Fisheries / COMPLETED PROJECTS

Status and Life History of the Amblyopsid in Kentucky

Matthew L. Niemiller and able or imperiled (Culver et al. 2000) (Typhlichthys subterraneus). Benjamin M. Fitzpatrick, because of restricted geographic Despite large distributions in central Department of Ecology and distributions (Culver et al. 2000, Kentucky, little is known regarding life Evolutionary Biology, University 2003) and a number of threats, such history of these , particularly of as groundwater pollution and habitat the obligate cave-dwelling A. spelaea of Tennessee degradation (Elliott 2000; Danielopol and T. subterraneus. Pursuant with et al. 2003; Boulton 2005). Unfortu- Kentucky’s priority research and survey KDFWR Contact: Ryan Oster nately, the distribution and status of needs, the objectives of this study were many species is incomplete or lack- to (1) conduct baseline surveys and ing entirely, making conservation and status assessments of each amblyopsid management decisions difficult. Here species to determine their distribution Introduction we investigate the distribution, ecol- and conservation status in the state, (2) The Appalachians and Interior ogy, conservation status, and threats obtain cavefish biology information, Plateau support the highest aquatic to three cave-associated fish species such as habitat requirements, ecology, subterranean biodiversity within the in the family in Ken- and demography for each species, and continental U.S. (Culver et al. 2003). tucky (Fig 1): the (3) identify potential threats to existing However, over 95% of subterranean ( spelaea), Spring Cavefish and significant populations of each spe- species in North America are vulner- (Forbesichthys agassizii) and Southern cies and develop recommendations for

Cavefish / Dante Fenolio Figure 1: Four species of amblyopsid cavefishes occur in Kentucky: the Spring Cavefish (Forbesichthys papilliferus) (top left), Northern Cavefish (Amblyopsis spelaea) (top right), Southern Cavefish yphlichthys(T subterraneus) (bottom left), and Kentucky Cavefish (bottom right).

Annual Research Highlights 2011 9 COMPLETED PROJECTS / Fisheries

Figure 2: The distributions of amblyopsid cavefishes in Kentucky are confined to cave- and karst- bearing regions. Ecoregions are colored. status evaluations and monitoring. stream channel and thoroughly scanned stream pool, stream riffle, rimstone the streambed with the beams of our pool; terrestrial: mud bank, bank-cut, Methods headlamps. We also carefully lifted flat crevice), substrate (mud, sand, cobble, Field Surveys rocks, small cobble, and detritus under gravel, bedrock, organic debris, arti- We searched for Northern Cave- which smaller individuals might seek ficial), cover type (rock, log, crevice, fish, Southern Cavefish, and Spring refuge. Lifted rocks were returned to organic debris), and other aspects of Cavefish from May 2007 through their original positions to minimize life history (diet, behavior, community September 2011 in caves, springs, and habitat disturbance. A similar approach associates). Additionally, we excised spring-fed streams throughout the Inte- was taken in surface springs, streams, a small tissue sample from the right rior Plateau and along the Cumberland and ponds while surveying for Spring pectoral fin or caudal fin of one or more Plateau of Kentucky, including sev- Cavefish. We used large dipnets to cavefish captured at each locality (up eral historic localities. We conducted search through aquatic vegetation and to 15 at a given locality) for subsequent surveys during all months of the year, detritus where Spring Cavefish might genetic analyses. but concentrated during periods of seek refuge during the day. We also favorable conditions in subterranean searched beneath rocks, logs, and other Results and Discussion streams (i.e., shallow, clear water with potential cover objects. A tally of each Spring Cavefish Distribution little flow) or during spring when water individual found was kept, and a con- levels were higher and Spring Cavefish certed effort was made to capture, with Within Kentucky, Spring Cavefish can be found in surface habitats. Sur- small bait nets, each cavefish encoun- have been reported from at least 48 veys for cave species (i.e., Northern tered. localities in 17 counties, including at Cavefish and Southern Cavefish) were Captured fish were placed in clear least seven records from caves (Fig. 2). temporarily discontinued in 2009 and plastic bags until standard length (SL) We did not observe Spring Cavefish 2010 because of concerns of the spread was measured to the nearest mm using during any cave surveys; however, of White Nose Syndrome affecting a small metric rule or digital calipers. the species has been reported from a cave-roosting bats. Other data were gathered from each few caves in the Western Pennyroyal To locate cavefish, we donned captured fish if possible, including sex, Karst. Spring Cavefish occur in four wetsuits and slowly walked along, condition (e.g., injuries, growths, or ecoregions in Kentucky. This distri- waded through, or crawled in the cave presence of parasites), habitat (aquatic: bution extends through much of the

