello, & Gotelli, 2007) and standardize sampling procedures (Duelli, Obrist, & Schmatz, 1999; Spafford & Lortie, 2013) for surfacedwelling arthro pods, no systematic efforts have been undertaken to identify the best practices for surveying terrestrial cavernicolous (cavedwelling) arthropod communities. However, DoleOliver et al. (2009) devel oped and tested standardized sampling procedures for aquatic cavedwelling arthropods, and Hunt and Millar (2001) described several cavedwelling terrestrial invertebrate sampling methods and pro vided recommendations for their best use. To our knowledge, only one study, Weinstein and Slaney (1995), qualitatively described the efficacy of different arthropod sampling techniques for complete bi ological surveys in caves. A review of cavernicolous arthropod studies revealed an array of sampling techniques designed to capture specific taxa. Paquin and Provost (2010) suggested that “hand collecting” (i.e., visual searching for and collecting arthropods using paint brushes and/ or aspirators) is preferred for cavedwelling arthropods due to its selectivity. For wellknown or easily distinguishable taxa, research ers can use this approach as it results in fewer animals being re moved from the ecosystem. Early on, Barber (1931) and Valentine (1941) favoured baited pitfall traps for capturing omnivorous and carrion beetles due to the quick return rate, while deemphasizing visual searching due to low returns. Because not all arthropods at tracted to pitfall traps will be captured, searches around traps prior to removal have been applied by previous workers (Campbell et al., 2011; Martn & Orom, 1986; Poulson & Culver, 1969; Wynne, 2013; Wynne et al., 2014). For sampling previously unstudied caves in Hawai’i and Australia, Howarth (1980, 1988, respectively) applied a variety of techniques including intensive direct intuitive searches (in promising microhabitats), deploying bait stations (using naturally occurring organic material, tubers, rotting meat, cheese and grains) and to a lesser extent baited pitfall traps—the latter in cluded both live and kill pitfalls. Researchers have emphasized logistical difficulties in sampling terrestrial cavernicolous arthropods (Howarth, James, McDowell, Preston, & Imada, 2007; Krejca & Weckerly, 2007; Weinstein & Slaney, 1995; Zagmajster, Culver, & Sket, 2008), which present chal lenges for effectively inventorying, managing and conserving sensi tive cavernicolous arthropod communities. These include technical access (e.g., caves on cliff faces, deep canyons and rugged moun tainous terrain, as well as moving equipment and personnel through constricted cave passageways, deep fissures, pits and vertical pas sageways often while on rope), considerable equipment requirements (e.g., technical caving, safety and scientific equipment), and awareness and avoidance of sensitive archaeological and geological resources. Effectively detecting troglomorphic and obligate troglophiles (caverestricted) species presents further logistical consider ations. These animals are typically endemic to a single cave or region (Borges et al., 2012; Christman, Culver, Madden, & White, 2005; Culver et al., 2000; Deharveng et al., 2008; Harvey, Berry, Edward, & Humphreys, 2008; Harvey et al., 2011; Niemiller & Zigler, 2013; Reddell, 1981; Shear, Taylor, Wynne, & Krejca, 2009; Tian, 2011; Wynne & Shear, 2016; Wynne et al., 2014) and often represented by small populations (Mitchell, 1970; Niemiller et al., 2017). Recognizing temporal and spatial heterogeneity of cave hab itats (Chapman, 1983; Kane & Poulson, 1976; Pellegrini & Ferreira, 2013; Trontelj et al., 2012) is another important consideration. If bi ological inventories or monitoring is conducted at an inappropriate time of year, in the wrong locations, or lack the necessary sampling intensity, sensitive species populations may not be detected. For example, Krejca and Weckerly (2007) reported that despite inten sive surveys by trained cave biologists, a new caveadapted species were discovered upon the 40th visit to a Texas cave. Sket (1981) and Culver et al. (2004) reported a new genus of stygobiont after over one hundred site visits to a wellstudied Slovenia cave. Because it’s impossible for most studies to conduct 40 to 100 site visits per cave, these examples underscore the difficulties with detecting more cryptic species. Numerous human activities negatively impact cave ecosys tems and their rare and endemic inhabitants—heightening the need to conduct thorough biological inventories. These impacts may include, but are not limited to, forest conversion (Culver, 1986; Trajano, 2000), mining (Elliott, 2000), groundwater pollu tion (Aley, 1976; Graening & Brown, 2003; Notenboom, Plnet, & Turquin, 1994), water extraction and water impoundments (Lisowski, 1983; Olson, 2005; Ubick & Briggs, 2002), alien spe cies introductions (Elliott, 1992; Howarth, 1981; Howarth et al., 2007; Reeves, 1999; Taylor, Krejca, Smith, Block, & Hutto, 2003), global climate change (Badino, 2004; Chevaldonn & Lejeune, 2003), livestock grazing (Borges et al., 2012) and recreational use (Culver, 1986; Howarth & Stone, 1993; PulidoBosch, MartnRosales, LpezChicano, RodrguezNavarro, & Vallejos, 1997). This underscores the urgency for developing a methodological approach to efficiently and systematically sample cavernicolous arthropod communities. Such an approach is important for interpreting landscapescale patterns, as well as further developing the field of cave ecology and making evidence-based management decisions. The lack of a

PAGE 3

| standardized sampling protocol (specifically targeting the highest efficiency and completeness possible) for sampling cavedwelling terrestrial arthropods challenges our ability to make meaningful regional and potentially global comparisons between caves and across geographic regions. Additionally, comparable data sets across landscapes will further assist ecologists in better quantify ing cave communities, as well as furthering our understanding of the relationship between cave habitats and cavedwelling arthro pod diversity. Sampling techniques should be highly efficient in return per unit effort and should yield as complete a record of the cavedwelling arthropod community as possible. Schneider and Culver (2004) employed a systematic sampling protocol (baited pitfall traps and direct intuitive searches in cave deep zones) to evaluate completeness of these techniques using rarefaction curves. None of their curves for their study caves reached an asymptote. They suggested, due to issues of rarity, roughly half the species likely to exist were detected (Schneider & Culver, 2004). Similarly, in a land scape scale study of subterranean beetle diversity in the Dinaric karst region of Western Europe, Zagmajster et al. (2008) concluded more intensive sampling was required to detect most of the hypo gean beetles in the region. In this study, we examined differences across a standardized set of terrestrial cavernicolous arthropods sampling techniques in terms of efficiency (return per unit effort) and completeness (in detecting as many species as possible). We also investigated differences in de tectability of select taxonomic groups and species of management concern across all sampling techniques. Our aim was to provide a recommended set of standardized procedures for surveying cavedwelling arthropods.| |We studied caves in El Malpais National Monument, New Mexico and Grand CanyonParashant National Monument, Arizona, United States and Rapa Nui (Easter Island) National Park, Rapa Nui, Chile (Figure 1). El Malpais National Monument (ELMA; 34.9006N, 107.9943W) encompasses ~1,522 km2 in westcentral New Mexico, United States. The region consists of at least eight major lava flows ranging from 3,000 to 100,000 years in age (Cascadden, Kudo, & Geisman, 1997). Vegetation types include mixed conifer and pon derosa pine ( Pinus ponderosa ) forest at higher elevations gradat ing to pion–juniper woodlands and grasslands at lower elevations (Bleakly, 1997). All sites were lava tube caves ( n = 10) of various geo logical ages and ranged in elevation from 1,991 to 2,334 m above sea level (ASL). Grand CanyonParashant National Monument (PARA; 36.4015N, 113.6997W) in northwestern Arizona, United States, encompasses approximately 4,451 km2. The monument is characterized by deeply incised canyons, mesas and mountains. Study sites occurred in both juniper ( Juniperus spp.) shrublands ( n = 4 caves) and lowland Mojave Desert scrub ( n = 6). Except for one basaltic fissure cave, all caves were located within the Supai, Kaibab or Redwall limestone forma tions. Elevation of caves ranged from 736 to 1,590 m ASL. Rapa Nui National Park, located on Rapa Nui, Chile (RAPA; 27.1130S, 109.3496W), encompasses approximately 42 km2 in area. All sites were lava tube caves ( n = 6) and located within the Roiho lava flow, which is the island’s youngest lava flow (esti mated between 0.02 and 0.13 Ma years old; O’Connor, Stoffers, & Left panel, Western Hemisphere from the United States south through South America. Study areas designated a, b, c with black dot denoting general study area locations. Right panels from top to bottom: (a) Grand CanyonParashant National Monument, northwestern Arizona, United States; (b) El Malpais National Monument, westcentral New Mexico, United States; and (c) the island of Rapa Nui, Chile; For study sites, black dots represent general cave locations

PAGE 4

| McWilliams, 1995). The surrounding landscape is characterized by gently rolling hills (i.e., extinct scoria cones) with a coastal cliff es carpment flanking the western boundary. Vegetation consisted of grassland and guava ( Psidium guajava ) shrub with fig ( Ficus spp.), av ocado ( Persea americana ), eucalyptus ( Eucalyptus spp.) and banana ( Musa paradisiaca ) trees occurring within many of the cave en trances. Elevation ranged from 32 to 126 m ASL.|Study sites were sampled from 2007 through 2011. ELMA work was conducted during 06–14 October 2007 and 07–15 October 2008. PARA caves were investigated during 01–06 May 2007, 16–25 August 2007, 12–21 May 2008, 05–12 March 2009, 17–19 November 2009, 05–07 May 2010 and 01–03 September 2011. We studied caves on RAPA during 16–21 August 2008, 04–08 December 2008, 28 June–17 July 2009 and 01–07 August 2011. To avoid the risk of overcollecting possible endemic and/or cave-adapted/restricted species, all specimens were collected judi ciously. When observers were able to readily identify an individual to species, those observations were tallied on field forms, and the in dividuals were released. To further reduce the likelihood of unneces sary mortalities, we used “live capture” baited pitfall traps, and deep zone baits were placed directly on the cave floor, walls and ceiling.|We compared the relative efficacy of opportunistic collecting (OC), timed searches (TS) and baited pitfall trapping (PT), as well as a combination of systematic techniques (TS and PT, hereafter re ferred to as interval search or IS) in terms of sampling efficiency and completeness. For all caves, maps were used to determine our sampling inter val. Because we strived toward obtaining the most comprehensive representation of arthropod communities within our study caves, each site was sampled from entrance (i.e., drip line) to the back of the cave. Our sampling interval was 10% of the total cave length (e.g., for a 100mlong cave, the sampling interval was 10 m). At each sampling interval, three sampling stations (one at each wall and one at cave centreline) were established and represented an individual sampling array. Fewer than three sampling stations for a given array occurred in two cases: (a) when the cave passageway width was (b) when exposed bedrock was encountered and no materials were available to aid in creating ramp-ways into the trap, the sample sta tion was excluded. At each sampling station, we deployed live capture baited PT. Each trap consisted of two ~900ml nested plastic containers (13.5 cm high, 10.8cmdiameter rim and 8.9cmdiameter base). A teaspoon (~5 g) of peanut butter (Santa Cruz Organic Creamy at ELMA and PARA; Jif Creamy for RAPA) was placed in the bottom of the exterior container. At the bottom of each inner container, we made several dozen ~1mmdiameter holes so the bait could “breathe” to attract arthropods. Attempts were made to counter sink each pitfall trap within the cave sediment. When this was not possible, we built ramps using local materials (e.g., rocks and wooden debris) so arthropods could access the trap and fall in (Ashmole, Orom, Ashmole, & Martn, 1992). To maximize trap effectiveness and discourage small mammals, we placed small rocks around the edges of the trap and then covered the mouth of the trap with a cap rock. Pitfall traps were deployed for 3 to 4 days (3day deployment durations were driven by logistical constraints). Before removal of traps, we inspected the external cup in the event that arthropods were able to move through the holes of the inner cup. For TS, we searched for arthropods within a 1m radius around each sampling station. A 1to 3min timed search (1 min if no arthro pods were observed, three if arthropods were detected) was con ducted before PT deployment and prior to trap removal. The total number of sample stations dictated time expenditure per cave. For example, when a total of 30 sampling stations were deployed, TS ranged from 30 to 90 min per cave. OS is an approach for sampling cavedwelling arthropods as encountered (e.g., Peck, 1989; Ferreira, Martins, & Yanega, 2000; Reeves, Jensen & Ozier, 2000; Wynne & Pleytez, 2005); by design, the method was often applied in an unsystematic fashion. In this study, OC occurred as the survey team traversed the length of each cave, between sampling arrays, and while deploying and removing PT, and conducting TS. This technique was applied at least twice per cave (during PT deployment and retrieval trips) with at least two trained observers collecting and/or recording arthropods as encoun tered. For example, a cave with three observers would result in three individual OC per site visit. Because we conducted two site visits per cave, this resulted in six OC samples for each cave.|Because caves are highly heterogeneous in their distribution and occurrence of microhabitats, it is not possible to apply an interval sampling approach without either missing or ineffectively sampling areas that may support unique arthropod communities. We encoun tered moss gardens within entrances and beneath skylights of two (ELMA0008 and ELMA0012; Wynne, 2013; Wynne & Shear, 2016) and root curtains from two (ELMA0303 and ELMA0315; Wynne, 2013) ELMA caves. Fernmoss gardens occurred within entrances and beneath cave skylights of five of six lava tube caves on Rapa Nui (Wynne et al., 2014). For most caves containing moss or fernmoss gardens, we conducted direct intuitive searches (DIS) with an unlim ited search radius in which we intensively searched for arthropods by examining moss and ferns, looking on and beneath cobbles and boulders and inspecting cave sediment. For root curtains, we gently searched root masses protruding from cracks within cave ceilings of cave deep zones (Wynne, 2013). DIS were conducted for 1 hr (20 min 3 observers searching) within each of these habitats. Cave deep zones were differentially sampled. This environmen tal zone is characterized by complete darkness, stable temperature, a nearwater saturated atmosphere and low to no airflow (as in

PAGE 5

| Howarth, 1980). For select PARA and RAPA caves, we used com bined bait sampling (B) and DIS to examine deep zones of three RAPA caves (Q15074, Q15076/078 and Q15113) and two PARA caves (PARA1001 and PARA2204). Because B and DIS sites were often colocated within deep zones, we referred to these sampling techniques collectively as B/DIS. Deploying multiple bait types in cave deep zones is a common method for inventorying troglomor phic arthropods (Howarth et al., 2007; Peck & KukalovaPeck, 1981; Reeves, 2001; Reeves, Jensen, & Ozier, 2000). Baits were placed di rectly on the cave floor and within cracks and crevices on the floors, ceilings and walls and deployed for 4 to 5 days. We used different bait types to maximize attracting different taxonomic groups includ ing: (a) sweet potato ( Ipomoea batatas ), (b) fresh chicken livers for PARA and longfin bigeye fish ( Cookeolus japonicus ) entrails for RAPA and (c) small diameter branches from locally occurring plants—juni per trees ( Juniperus spp.) for PARA and hibiscus ( Hibiscus rosa-sinen sis ) and ngaoho ( Caesalpinia major ) for RAPA. Two to three stations of each bait type were deployed within the deep zone of each Rapa Nui cave, and four bait stations of each type were deployed in the two Parashant caves. Also, we conducted one DIS within each deep zone by inspecting the cave floor (each at 10 min per 1 m2). Finally, we conducted one DIS (1 hr 1 observer searching) within the deep zone of PARA3504.|All specimens were identified to the lowest taxonomic level possi ble. Across all study areas, specimens representing both accidentals and cavedwelling arthropods were included in this analysis. We at tempted to separate larval Lepidoptera specimens into operational taxonomic units for inclusion in the analyses, while specimens of lar val stage Diptera were excluded from analysis. Specieslevel identifications for many of the specimens and new species designations were provided by Drs. Rolf Aalbu (Tenebrionidae), Max Barclay (Coleoptera), Thomas Barr (Carabidae), Glen Mynhardt (Coleoptera: Ptinidae), Stewart Peck (Coleoptera: Leiodidae), Ernest Bernard (Collembola), Theodore Cohn (Orthoptera: Rhaphidophoridae), Jostein Kjaerandsen and Sarah Oliveira (Diptera: Mycetophilidae), Jon Gelhaus and Chen Young (Diptera: Tipulidae), Mark Harvey (Pseudoscorpiones: Chernetidae), Robert Johnson (Hymenoptera: Formicidae), Lynn Kimsey (Hymenoptera: Tiphiidae), Edward Mockford (Psocoptera), Barry O’Connor (Acari), Pierre Paquin (Araneae), William Shear (Myriapoda, Opiliones) and Stefano Taiti (Isopoda).|We examined differences between how both morphospecies and coarse level taxonomic groups (i.e., typically at the order level) were detected across sampling techniques for all study areas (ELMA, PARA and RAPA) using nonmetric multidimensional scaling (NMDS) and analysis of similarity (ANOSIM) with the Bray–Curtis distance index. Both tests were conducted using R’s “vegan” package. Because caves are highly heterogeneous environments (Chapman, 1983; Kane & Poulson, 1976; Pellegrini & Ferreira, 2013; Trontelj et al., 2013) and arthropod sampling was conducted randomly and systematically, we sampled areas where arthropods did not occur and/or were not detected. This resulted in a large number of samples where no arthropods were encountered. As only 19% to 38% of our samples detected arthropods, we pooled samples per method per cave. This resulted in collapsing the total number of samples from between 323 and 666 to between 24 and 30 samples across the three study areas. Our last step was to then remove any remaining “no data” samples prior to running the analysis. We then conducted our analysis by grouping (or “blocking”) each study area data set at the cave level. For NMDS, each data set was square root transformed and stan dardized using Wisconsin double standardizations. Both transforma tions improve performance (Oksanen et al., 2017). We ran NMDS at both the morphospecies and coarse taxonomic groups for each study area data set using 500 random starts. The higher the R2 value, the better the goodness of fit of the data. ANOSIM is a nonparamet ric permutation procedure used to examine dissimilarities in diver sity between groups of treatments (Clarke, 1993), which enabled us to examine diversity levels across sampling methods; the number of permutations was set at 999. For both of these analyses, we exam ined differences between OC, TS and PT. We then used Wilcoxon signedrank tests (test statistic = W ; = 0.05 for all tests) to examine finerscale differences between sampling techniques across all paired comparisons for (a) ob served richness ( Sobs), (b) sampling efficiency by effort, (c) sam pling efficiency by time expended and (d) completeness. Paired comparisons were made for each study area and for study areas combined. Because we did not apply DIS or B/DIS uniformly in all caves, we examined differences between OC, TS, PT and com bined PT and TS (or IS) only. For Sobs, we used the Sobs value for each technique (OC, TS, PT and IS) from each cave to examine all paired comparisons. This approach was repeated for efficiency by sampling effort (rate of return by dividing Sobs by the total num ber of samples per technique) and time expended (rate of return by dividing Sobs by the estimated number of minutes required to apply each technique). For pitfall trapping, we did not use the total amount of time traps were deployed; rather we summed the total time required to deploy and retrieve pitfall traps and process each traps’ contents upon retrieval. Unless otherwise noted, all analy ses were conducted using RStudio 1.1.383 (R Studio Core Team, 2017). For completeness, we used two approaches to examine how effectively each technique captured richness. First, we used ES-TIMATES 9.1 (Colwell, 2013) to calculate a suite of species rich ness estimators using the data from all sampling techniques combined for each cave. Estimators examined include Chao 2, ICE, Jackknife 1, Jackknife 2 and Bootstrap (refer to Magurran, 2004 for estimator details). Because their performance may vary across datasets (Walther & Moore, 2005), examining the results of multiple estimators is recommended (Hortal, Borges,