10 Kentucky Department of Fish & Wildlife Resources Fisheries / COMPLETED PROJECTS

southern Interior Plateau in the central Appalachians. Included in the distribu- of these studies have focused on caves part of the state, including the Western tion of the Southern Cavefish in Ken- that are known to contain relatively Highland Rim, Eastern Highland Rim, tucky are five HUC8 watersheds. We large populations. Other studies for Crawford-Mammoth Cave Uplands documented several new populations which the most reliable estimates of and Western Pennyroyal Karst from the in Pulaski County in the Upper Cum- abundance have been obtained have Mammoth Cave region in Edmonson berland watershed. Prior to this study, focused on the species of conservation and Hart counties south to the Tennes- Typhlichthys were confirmed from only concern. Additional demographic stud- see border and west to Trigg, Lyon, Sloans Valley Cave (Cooper and Beiter ies, including long-term censuses, are and Livingston counties along the 1972), which is partially inundated by needed for both surface and subterra- Cumberland River. At least six records Lake Cumberland. With the assistance nean populations. exist within the Caseyville Hills of the of the Greater Cincinnati Grotto, we Historically, Spring Cavefish has Interior River Valleys and Hills. The discovered new populations in three been considered rare to uncommon highest density of Spring Cavefish lo- nearby cave systems with unconfirmed throughout much of its range. In Ken- calities occurs in the Land Between the reports from two additional cave sys- tucky, this species has been widely Lake area in Lyon and Trigg counties, tems. All localities occur within the reported but most localities yield fewer as well as south of the Bowling Green Plateau Escarpment of the Southwest- than ten fish during a single survey area in Warren County. Included in the ern Appalachians and are isolated from (Fig. 3). To our knowledge, Spring distribution of the Spring Cavefish in other populations in both Kentucky and Cavefish have only been observed Kentucky are nine HUC8 watersheds, Tennessee. This set of populations only in excess of 25 fish at two localities: including the Cumberland, Green, has an EOO of 38.3 km2 and an AOO a ditch off of Morton Road in Todd Lower Ohio, and Lower Tennessee ba- of 80.0 km2. County and Rich Pond in Warren sins. Spring Cavefish have the largest County. Most surveys yield just a few geographic extent of all amblyopsids in Northern Cavefish Distribution fish; however, this likely is an artifact Kentucky with an extent of occurrence of habitats sampled, as many ichthyo- (EOO) of 14,786.2 km2, and an area of Within Kentucky, Northern Cave- logical surveys focus on streams and occupancy (AOO) of 720.0 km2 (based fish have been reported from at least 39 other larger bodies of water rather than on 4 x 4 km grid cells). We discovered localities, including 38 caves and one spring runs and springs. Moreover, one new locality in Todd County in the spring in five counties (Fig. 2). North- most springs are located on private Red River watershed. ern Cavefish occur in three ecoregions property and consequently have been in the state: the Crawford-Mammoth poorly sampled. Because Spring Cave- Southern Cavefish Distribution Cave Uplands and Mitchell Plain of fish return and persist in spring heads the Interior Plateau and the Caseyville and underground waters when their Within Kentucky, Southern Cave- Hills of the Interior River Valley and surface habitats dry in late summer and fish have been reported from at least Hills (two localities). The highest autumn, the best chance of detecting 29 localities, including 27 caves, one density of Southern Cavefish locali- this species occurs when water levels spring, and one well in eight coun- ties occurs in the Sinking Creek val- are high in late winter and early spring. ties (Fig. 2). The highest density of ley in Breckinridge County and in the We discovered a new, significant Southern Cavefish localities occurs in Mammoth Cave region in Edmonson population of Spring Cavefish in a Edmonson County. Southern Cavefish County. Included in the distribution in spring-fed ditch off of Morton Road in occur in four ecoregions in Kentucky. Kentucky are three HUC8 watersheds: Todd County (Fig. 4). This stream has This distribution extends through much the Rough and Upper Green watersheds been channelized for irrigation and av- of the southern Interior Plateau in the of the Green River Basin and the Blue- erages ca. 2 m wide. It is full of aquatic central part of the state, including the Sinking Watershed of the Lower Ohio vegetation, which provides ample cover Crawford-Mammoth Cave Uplands River Basin. Northern Cavefish have for the species. During our first visit on and Western Pennyroyal Karst from the an EOO of 2700.6 km2 and an AOO of 31 Mar 2010, we captured 77 fish in Mammoth Cave region in Edmonson 432.0 km2 in the state. the 30 m stretch upstream of the road and Hart counties south to the Tennes- crossing and we estimated a population see border and west to Trigg County. Relative abundance, population size density of 12,833 fish per hectare at this Southern Cavefish in this region have and trends locality. However, the number of fish an EOO of 4,547.9 km2 and an AOO dramatically decreased in subsequent of 320.0 km2. A disjunct cluster of Few studies have attempted to weeks as water levels began to decrease populations occurs in Plateau Escarp- quantify population sizes and relative and fish presumably moved upstream. ment ecoregion of the Southwestern abundance of amblyopsids, and most By mid-June in both 2010 and 2011,