PAGE 6

| & Gaspar, 2006). All data sets were randomized 1,000 times. Next, we selected the most parsimonious estimators based upon typical species accumulation curve shape expressing most as ymptotic behaviour (i.e., curve increases as most common spe cies are detected, then begins to asymptote as more rare species values per technique per cave were calculated by dividing Sobs by the most parsimonious estimator value ( sensu King & Porter, 2005). Because sample sizes were small for many of the caves surveyed, we generated Mao Tau species accumulation curves at the study area level to examine relative completeness for each sampling technique and combination of all sampling techniques. We examined regional completeness by visually examining curve shape. Finally, to identify any taxonomic groupspecific patterns across techniques, we summarized detections for both individual study area and study areas combined using both absolute values and de scriptive statistics. We also wanted to ascertain whether species of management concern were detected differentially across tech niques, so we identified the extent to which each technique resulted in the detection of undescribed species (i.e., species new to science), range expansions for known species (for ELMA and PARA) and newly catalogued alien species (for RAPA).| We tallied the total number of individuals, morphospecies and coarse taxonomic groups for each study area. At ELMA, we detected 400 individuals representing 62 morphospecies within 16 orders. We detected 1,873 individuals (1,180 individuals of one cricket mor phospecies within one cave) representing 63 morphospecies in 17 orders within PARA caves. For RAPA caves, we detected 1,940 indi viduals representing 53 morphospecies within 16 orders (plus three cases where specimens were identified to “Class” level). Refer to Additional Supporting Information, Appendix S1, Tables S1.1–S1.3, for species lists of each study area. For comparisons by technique, NMDS analysis reported low stress for both morphospecies (MS) and coarse taxonomic groups (CTG; Figure 2) across all study areas: ELMA MS ( R2 = 1, stress = 0.0001) and CTG ( R2 = 0.997, stress = 0.004), PARA MS ( R2 = 0.994, stress = 0.074) and CTG ( R2 = 0.981, stress = 0.139) and RAPA MS ( R2 = 0.957, stress = 0.078) and CTG ( R2 = 0.991, stress = 0.093). Initially, NMDS for ELMA MS and RAPA MS analysis showed stress nearing zero and point clustering vertically along the left side of the y axis. This result may be indicative of the large number of zeroes (i.e., resulting from a large number of individual records where a technique was applied but few species were detected) potentially increasing the NMDS results of cavedwelling arthropod sampling data by morphospecies (left) and coarse taxonomic groups (i.e., Order level or higher; right) for (a) El Malpais National Monument, (b) Grand CanyonParashant National Monument and (c) Rapa Nui National Park. Graphs represent Bray– Curtis dissimilarity ordinations for the three sampling techniques—opportunistic collecting (OC), baited pitfall trapping (PT), and timed searches (TS). All data sets were square root transformed and standardized using Wisconsin double standardizations. To obtain the lowest stress result, we applied the recommended 500 random starts to each data set

PAGE 7

| likelihood of outliers. We successfully removed the outlier for RAPA MS to obtain an ordination. ELMA MS ordination was not possible, given the paucity of data. However, when we collapsed the MS data sets to create the CTG data sets, this greatly reduced the number of zerovalues. The resultant ELMA CTG output exhibited less ques tionable behaviour as both had low stress and moderate clustering of data points by technique. ANOSIM results (Table 1) were significant across all study areas: ELMA ( R = 0.019, p = 0.041); PARA ( R = 0.111, p = 0.001); and, RAPA ( R = 0.284, p = 0.001). To further examine the differences across sampling techniques, we examined paired comparisons using Wilcoxon signedrank tests ( W ) for Sobs, sampling effort by number of samples, sampling effort by time expended and completeness (Table 2). For Sobs, OC and IS were significantly different for ELMA ( W = 50, p = 0.024) and all regions combined ( W = 280, p = 0.008). For efficiency by effort for both number of samples collected and time expended, over half (or >59%; 19 of 32) of these paired comparisons were significantly differ ent (refer to Table 2). For completeness, we observed significant dif ferences for OC and IS for all regions combined ( W = 170, p = 0.0136). When identifying the most parsimonious richness estima tors for each cave for completeness, we found second order jack knife (Burnham & Overton, 1978, 1979) and bootstrap (Burnham & Overton, 1978; Smith & van Belle, 1984) estimators performed best for all but one cave (Additional Supporting Information, Appendix S2, Table S2.1). For six caves (four at ELMA and two in PARA), we detected between one to four morphospecies. These resultant estimators were either close to the actual number, or equalled the number, of species detected and curve shape exhibited exponential behaviour. Because this is likely a small sample size effect, we did not calculate complete ness values, and these caves were excluded from further analysis. None of the species accumulation curves for any of the study areas for single techniques or all sampling techniques combined approached an asymptote (Figure 3). In all cases, more sampling is required to detect all species theoretically expected (or to attain asymptotic behaviour). Curves for OC across all study areas and DIS for ELMA exhibited the least asymptotic behaviour. For all study areas, many morphospecies were detected by only one sampling technique. The highest rate of single technique detections was for PARA at 66.7%. ELMA was second at 59.7%, while RAPA had the lowest level of single technique detections at 39.6%. For the most speciose groups, we examined the relationship between detectability and sampling technique. Across study areas, spiders were most often detected using visual search techniques and rarely detected using PT. Of the 37 spider morphospecies, five (~14%) were detected with PT (among other techniques) and none were detected only with PT. Regarding spider abundance, of the 333 individuals detected, only six individuals (<2%) were detected using PT. For the 24 beetle morphospecies, six (24%) were detected with PT only, 11 (~46%) were detected with PT and at least one visual sampling technique, and seven (~29%) were detected by one or more of the visual sampling techniques (and not detected by PT). For Collembola (13 springtail morphospecies across all study areas), four (~31%) were detected by PT only, four (~31%) with PT and at least one visual sampling technique and five (~38%) using one or more visual sampling techniques. Of note, two RAPA collembolans, representing undescribed endemic species, were detected with only DIS within fernmoss gardens (Bernard et al., 2015). For Diptera ( n = 22, across all study areas), most morphospecies (68.2%; 15 of 22 morphospecies) were detected through a combination of visual search techniques. Four morphospecies (~18%) were detected using B/DIS including a new species of keroplatid fly at PARA, which is troglomorphic in its larval stage. Only three morphospecies (~14%) were detected with both PT and visual search techniques. Most isopod morphospecies ( n = 6) were detected on Rapa Nui, while two cave-restricted morphospecies were detected in two Parashant caves. For the six RAPA morphospecies, half were de tected by both PT and visual search techniques and half using only visual search techniques. Of these, the new Polynesian endemic species, Styloniscus manuvaka , Taiti & Wynne, 2015; was detected using TS only, while the new island endemic, Hawaiioscia rapui Taiti & Wynne, 2015; was detected using both PT and OC (Taiti & Wynne, 2015). Both cave-restricted PARA isopod morphospecies were detected using DIS within each cave’s deep zone. Regarding management concern species, nine new undescribed morphospecies were discovered, and two range expansions were documented on ELMA (Wynne, 2013). We detected three mor phospecies using DIS only—a troglomorphic dipluran, Haplocampa n. sp.? (which was first observed by Northup & Welbourn, 1997), a potentially undescribed caveadapted species of planthopper, Fulgoroidea n. sp.? (probably family Cixiidae) from the root curtains of two caves (Wynne, 2013) and a relict/caverestricted millipede, Austrotyla awishoshola , Wynne & Shear, 2016; within the moss gar dens of one cave (Wynne & Shear, 2016). For the remaining six un described species, two were detected by all techniques, one with both OC and DIS, two with TS only and one solely with PT. Range expansions for two tiphiid wasp species ( Tiphia andersoni Allen, 1971 and Tiphia nona Allen, 1965) were confirmed with DIS in moss gardens (Wynne, 2013). For Parashant, 13 new undescribed species and one range ex pansion of one cave-restricted species were documented. Four ANOSIM results for morphospecies and coarse taxonomic groups (order level or higher) for El Malpais (ELMA) and Grand CanyonParashant (PARA) National Monuments, United States and Rapa Nui (RAPA) National Park, Chile Morphospecies Coarse taxonomic group ELMA 0.121 (0.045) 0.156 (0.029) PARA 0.256 (0.001) 0.353 (0.001) RAPA 0.258 (0.006) 0.235 (0.004) All analyses were conducted by blocking by cave. We sampled 10 caves at both ELMA and PARA, and six RAPA caves. R statistic is provided as first value with p values appearing in parentheses. All tests were signifi cant at p < 0.05.

PAGE 8

| undescribed troglomorphic morphospecies were detected using B/ DIS within cave deep zones. Six undescribed species were identified with both PT and visual search techniques, one morphospecies was detected using PT only, one using a combination of visual search techniques, and one with B/DIS and visual search techniques. The range expansion of the cave-restricted millipede, Pratherodesmus voylesi Shear 2009, was determined using B/DIS within cave deep zones. Nine new/undescribed morphospecies and 12 newly catalogued alien species were identified on Rapa Nui. Detections of undescribed species included two morphospecies detected with DIS within fernmoss gardens, two with PT only, two with OC only and the remaining three morphospecies using two or more visual search techniques. For the 12 newly catalogued alien species, four were identified only by TS. The remaining eight were detected by more than one sam pling technique.| Developing a standardized sampling protocol for terrestrial cavedwelling arthropods is important for interpreting landscapescale patterns, furthering our understanding of cave ecology and mak ing evidence-based conservation and management decisions. Without the development of a standardized sampling protocol (specifically targeting the highest efficiency and completeness possible) for terrestrial cave-dwelling arthropods, it also becomes more challenging to make meaningful regional and global compari sons between caves. This study represents the first attempt to quantitatively ex amine the efficacy of various techniques for sampling terrestrial cavedwelling arthropods and examine the utility of various sam pling techniques within a conservation and management context. While we do not claim these techniques represent “the” protocol for sampling cavedwelling arthropod diversity, we are hopeful our findings will be useful in establishing a foundational framework for addressing both terrestrial cavedwelling arthropod community research questions, as well as improving biological inventory sampling designs. We observed that different sampling techniques were better suited for targeting different taxonomic groups and species of man agement concern, which underscored the importance of applying multiple sampling techniques to best capture cavedwelling arthropod diversity. ANOSIM analysis of morphospecies and coarse tax onomic groups indicated dissimilarities across sampling techniques, and we observed differences when examining Wilcoxon signedrank tests for Sobs, sampling efficiency by effort, sampling efficiency by time expended and sampling completeness. Because the sampled obs) El Malpais 19 (0.43) 37.5 (0.084) 32 (0.28) Parashant 25 (0.36) 21 (0.723) 15.5 (0.778) 41 (0.184) Rapa Nui 11.5 (0.916) 6.5 (0.462) 4.5 (0.248) 13 (0.688) All regions 154 (182) 154.5 (0.625) 116 (0.512) El Malpais 9.5 (0.138) 6 (0.058) 10 (0.084) Parashant Rapa Nui 2 (0.094) 1 (0.063) All regions El Malpais 31 (0.343) Parashant 23 (0.695) Rapa Nui 1 (0.063) 8 (0.688) 20 (0.063) 12 (0.844) All regions 121 (0.27) 211 (0.38) Completeness El Malpais 8 (0.361) 18 (0.156) 12 (0.844) 20 (0.063) Parashant 21 (0.272) 10 (1) 8 (0.353) 26 (0.313) Rapa Nui 13 (0.688) 7 (0.566) 4 (0.219) 14 (0.563) All regions 114 (0.08) 84 (0.965) 61 (0.178) Results for Sobs, efficiency by sampling effort ( Sobs divided by number of samples per technique), efficiency by time expended ( Sobs divided by number of minutes required to apply technique) and sampling completeness ( Sobs divided by estimated richness) are presented. We set = 0.05 and con sidered all tests significant at p W is provided with p values included within parentheses. Significance is presented as values in bold font.Summary of Wilcoxon signedranks tests ( W ) with p values for all paired comparisons for opportunistic collecting (OC), interval sampling (timed searches and baited pitfall trapping; IS), pitfall trapping (PT) and timed searches (TS)

PAGE 9

| cave deep zones and nutrient resource sites did not occur uniformly within all 26 caves, the resultant data were not included in the quan titative analysis. However, we still found compelling differences across all tech niques using descriptive statistics. For coarse taxonomic groups, had we relied on pitfall trapping for spiders or visual searches for beetles, we would have detected only ~14% and ~30% of their di versity, respectively. The use of B/DIS in deep zones and DIS in nu trient resource sites was shown to be of considerable importance. Without these techniques nearly one third of the cave-adapted/ restricted species would not have been detected. On Parashant, we detected four troglomorphic species by applying two addi tional sampling techniques (B/DIS) within select cave deep zones. In one cave (PARA1001), we detected two new troglomorphic species—a pseudoscorpion ( Hesperochernes bradybaughi Harvey & Wynne, 2014) and a fungus beetle ( Ptomaphagus parashant , Peck & Wynne, 2013)—both are posited to be single cave endemics (Harvey & Wynne, 2014; Peck & Wynne, 2013). We also identified poten tially two new species of a cave-restricted isopods ( Brackenridgia n. sp. 1 & 2; Stefano Taiti, in litt .; Wynne, 2010) from PARA2204 to PARA3507 caves. For the recently described cave-restricted mil lipede, Pratherodesmus voylesi , we expanded its range by ~64 km to the south by southwest (Wynne, 2010; refer to Shear et al., 2009) using B/DIS sampling at PARA2204 and PARA3507 caves. Finally, we discovered a new species of fly (likely of the family Keroplatidae), which is troglomorphic in its larval state (Wynne, 2010) from the deep zone of PARA3507. All of these animals should be considered species of management concern. By sampling nutrient resource sites, we identified several species that were not revealed using IS and OC. Moss gardens at both ELMA (Northup & Welbourn, 1997) and select Oregon caves (Benedict, 1979) were identified as relict habitats of the Last Glacial Maximum and may support once widespread species now restricted to this environment. On Rapa Nui, sampling fernmoss gardens resulted in the discovery of at least eight new species (Bernard, SotoAdames, & Wynne, 2015; Mockford & Wynne, 2013; Taiti & Wynne, 2015). These animals represent endemic relict species of a severely envi ronmentally degraded island ecosystem and thus should be consid ered critically important to conservation and management (Wynne et al., 2014). Fernmoss garden habitat supported the highest con centration of endemic species on the island (Wynne et al., 2014) and is therefore considered one of the most important terrestrial habi tats on Rapa Nui (Wynne et al., 2016). Stone, Howarth, Hoch, and Asche (2005) underscored the im portance of tree roots as significant habitats for caveadapted ani mals. Using DIS in root curtains of two ELMA caves, we discovered a potentially undescribed species of troglomorphic planthopper (Wynne, 2013). Had DIS not been applied in these areas, the poten tial importance of RAPA fernmoss gardens would not have been recognized, nor would the ELMA planthopper have been detected. |Including a visual search technique with any trapping effort greatly improves sampling completeness. Other workers (Humphreys, 1991; Martn & Orom, 1986; Peck, 1976; Poulson & Culver, 1969; Mao Tau species accumulation curves per technique for (a) El Malpais National Monument, (b) Grand CanyonParashant National Monument and (c) Rapa Nui National Park. Curves are provided for all techniques (All), opportunistic collecting (OC), timedsearch (TS), pitfall trapping (PT) and interval sampling (IS; timedsearch and pitfall trapping combined), bait sampling and direct intuitive searches combined (B/DIS) of cave deep zones, and direct intuitive searches (DIS) of moss and fernmoss gardens. 95% confidence intervals are provided for all techniques only. Confidence intervals for other curves were similar. Each point along each curve line represents an average of 1,000 randomizations