Annual Research Highlights 2011 11 COMPLETED PROJECTS / Fisheries

we were unable to capture a single in a ca. 300 m section of the stream. system have been identified as popula- fish at this site. A similar phenomenon This population is unusual in that cave- tion centers for Amblyopsis in Ken- was observed at Rich Pond in Warren fish are found in the cave stream with tucky (Pearson and Boston 1995). Our County. We surveyed a 50 m section of considerable flow, often underneath surveys focused primarily on the north- stream upstream of the road crossing on rocks in the middle of the channel or ern population center in Breckinridge several occasions throughout the year. under undercut ledges around bends County where we observed significant The stream at Rich Pond issues from and meanders. During two surveys, numbers in several caves, including a series of small springs then flows water levels were elevated with low Under the Road Cave, which may have for a few hundred meters through an visibility and we observed few cave- experienced a population decline (Pear- agricultural field before issuing into a fish. We estimated a population density son and Boston 1995). Webster’s Cave large depression in an agricultural field. of 450 cavefish per In the spring during high water levels, hectare in this sec- the water from the stream issues into tion of stream. We this depression forming a large pond discovered a new (up to 340 acres in size). However, as significant popula- the season progresses, water levels drop tion of Typhlichthys and flow is usually reduced to a small in Pulaski County at stream that eventually goes completely Drowned Rat Cave. dry by July or August. During our sur- We searched ca. 400 veys, we observed as few as zero and m of stream pas- as many of 203 Spring Cavefish in this sage on four occa- 50 m section amidst aquatic vegetation. sions and observed We estimated a population density up 31, 17, 24, and 14 to 27,067 fish per hectare in the spring cavefish, including but dropping to 0 fish per hectare in the presumably young- autumn when fish move underground of-the-year fish. We and the stream dries. estimate a popula- Sixty-three percent of reported tion density of 258 Southern Cavefish localities yield cavefish per hectare fewer than ten fish during a single in this section of survey. Only Hawkins River in Mam- stream. moth Cave, Hidden River Cave in Hart Like Southern County, and L & N Railroad Cave in Cavefish, most Barren County have historically pro- Northern Cavefish duced 25+ fish during a single survey localities yield few (Fig. 3). Although Southern Cavefish cavefish, as ten or have been found in many portions of fewer cavefish have the Mammoth Cave system, the vast been observed from majority of cavefish observed are from 64% of localities in the Proctor Cave section of the system, Kentucky (Fig. 3). and more specifically Hawkins River. The largest popula- Pearson and Boston (1995) observed up tions exist in Breck- to 104 Typhlichthys during several sur- inridge County, veys in 1993 and 1994. We visited the including Webster’s Logsdon River section and observed 19 Cave, Penitentiary cavefish in 2010 even though water lev- Cave, Amblyopsis els were slightly elevated from recent Cave, and Under the rainfall. Pearson and Boston (1995) Road Cave where observed up to 45 cavefish during sev- over 100 individuals Figure 3: Proportion of Spring Cavefish (top), Southern eral surveys of L & N Railroad Cave in have been observed Cavefish (middle), and Northern Cavefish (bottom) 1993 and 1994. We visited the cave on during a single sur- localities in Kentucky categorized by the maximum four occasions and observed between vey. This area and number of cavefish observed during a single survey. 8, 15, 22, and 27 cavefish, respectively, the Mammoth Cave Most localities yield few individuals.