PAGE 10

| Schneider & Culver, 2004) have used various visual search tech niques around traps as part of their sampling design. Whereas our analysis qualitatively found some differences in paired comparisons between TS and PT, the most compelling result was the detection of four undescribed species (including one troglomorphic spider from ELMA and one caverestricted isopod from RAPA) and two newly catalogued alien species (on RAPA) by TS only. Because the addi tional time required per cave is relatively low and the results can be rather important from a management perspective, TS with a defined search radius should be coupled with all trapping efforts. For this study, TS (conducted within a 1 m radius of each PT) were limited from 1 to 3 min. While a slight refinement from Peck (1976), a further improvement would be to extend TS time per sampling sta tion. Campbell et al. (2011) searched for arthropods for 5 min using a much larger radius (5 m) around each pitfall trap. Because of the significance of TS in the detection of management concern species, we propose at least a 3-5 min search (using the same decision rule discussed above) prior to trap deployment and removal.|We observed no clear patterns to suggest IS was a better sampling technique than OC for observed richness, efficiency by number of samples, efficiency by amount of time expended or completeness. However, IS techniques typically detected more individuals and morphospecies than OC, but neither approach identified as many species as all techniques combined—especially when we incorpo rated the additional morphospecies detected by sampling cave deep zones and nutrient resource sites within cave entrances. These pat terns were observed for both individual caves and study areas. For most of our study caves, the number of observed morphospecies more than doubled for all techniques combined compared with the number of morphospecies detected by any single technique. The accumulation curves for all techniques combined were most asymptotic (i.e., flatter) for all techniques combined than for curves generated using data from single techniques. These findings indi cate that using multiple techniques is essential, but also that more sampling (i.e., additional site visits) is required before “all species” are detected and the theoretical asymptote is achieved. In other cavedwelling arthropod studies, nonasymptotic curve shape was partially attributed to the cryptic behaviour and/or small popula tion sizes of caveadapted organisms. Schneider and Culver (2004) reported none of their species accumulation curves neared asymp totic behaviour in their analysis of West Virginia troglomorphic ar thropods. Reporting similar asymptotic behaviour of their species accumulation curves, Gallo and Bichuette (2015) emphasized that sensitive caveadapted species may be overlooked due to limited sampling; this could result in incorrect management decisions being made based upon incomplete information. Additionally, we recommend using a multiyear study design with multiple site visits per year during the most biologically pro ductive, warmest and wettest time of year (e.g., postmonsoons in the southwestern United States, early austral summer for RAPA) as a necessary study design component. Although caves are some what buffered environments, subterranean animal communities do respond to seasonal changes (e.g., Biswas, 2010; DoleOliver et al., 2009; Ducarme, Andre, Waulty, & Lebrun, 2004; Hancock & Boulton, 2009; Howarth, 1982). For example, the desiccating effect of cold dry air on caves (i.e., the winter effect; Barr, 1968; Howarth, 1980) frequently drives caveadapted organisms into small cracks and fissures where they are more difficult to detect. Thus, sampling for troglomorphic taxa during a less optimal time of year may result in detecting fewer troglomorphic species. Several additional techniques, not evaluated in this study, may be considered to further improve the ability to better characterize diversity and detect troglomorphic species of management con cern. These techniques include the direct sampling of mesocaverns (Howarth, 1983; Ueno, 1977), DNA analysis to further refine species designations (Paquin & Hedin, 2004; Wessel et al., 2013), introduc ing irritants (e.g., smoke and diluted alcohol) into crevices to flush out animals, and building artificial mesocaverns in otherwise barren passages to draw troglomorphic arthropods out of cave pore spaces so they may be detected (Howarth, 1983; Ueno, 1977).|Through the application of standardized procedures even when sam pling for specific taxonomic groups, those workers (e.g., biologists and taxonomists) may contribute to both addressing ecological questions and providing information to better guide management decisions. Presently, no standardized sampling protocols exist for any terres trial cavedwelling taxonomic group. While coleopterists (Aalbu, 1989; Barber, 1931; Valentine, 1941) have favoured pitfall traps for the inventory of beetles, no efforts have been made to standardize this approach. For sampling millipedes, Mesibov, Taylor, and Brereton (1995) reported that 4 weeks of hand collecting during the fall was more efficient than an intensive pitfall trapping effort involving trap ping 7 days per month for over 1 year. Cardoso (2009) tested and proposed a standardized spider sampling protocol for surface spe cies. The findings of Mesibov et al. (1995) and Cardoso (2009) hold promise for further shaping standardized sampling in cave systems. As taxonspecific systematic sampling procedures are developed and tested, these techniques may facilitate more meaningful statis tical comparisons at local and landscape scales and offer additional insights towards the development of an all cavernicole sampling pro tocol. Importantly, we encourage cave biologists to record and de scribe, in as much detail as possible, the specific collection methods applied during their surveys.| Presently, a universal set of standardized methods to survey terres trial biodiversity of caves does not exist. We reported that methods useful for the inventory of one taxonomic group might not be as ef fective in detecting other taxonomic groups. Our fieldwork results

PAGE 11

| provide a measure of the relative effectiveness of a variety of sur vey techniques and show that no single technique is as effective as a combination of techniques in achieving maximum efficiency or completeness in surveying terrestrial cave arthropod communities. We hope that our work provides part of the foundation towards a more standardized survey protocol that will both optimize survey outcomes and facilitate more robust statistical comparisons across cave communities. Until additional studies comparing techniques are conducted and new sampling techniques are developed and tested, we recommend using the entire sampling design applied here—factoring in the rec ommended improvements. Each technique contributed significantly to the detection of arthropod species, and thus, all techniques should be considered important. Furthermore, always targeting cave deep zones and nutrient resource sites, as well as additional time for timed searches and multiyear sampling (during the most biologically productive/wettest and warmest times of year), are necessary re finements that will result in more complete survey outcomes. We further suggest each sampling technique interacted in a unique way with distinct arthropod behaviours, likely due to habitat requirements, movement patterns and ecology of different species. Thus, applying multiple sampling techniques resulted in maximizing the number of cavedwelling species detected. Additionally, a multi ple technique/multiyear approach will be required to adequately detect species of potential management concern (e.g., cave-adapted/ restricted species and newly arriving alien species) and develop a longterm framework to monitor changes related to human impacts (e.g., global climate change and alien species’ introductions). We have demonstrated there is no easy recipe to address the detection of cryptic taxa, and sampling cave systems is challeng ing on many levels. Our hope is that the techniques provided here, and the proposed improvements, will be further tested, scrutinized and improved for broader use in other geographic regions. While imperfect, we view this as a critical first step towards developing a comprehensive strategy for systematically sampling arthropods in terrestrial cave ecosystems. Special thanks to Kayci Cook Collins, Dave Hays, Dana Sullivan (El Malpais National Monument), Jeff Alford (Bandera Ice Caves), Jeff Bradybaugh, Rosie Pepito, Ethan McIntyre and Ray Klein (Parashant National Monument), Ninoska Cuadros Hucke, Susana Nahoe and Enrique Tucky (Rapa Nui National Park), Sergio Rapu, the mem bers of Consejo de Monumentos (Rapa Nui) and Cristian Tambley (CampoAlto Operaciones) for their assistance and logistical support of this research. Jabier Les, Sociedad de Ciencias Espeloggicas Alfonso Antxia and Andrzej Ciszewski, Polish Expedition Team provided cave maps for the Rapa Nui work. Jason Ballensky, Tama and John Cassidy, Christina Colpitts, Zach Fitzner, Greg Flores, Nicholas Glover, Michael Gowan, Luke Hanna, Lynn Hicks, Bruce Higgins, Christopher Holcomb, Alicia Ika, Jonathan Kalman, Talina Konotchick, Ara Kooser, Scott Nicolay, Knutt Petersen, Lazero Pakarati, Vicky Pakarati Hotus, Peter Polsgrove, Dan Ruby, Elizabeth Ruther, Ben Solvesky and Ty Spatta provided assistance in the field. The Illinois and San Bernardino Cave Rescue Teams remained on emergency standby status during most of this work. The Explorers Club graciously recognized most of these research trips as Explorers Club Flag Expeditions. Jeff Foster and Tom Sisk provided valuable comments leading to the improvement of this manuscript. Jacqueline Beggs and Paulo Borges peerreviewed this work. Special acknowledgments are also extended to the taxo nomic specialists listed in Method’s section. Christine Albano and Derek Sonderegger assisted with developing part of the analytical framework and provided R coding guidance. This work was sup ported by the National Park Service (Grand CanyonParashant and El Malpais National Monuments) through CPCESU cooperative agreements, The Explorers Club Exploration Fund, the American Museum of Natural History and the National Speleological Society’s International Exploration Fund. J. Judson Wynne http://orcid.org/0000-0003-0408-0629 Aalbu, R. L. (1989). An analysis of the coleoptera of Mitchell caverns, San Bernardino County, California. National Speleological Society Bulletin , 51 , 10. Aley, T. J. (1976). Hydrology and surface management. In The National cave management symposium proceedings (pp. 445). Albuquerque, New Mexico: National Cave & Karst Management Symposia. Ashmole, N. P., Orom, P., Ashmole, M. J., & Martn, J. L. (1992). Primary faunal succession in volcanic terrain: Lava and cave studies on the Canary Islands. Biological Journal of the Linnean Society , 46 , 207–234. https://doi.org/10.1111/j.1095-8312.1992.tb00861.x Badino, G. (2004). Cave temperatures and global climate change. International Journal of Speleology , 33, 103–114. https://doi. org/10.5038/1827-806X Barber, H. S. (1931). Traps for cave inhabiting insects. Journal of the Mitchell Society , 46 , 25966. Barr, T. C. Jr. (1968). Cave ecology and the evolution of troglobites. Evolutionary Biology , 2 , 35–102. Benedict, E. M. (1979). A new species of Apochthonius Chamberlin from Oregon (Pseudoscorpionida, Chthoniidae). Journal of Arachnology , 7 , 79–83. Bernard, E. C., Soto-Adames, F. N., & Wynne, J. J. (2015). Collembola of Rapa Nui (Easter Island) with descriptions of five endemic caverestricted species. Zootaxa , 3949 , 239–267. https://doi. org/10.11646/zootaxa.3949.2.6 Biswas, J. (2010). Kotumsar cave biodiversity: A review of cavernicoles and their troglobiotic traits. Biodiversity Conservation , 19 , 275–289. https://doi.org/10.1007/s10531-009-9710-7 Bleakly, D. L. (1997). Plate life on the lava—the vegetation of El Malpais. In K. Mabery (Ed.), Natural history of El Malpais national monument (pp. 113–122). New Mexico Bureau of Mines & Mineral Resources 156. Borges, P. A. V., Cardoso, P., Amorim, I. R., Pereira, F., Constncia, J. P., Nunes, J. C., Dapkevicius, M. L. (2012). Volcanic caves: Priorities for conserving the Azorean endemic troglobiont spe cies. International Journal of Speleology , 41 , 101–112. https://doi. org/10.5038/1827-806X

PAGE 12

| Burnham, K. P., & Overton, W. S. (1978). Estimation of the size of a closed population when capture probabilities vary among animals. Biometrika , 65 , 625–633. https://doi.org/10.1093/biomet/65.3.625 Burnham, K. P., & Overton, W. S. (1979). Robust estimation of popula tion size when capture probabilities vary among animals. Ecology , 60 , 927–936. https://doi.org/10.2307/1936861 Campbell, J. W., Woods, M., Ball, H. L., Pirkle, R. S., Carey, V., & Ray, C. H. (2011). Terrestrial macroinvertebrates captured with a baited ramppitfall trap from five limestone caves in North Alabama and Georgia (USA) and their association with soil organic matter. Journal of Natural History , 45 , 2645–2659. https://doi.org/10.1080/002229 33.2011.597884 Cane, J. H., Minckley, R. L., & Kervin, L. J. (2000). Sampling bees (Hymenoptera: Apiformes) for pollinator community studies: Pitfalls of pantrapping. Journal of the Kansas Entomological Society , 1 , 225–231. Cardoso, P. (2009). Standardization and optimization of arthropod inven tories—the case of Iberian spiders. Biodiversity and Conservation , 18 , 3949–3962. https://doi.org/10.1007/s10531-009-9690-7 Cardoso, P., Crespo, L. C., Carvalho, R., Rufino, A. C., & Henriques, S. S. (2009). Adhoc vs. standardized and optimized arthropod diversity sampling. Diversity , 1 , 36–51. https://doi.org/10.3390/ d1010036 Cascadden, T. E., Kudo, A. M., & Geisman, J. W. (1997). Discovering the relationships in a family of volcanoes—Cerro Candelaria, Twin Craters, Lost Woman Crater and Lave Crater. In K. Mabery (Ed.), Natural History of El Malpais National Monument. New Mexico Bureau of Mines & Mineral Resources 156 (pp. 41–52). New Mexico: New Mexico Bureau of Mines & Mineral Resources Socorro. Chapman, P. R. J. (1983). Species diversity in a tropical cave ecosys tem. Proceedings of the University of Bristol Speleological Society , 16 , 201–213. Chevaldonn, P., & Lejeune, C. (2003). Regional warminginduced spe cies shift in northwest Mediterranean marine caves. Ecology Letters , 6 , 371–379. https://doi.org/10.1046/j.1461-0248.2003.00439.x Christman, M. C., Culver, D. C., Madden, M. K., & White, D. (2005). Patterns of endemism of the eastern North American cave fauna. Journal of Biogeography , 32 , 1441–1452. https://doi. org/10.1111/j.1365-2699.2005.01263.x Clarke, K. R. (1993). Nonparametric multivariate analyses of changes in community structure. Australian Journal of Ecology , 18 , 117–143. https://doi.org/10.1111/j.1442-9993.1993.tb00438.x Colwell, R. K. (2013). EstimateS: Statistical estimation of species richness and shared species from samples. Version 9, Last Revised June 14, 2013. User’s Guide and application: http://purl.oclc.org/estimates [date accessed 30 October 2017]. Culver, D. C. (1986). Cave Faunas. In M. Soul (Ed.), Conservation biology (pp. 427–443). Sunderland, MA: Sinauer Associates Inc. Culver, D. C., Christman, M. C., Sket, B., & Trontelj, P. (2004). Sampling adequacy in an extreme environment: Species richness patterns in Slovenian caves. Biodiversity and Conservation , 13 , 1209–1229. https://doi.org/10.1023/B:BIOC.0000018153.49280.89 Culver, D. C., Master, L. L., Christman, M. C., & Hobbs, H. H. III (2000). Obligate cave fauna of the 48 contiguous United States. Conservation Biology , 14 , 38601. https://doi.org/10.1046/j.1523-1739.2000.99026.x Culver, D. C., & Sket, B. (2000). Hotspots of subterranean biodiversity in caves and wells. Journal of Cave and Karst Studies , 62 , 11–17. Deharveng, L. O., Brhier, F., Bedos, A. N., Tian, M. Y., Li, Y. B., Zhang, F., Tan, X. F. (2008). Mulun and surrounding karsts (Guangxi) host the richest cave fauna of China. Subterranean Biology , 6 , 75–79. Dole-Oliver, M.-J., Castellarnii, F., Coineau, N., Galassi, D. M. P., Martin, P., Mori, N., Gilbert, J. (2009). Towards an optimal sampling strat egy to assess groundwater biodiversity: Comparison across six European regions. Freshwater Biology , 54 , 777–796. https://doi. org/10.1111/j.1365-2427.2008.02133.x Ducarme, X., Andre, H. M., Waulty, G., & Lebrun, P. (2004). Comparison of endogenic and cave communities: Microarthropod density and mite species richness. European Journal of Soil Biology , 40 , 129–138. https://doi.org/10.1016/j.ejsobi.2004.10.003 Duelli, P., Obrist, M. K., & Schmatz, D. R. (1999). Biodiversity evalua tion in agricultural landscapes: Aboveground insects. Agriculture, Ecosystems & Environment , 74 , 33–64. https://doi.org/10.1016/ S0167-8809(99)00029-8 Elliott, W. R. (1992). Fire ants invade Texas caves. American Caves, Winter 5, 13. Elliott, W. R. (2000). Conservation of the North American cave and karst biota. In H. Wilkens, D. C. Culver, & W. F. Humphreys (Eds.), Subterranean ecosystems, ecosystems of the world, 30 (pp. 665–689). Amsterdam: Elsevier. Ellison, A. M., Record, S., Arguello, A., & Gotelli, N. J. (2007). Rapid inven tory of the ant assemblage in a temperate hardwood forest: Species composition and assessment of sampling methods. Environmental Entomology , 36 , 766–775. https://doi.org/10.1093/ee/36.4.766 Ferreira, R. L., Martins, R. P., & Yanega, D. (2000). Ecology of bat arthro pod communities in a Brazilian cave. Ecotropica , 6 , 105–116. Gallo, J. E., & Bichuette, M. E. (2015). Taxonomic distinctness and con servation of a new high biodiversity subterranean area in Brazil. Anais da Academia Brasileira de Cincias , 87 , 209–217. https://doi. org/10.1590/0001-3765201520140312 Graening, G. O., & Brown, A. V. (2003). Ecosystem dynamics and pol lution effects in an Ozark cave stream. Journal of the American Water Resources Association , 39 , 1497–1507. https://doi. org/10.1111/j.1752-1688.2003.tb04434.x Hancock, P. J., & Boulton, A. J. (2009). Sampling groundwater fauna: Efficiency of rapid assessment methods tested in bores in east ern Australia. Freshwater Biology , 54 , 902–917. https://doi. org/10.1111/j.1365-2427.2007.01878.x Harvey, M. S., Berry, O., Edward, K. L., & Humphreys, G. (2008). Molecular and morphological systematics of hypogean schizomids (Schizomida: Hubbardiidae) in semiarid Australia. Invertebrate Systematics , 22 , 167–194. https://doi.org/10.1071/IS07026 Harvey, M. S., Rix, M. G., Framenau, V. W., Hamilton, Z. R., Johnson, M. S., Teale, R. J., Humphreys, W. F. (2011). Protecting the innocent: Studying shortrange endemic taxa enhances conservation outcomes. Invertebrate Systematics , 25 , 1–10. https://doi.org/10.1071/IS11011 Harvey, M. S., & Wynne, J. J. (2014). Troglomorphic Pseudoscorpions (Arachnida: Pseudoscorpiones) of northern Arizona, with descrip tions of two new shortrange endemic species. Journal of Arachnology , 42 , 205–219. https://doi.org/10.1636/K14-34.1 Hortal, J., Borges, P. A. V., & Gaspar, C. (2006). Evaluating the per formance of species richness estimators: Sensitivity to sample grain size. Journal of Animal Ecology , 75 , 274–287. https://doi. org/10.1111/j.1365-2656.2006.01048.x Howarth, F. G. (1980). The zoogeography of specialized cave ani mals: A bioclimatic model. Evolution , 34 , 394–406. https://doi. org/10.1111/j.1558-5646.1980.tb04827.x Howarth, F. G. (1981). Community structure and niche differentiation in Hawaiian lava tubes. In D. Mueller-Dombois, K. W. Bridges, & H. L. Carson (Eds.), Island Ecosystems: Biological Organization in Selected Hawaiian Communities, US/IBP Synthesis Series , Vol. 15 (pp. 318–336). PA: Hutchinson Ross Publishing Co. Howarth, F. G. (1982). Bioclimatic and geological factors governing the evolution and distribution of Hawaiian cave insects. Entomologia Generalis , 8 , 17–26. Howarth, F. G. (1983). Ecology of cave arthropods. Annual Review of Entomology , 28 , 365–389. https://doi.org/10.1146/annurev. en.28.010183.002053 Howarth, F. G. (1988). Environmental ecology of North Queensland caves: Or why are there so many troglobites in Australia. In 17th