12 Kentucky Department of Fish & Wildlife Resources Fisheries / COMPLETED PROJECTS

in Breckinridge County also supports a erwise dry cave may not be representa- experienced a population decline after large population of Amblyopsis. Louis tive of the habitat that most cavefish excessive collections in the late 1800s (1999) estimated a population size of inhabit. Cavefish can disperse through (Niemiller and Poulson 2010) but there 211 ± 37 individuals in a 2530 m sec- and occupy submerged passages inac- is no evidence to suggest that current tion of stream passage at Webster’s cessible to humans but these habitats population densities are any less than Cave using mark-recapture with visual are probably neither usual for the fish those in the mid 1800s when the spe- implant elastomers. However, this es- nor optimal. These habitats likely act as cies was first discovered in the Mam- timate likely is conservative given that corridors for dispersal. Given their lon- moth Cave system. Pearson and Boston (1995) observed gevity, low metabolic rates, and forag- 162 individuals during a single survey ing efficiency, cavefish likely can move Management and estimated a population size of 456 relatively long distances but data are cavefish. We observed as many as 51 lacking to support this hypothesis. Recommendations individuals during our surveys of the Determining the actual popula- Several conservation measures first 1200 m of this passage. Based on tion sizes of amblyopsid cavefishes is have been proposed or implemented Pearson and Boston’s (1995) data, we extremely challenging because of the for populations of cave amblyopsids estimated a population density of 64 difficulty and inaccessibility of the hab- in Kentucky. Fencing or gating of cave cavefish per hectare in the surveyed itats that each species inhabits. Only a entrances have been proposed or imple- portion of Webster’s Cave, but it should fraction of the actual census population mented to reduce and control human be noted that cavefish have been ob- likely is sampled during a given survey; visitation to sensitive cave ecosystems, served all throughout the cave system, however, even estimating that fraction such as the many entrances to the including areas not subject to survey sampled is not trivial. Here we apply Mammoth Cave system in Edmonson (Chris Anderson, personal communica- an order of magnitude scaling factor for County, Thornhill Cave in Breckinridge tion). estimating population size, but recog- County and Parker Cave in Barren Although these results might be nize that actual population sizes could County. Special bat gates are needed a reflection of true abundance, the be lesser or greater than our coarse to allow entry and exit by bats but stop distribution and abundance of the estimates. We estimate a minimal popu- human entry. Bat Conservation Inter- cave-dwelling amblyopsids likely is lation size of over 12,000 individuals national and The National Speleologi- greater than currently realized. Locali- for Spring Cavefish, 3,200 individuals cal Society have been leaders in the ties for which Southern Cavefish and for Southern Cavefish, 500 individuals improvement and installation of such Northern Cavefish have been reported for new lineage of Typhlichthys in Pu- gates on an increasing number of bat represent but a fraction of total avail- laski County and 14,900 individuals for caves. At other caves, such as Wells able habitat accessible to cavefish. This Northern Cavefish in Kentucky. Cave in Pulaski County, signs have was clearly illustrated during a fertil- Trends refer to directional change been posted to help reduce illegal visi- izer pipeline break within the recharge over the short-term (within three gen- tation. Protection of cave surface and zone of Meramec Spring that resulted erations) and long-term (within 100 subsurface watersheds is probably the in the death of at least 1,000 Southern years) in population size, EOO, AOO, most important intervention for cave- Cavefish and likely many more. This or number of occurrences. There is no fish localities. Watershed protection has unfortunate kill is informative because current evidence to suggest that there included establishing preserves as well the drainage basin had no records for have been substantial changes in any as institution of best land management the species previously. The problem of these factors over the short-term or practices around sinkholes and sinking with inferring population densities long-term for amblyopsid cavefishes creeks, including reforestation. Indeed, from such fish kills is that we do not in Kentucky, although these factors a number of cave systems receive know the volume or extent of habitat should be reassessed every 5–10 years. some protection by occurring on state impacted. Most observations of South- The population of Typhlichthys at or federally owned land or are owned ern Cavefish and Northern Cavefish are Sloans Valley Cave has not been con- or leased by conversation agencies. In restricted to caves near the surface and firmed since the late 1960s, but cavers other cases, water tracing has identified there is some controversy as to whether have reported seeing white, the source of pollutants and so allowed even the best cavefish caves are sources in the same pools where Cooper and legal action that remedied the situation. or sinks (Niemiller and Poulson 2010). Beiter (1972) collected cavefish over Hidden River Cave in Hart County, Habitats where few or no cavefish are the past two decades. If this population Kentucky is one example. We suggest observed likely represent population was extirpated, a significant reduction that demographic source caves deserve sinks and not sources. Wells and short in EOO and AOO would occur for this complete protection of their water- stream segments encountered in an oth- lineage. Northern Cavefish may have sheds, such as Northern Cavefish local-