PAGE 13

| Biennial Australian Speleological Federation Tropical Conference, Lake Tinaroo, Far North Queensland (pp. 764). Cairns: Australian Speleological Federation. Howarth, F. G., James, S. A., McDowell, W., Preston, D. J., & Imada, C. T. (2007). Identification of roots in lava tube caves using molecular techniques: Implications for conservation of cave arthropod faunas. Journal of Insect Conservation , 11 , 251–261. https://doi.org/10.1007/ s10841-006-9040-y Howarth, F. G., & Stone, F. D. (1993). Conservation of Hawaii’s speleo logical resources. In W. R. Halliday (ed.) Proceedings of the third in ternational symposium on volcanospeleology, Bend, Oregon, 1982 (pp. 124–126). Seattle, WA: International Speleological Foundation. Humphreys, W. F. (1991). Reestablishment of pulsedriven populations in a terrestrial troglobite community. Journal of Animal Ecology , 60 , 609–623. https://doi.org/10.2307/5301 Hunt, M. R., & Millar, I. (2001). Cave invertebrate collecting guide. New Zealand Department of Conservation Technical Series 26. Pp. 28. Kane, T. C., & Poulson, T. L. (1976). Foraging by cave beetles: Spatial and temporal heterogeneity of prey. Ecology , 57 , 793–800. https://doi. org/10.2307/1936192 King, J. R., & Porter, S. D. (2005). Evaluation of sampling methods and species richness estimators for ants in upland ecosystems in Florida. Environmental Entomology , 34 , 1566–1578. https://doi. org/10.1603/0046-225X-34.6.1566 Krejca, J. K., & Weckerly, B. (2007). Detection probabilities of karst invertebrates. In W. R. Elliott (Ed.), Eighteenth National Cave and Karst management symposium, St. Louis, Missouri, USA, 8–12 October 2007 (pp. 283–289). Austin, Texas, USA: Texas Parks and Wildlife Department. Lisowski, E. A. (1983). Distribution, habitat, and behaviour of the Kentucky cave shrimp Palaemonias ganteri Hay. Journal of Crustacean Biology , 3 , 88–92. https://doi.org/10.2307/1547855 Magurran, A. E. (2004). Measuring biological diversity . Malden, MA: Blackwell Publishing. Martn, J.-L., & Orom, P. (1986). An ecological study of Cueva de los Roques lava tube (Tenerife, Canary Islands). Journal of Natural History , 20 , 375–388. https://doi.org/10.1080/00222938600770281 Mesibov, R., Taylor, R. J., & Brereton, R. N. (1995). Relative efficiency of pitfall trapping and handcollecting from plots for sampling of millipedes. Biodiversity and Conservation , 4 , 429–439. https://doi. org/10.1007/BF00058426 Mitchell, R. W. (1970). Total number and density estimates of some species of cavernicoles inhabiting Fern Cave, Texas. Annales de Splologie, 25 , 73–90. Mockford, E. L., & Wynne, J. J. (2013). Genus Cyptophania Banks (Psocodea: Lepidopsocidae): Unique features, augmented descrip tion of the generotype, and descriptions of three new species. Zootaxa , 3702 , 437–449. https://doi.org/10.11646/zootaxa.3702.5.3 Niemiller, M. L., & Zigler, K. S. (2013). Patterns of cave biodiversity and en demism in the Appalachians and Interior Plateau of Tennessee, USA. PLoS ONE , 8 , e64177. https://doi.org/10.1371/journal.pone.0064177 Niemiller, M. L., Zigler, K. S., Ober, K. A., Carter, E. T., Engel, A. S., Moni, G., Stephen, C. D. (2017). Rediscovery and conservation status of six shortrange endemic Pseudanophthalmus cave beetles (Carabidae: Trechini). Insect Conservation and Diversity , 10 , 495–501. https://doi. org/10.1111/icad.12263 Northup, D. E., & Welbourn, W. C. (1997). Life in the twilight zone—Lava tube ecology, natural history of El Malpais National Monument. New Mexico Bureau of Mines and Mineral Resources, Bulletin , 156 , 69–82. Notenboom, J., Plnet, S., & Turquin, M.-J. (1994). Groundwater contam ination and its impact on groundwater ecosystems. In J. Gilbert, D. L. Danielopol, & J. A. Stanford (Eds.), Groundwater ecology (pp. 477– 504). London: Academic Press Limited. https://doi.org/10.1016/ B978-0-08-050762-0.50025-5 O’Connor, J. M., Stoffers, P., & McWilliams, M. O. (1995). Timespace map ping of Easter Chain volcanism. Earth and Planetary Science Letters , 136 , 197–212. https://doi.org/10.1016/0012-821X(95)00176-D Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O’Hara, R. B., Wagner, H. (2017). vegan: Community Ecology Package, ver sion 2.4-4 (24 August 2017). Reference manual: https://cran.r-proj ect.org/web/packages/vegan/ [date accessed 05 November 2017]. Olson, R. (2005). The ecological effects of Lock and Dam No 6 in Mammoth Cave National Park. In D. Harmon (Ed.), People, places, and parks: Proceedings of the 2005 George Wright society conference on parks, protected areas, and cultural sites (pp. 294–299). Hancock, Michigan: The George Wright Society. Paquin, P., & Hedin, M. (2004). The power and perils of ‘molecular tax onomy’: A case study of eyeless and endangered Cicurina (Araneae: Dictynidae) from Texas caves. Molecular Ecology , 13 , 3239–3255. https://doi.org/10.1111/j.1365-294X.2004.02296.x Paquin, P., & Provost, P. (2010). Modification of the mouth aspiration for collection of cave arthropods. Speleobiology Notes , 2 , 1–3. Peck, S. B. (1976). The effect of cave entrances on the distribution of caveinhabiting terrestrial arthropods. International Journal of Speleology , 8 , 308–321. Peck, S. B., & Kukalova-Peck, J. (1981). The subterranean fauna and con servation of Mona Island (Puerto Rico). National Speleological Society Bulletin , 43 , 59–68. Peck, S. B. (1989). The cave fauna of Alabama: Part I: The terrestrial in vertebrates (excluding insects). National Speleological Society Bulletin , 40 , 39–63. Peck, S. B., & Wynne, J. J. (2013). Ptomaphagus parashant new species Peck and Wynne (Coleoptera: Leiodidae: Cholevinae: Ptomaphagini): The most troglomorphic cholevine beetle known from Western North America. The Coleopterists Bulletin , 67 , 309–317. https://doi. org/10.1649/0010-065X-67.3.309 Pellegrini, T. G., & Ferreira, R. L. (2013). Structure and interactions in a cave guanosoil continuum community. European Journal of Soil Biology , 57 , 19–26. https://doi.org/10.1016/j.ejsobi.2013.03.003 Poulson, T. L., & Culver, D. C. (1969). Diversity in terrestrial cave commu nities. Ecology , 50 , 153–158. https://doi.org/10.2307/1934678 Pulido-Bosch, A., Martn-Rosales, W., Lpez-Chicano, M., RodrguezNavarro, C. M., & Vallejos, A. (1997). Human impact in a tourist karstic cave (Aracena, Spain). Environmental Geology , 31 , 142–149. https://doi.org/10.1007/s002540050173 R Studio Core Team. (2017). RStudio: Integrated development for R . Boston, MA: RStudio, Inc., Version 1.1.383. http://www.rstudio.com [accessed 07 November 2017]. Reddell, J. R. (1981). A review of the cavernicole fauna of Mexico, Guatemala and Belize. Bulletin of the Texas Memorial Museum , 27 , 1–327. Reeves, W. K. (1999). Exotic species of North American Caves. In K. Henderson (Ed.), Proceedings of the 1999 National Cave and Karst management symposium (pp. 164–166). Chattanooga, Tennessee: National Cave & Karst Management Symposia. Reeves, W. K. (2001). Invertebrate and slime mold cavernicoles of Santee Cave, South Carolina, USA. Proceedings of the Academy of Natural Sciences of Philadelphia , 151 , 81–85. https://doi. org/10.1635/0097-3157(2001)151[0081:IASMCO]2.0.CO;2 Reeves, W. K., Jensen, J. B., & Ozier, J. C. (2000). New faunal and fungal records from caves in Georgia, USA. Journal of Cave and Karst Studies , 62 , 16979. Schneider, K., & Culver, D. (2004). Estimating subterranean species rich ness using intensive sampling and rarefaction curves in a high den sity cave region in West Virginia. Journal of Cave and Karst Studies , 66 , 39–45. Shear, W., Taylor, S., Wynne, J., & Krejca, J. (2009). Cave millipeds of the United States. VIII. New genera and species of polydesmidan millipeds from caves in the southwestern United States (Diplopoda,

PAGE 14

| Polydesmida, Polydesmidae and Macrosternodesmidae). Zootaxa , 2151 , 47–65. Sket, B. (1981). Niphargobates orophobata n.g., n.sp. (Amphipoda, Gammaridae s.l.) from cave waters in Slovenia (NW Yugoslavia). Bioloki Vestnik , 29 , 105–118. Smith, E. P., & van Belle, G. (1984). Nonparametric estimation of species richness. Biometrics , 40 , 119–129. https://doi.org/10.2307/2530750 Spafford, R. D., & Lortie, C. J. (2013). Sweeping beauty: Is grassland ar thropod community composition effectively estimated by sweep netting? Ecology and Evolution , 3 , 3347–3358. Stone, F. D., Howarth, F. G., Hoch, H., & Asche, M. (2005). Root commu nities in lava tubes. In D. C. Culver, & W. B. White (Eds.), Encyclopedia of caves (pp. 477–484). Burlington, Massachusetts: Elsevier. Taiti, S., & Wynne, J. J. (2015). The terrestrial Isopoda (Crustacea, Oniscidea) of Rapa Nui (Easter Island), with descriptions of two new species. ZooKeys , 515 , 27–49. https://doi.org/10.3897/zookeys.515.9477 Taylor, S. J., Krejca, J., Smith, J. E., Block, V. R., & Hutto, F. (2003). Investigation of the potential for red imported fire ant ( Solenopsis invicta ) impacts on rare karst invertebrates at Fort Hood, Texas: A field study. In Illinois Bexar County Karst Invertebrates Draft Recovery Plan Natural History Survey. Center for Biodiversity Technical Report , vol. 28 (pp. 1–153). Tian, M. (2011). A new subgenus and two new species of the troglo bitic genus Dongodytes Deuve from Guangxi, China (Coleoptera, Carabidae). Subterranean Biology , 8 , 57–64. https://doi.org/10.3897/ subtbiol.8.1232 Trajano, E. (2000). Cave faunas in the Atlantic tropical rain forest: Composition, ecology and conservation. Biotropica , 32 , 882–893. https://doi.org/10.1111/j.1744-7429.2000.tb00626.x Trontelj, P. A., Blejec, A., & Fier, C. (2012). Ecomorphological conver gence of cave communities. Evolution , 66 , 3852–3865. Ubick, D., & Briggs, T. S. (2002). The harvestman family Phalangodidae 4. A review of the genus Banksula (Opiliones, Laniatores). The Journal of Arachnology , 30 , 435–451. https://doi.org/10.1636/0161-8202(200 2)030[0435:THFPAR]2.0.CO;2 Ueno, S.-I. (1977). The biospeleological importance of non-calcareous caves. In T. D. Ford (Ed.), Proceedings of the 7th international speleo logical congress (pp. 407–408). Sheffield, UK: British Cave Research Association. Valentine, J. M. (1941). Trapping for cave beetles. National Speleological Society Bulletin , 2 , 4. Walther, B. A., & Moore, J. L. (2005). The definitions of bias, pre cision, and accuracy, and their use in testing the performance of species richness estimators, with a literature review of es timator performance. Ecography , 28 , 81529. https://doi. org/10.1111/j.2005.0906-7590.04112.x Weinstein, P., & Slaney, D. (1995). Invertebrate faunal survey of Rope Ladder Cave, Northern Queensland: A comparative study of sam pling methods. Journal of the Australian Entomological Society , 34 , 233–236. https://doi.org/10.1111/j.1440-6055.1995.tb01329.x Wessel, A., Hoch, H., Asche, M., von Rintelen, T., Stelbrink, B., Heck, V., Howarth, F. G. (2013). Founder effects initiated rapid species ra diation in Hawaiian cave planthoppers. Proceedings of the National Academy of Sciences , 110 , 9391–9396. https://doi.org/10.1073/ pnas.1301657110 Wynne, J. J. (2010). Preliminary results of arthropod baiting and surface sampling at PARA-2204 cave, Grand Canyon-Parashant National Monument, Arizona. Unpublished report on file with the National Park Service, Grand Canyon-Parashant National Monument, St. George, Utah. Pp. 11. Wynne, J. J. (2013). Inventory, conservation and management of Lava Tube Caves at El Malpais National Monument, New Mexico. Park Science , 50 , 45–55. Wynne, J. J., Bernard, E. C., Howarth, F. G., Sommer, S., Soto-Adames, F. N., Taiti, S., Pakarati-Hotus, V. (2014). Disturbance relicts in a rap idly changing World: The Rapa Nui (Easter Island) factor. BioScience , 64 , 711–718. https://doi.org/10.1093/biosci/biu090 Wynne, J. J., Ika, F., Yancovic-Pakarati, S., Gonzales, L., Lazo-Hucke, P., Manuheuroroa, S., Scherson, R. (2016). Island-wide inventory for en demic ground-dwelling arthropods in extreme environments of Rapa Nui (p. 28). New York, NY: Unpublished report on file with The Explorers Club. Wynne, J. J., & Pleytez, W. (2005). Sensitive ecological areas and species inventory of Actun Chapat Cave, Vaca Plateau, Belize. Journal of Cave and Karst Studies , 67 , 148–157. Wynne, J. J., & Shear, W. A. (2016). A new millipede ( Austrotyla awishashola, n. sp., Diplopoda, Chordeumatida, Conotylidae) from New Mexico, USA, and the importance of cave moss gardens as refugial habitats. Zootaxa , 4084 , 285–292. https://doi.org/10.11646/zootaxa.4084.2.8 Zagmajster, M., Culver, D. C., & Sket, B. (2008). Species richness pat terns of obligate subterranean beetles (Insecta: Coleoptera) in a global biodiversity hotspot—effect of scale and sampling in tensity. Diversity and Distributions , 14 , 95–105. https://doi. org/10.1111/j.1472-4642.2007.00423.x J. Judson Wynne is a cave research scientist and assistant re search professor at Northern Arizona University. His research interests presently focus on the conservation and manage ment of cavedwelling arthropods and bats, questions related to community dynamics and habitat selection of cavedwelling species and cave detection techniques using remote sensing. Author contributions: J.J.W. collected and analysed data, wrote the manuscript and coordinated manuscript submission; J.J.W. and K.V. conceived the study and developed and refined cave ar thropod sampling methods; K.V. assisted with data collection in Arizona and New Mexico; B.G. provided guidance in data analysis and contributed to the study design; J.J.W., S.S., F.G.H., B.G. and K.V. guided manuscript development and edited manuscript drafts. Additional supporting information may be found online in the Supporting Information section at the end of the article. Wynne JJ, Sommer S, Howarth FG, Dickson BG, Voyles KD. Capturing arthropod diversity in complex cave systems. Divers Distrib . 2018;24:1478–1491. https://doi.org/10.1111/ddi.12772


printinsert_linkshareget_appmore_horiz

Download Options

close


  • info Info

    There are only PDFs associated with this resource.

  • link PDF(s)



Cite this item close

APA

Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.

MLA

Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.

CHICAGO

Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.

WIKIPEDIA

Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.