Annual Research Highlights 2011 13 COMPLETED PROJECTS / Fisheries

ities located in Sinking Creek. Only a few caves have the vast majority of all Northern Cavefish ever censused. At- tention to protecting these caves should be a top priority for the near future. Likewise, source populations of Spring Cavefish, such as Rich Pond, should be identified and protected. To this end, several management policies should be implemented in the immediate recharge basins of significant cavefish popula- tions to protect the health and integrity of source populations: (1) alter land use practices and implement runoff control measures to reduce the input of sedi- ments and runoff into cave systems, (2) reduce or eliminate the use of toxic pesticides and herbicides known to negatively impact the fragile subter- ranean ecosystem, (3) identify and protect critical input points (sinkholes and sinking streams) into cave systems, and (4) limit access to areas within cave systems that support large cavefish subpopulations.

In light of the current state of knowledge regarding amblyopsid populations in Kentucky, we offer the following recommendations for future research and conservation management:

Spring Cavefish Newly found cavefish habitat / Matthew Niemiller 1. Identify and survey springs lo- cated on private property located Figure 4: This spring-fed ditch off of Morton Road in Todd County contains a within the suspected distribution newly discovered population of Spring Cavefish (F. papilliferus). of the species to discover additonal significant populations. in the Western Pennyroyal Karst. 5. Delineate the recharge zone and 2. Work to protect the Rich Pond Although dispersal ability in am- conduct annual monitoring water population through purchase of blyopsids is generally thought to be quality at Rich Pond. the spring and surrounding area, low, major flood events, such as the implementing habitat protection event during May 2010, may be im- Southern Cavefish strategies, or by obtaining a conser- portant for long distance dispersal in vation agreement with the private this species. 1. Delineate the recharge zones of landowner. known localities of the undescribed 4. Establish a yearly census at the species in Pulaski County, particu- 3. Additional population genetic two most significant localities (Mor- larly the Coral Cave system and analyses and long-term mark- ton Road in Todd County and Rich Hail Cave system. recaptured are warranted to deter- Pond in Warren County) during mine connectivity of populations April or May to monitor population 2. Additional surveys are needed and dispersal ability of the species and demographic trends over time. to document additional sites for the

14 Kentucky Department of Fish & Wildlife Resources Fisheries / COMPLETED PROJECTS

undescribed species in Pulaski and sures to reduce development and Elliott, W. R. 2000. Conservation of the determine if the distribution extends construction activities in the area. North American cave and karst biota. to the southwest along the escarp- Pp. 665–689 in Wilkens H, Culver ment of the Cumberland Plateau in 3. Implement a public awareness DC, Humphreys WF (eds.). Subter- Wayne County. program to inform landowners and ranean Ecosystems. Amsterdam: others of the harmful impacts of Elsevier. 3. Determine the point source of dumping into sinkholes on ground- groundwater contamination at water and life it contains. Louis, M. M. 1999. Age, growth and Friendship Cave in Warren County fin erosion of the northern cavefish, and initiate a chemical cleanup of 4. Conduct in situ studies to deter- Amblyopsis spelaea, in Kentucky and the cave if possible. mine if Rainbow Trout and Banded Indiana. Master’s thesis, University Sculpin successfully prey on sub- of Louisville, Louisville, Kentucky. 4. Implement a public awareness terranean fauna, including Northe 204p. program to inform landowners and Cavefish, in subterranean habitats others of the harmful impacts of and determine their influence on Niemiller M. L., and T. L. Poulson. dumping into sinkholes on ground- subterranean faunal abundance and 2010. Studies of the Amblyopsidae: water and life it contains. behavior. past, present, and future. Pp. 169– 280 in Trajano E., Bichuette M.E., 5. Remove the dilapidated pump Kapoor B.G. (eds). The biology of house and other debris at the en- Literature Cited subterranean fishes. Science Publish- trance of L & N Railroad Cave in Boulton, A. J. 2005. Chances and ers, Enfield, New Hampshire. Barren County to improve terrestrial challenges in the conservation of and aquatic habitat in the cave. groundwaters and their dependent Pearson, W. D., and C. H. Boston. ecosystems. Aquatic Conservation: 1995. Distribution and status of the Northern Cavefish Marine and Freshwater Ecosystems northern cavefish, Amblyopsis spe- 15: 319–323. laea. Final report, Nongame and En- 1. Surveys are needed of cave sys- dangered Wildlife Program, Indiana tems that occur between the main Cooper, J. E., and D. P. Beiter. 1972. Department of Natural Resources, centers of distribution for Amblyo- The southern cavefish,Typhlichthys Indianapolis. psis spealea in parts of Grayson, subterraneus (Pisces, Amblyopsi- Hardin and Hart counties to deter- dae), in the eastern Mississippian mine if the two main population Plateau of Kentucky. Copeia 1972: Funding Source: State Wildlife Grant centers in Kentucky are continuous 879–881. (SWG) and University of Tennessee at or isolated by the Hart County Knoxville Ridge. Additionally, future genetic Culver, D. C., Christman, M. C., El- work should focus on determin- liott, W. R., Hobbs, H. H., and J. R. KDFWR Strategic Plan. Goal 1. ing with relationships of southern Reddell. 2003. The North American Strategic Objective 5. Comprehen- populations of Amblyopsis in the obligate cave fauna: Regional pat- sive Wildlife Conservation Strategy: Mammoth Cave area with those to terns. Biodiversity and Conservation Appendix 3.9; Class the north in the the Sinking Creek 12: 441–468. and Cephalaspidomorphi: Taxa spe- area of Breckinridge County. This cific research project #1. latter recommendation is currently Culver, D. C., Master, L. L., Christman, underway. M. C., and H. H. Hobbs. 2000. Obli- gate cave fauna of the 48 contiguous 2. Because the populations in Sink- United States. Conservation Biology ing Creek in Breckinridge County 14: 386–401. represent the most significant popu- lation center of the species, efforts Danielopol, D. L., Griebler, C., Guna- should be made to protect these tilaka, A., and J. Notenboom. 2003. populations through landowner Present state and future prospects for agreements, the purchase of cave groundwater ecosystems. Environ- entrances and surrounding land mental Conservation 30: 104–130. within recharge zones, and mea-

Annual Research Highlights 2011 15