Mexistenasellus floridensis sp. n., the first stenasellid isopod discovered from the Floridan aquifer (Crustacea, Isopoda, Asellota)

Mexistenasellus floridensis sp. n. is described from Hole in Wall Cave, Jackson County, Florida. The discovery of this unique isopod nearly doubles the range of the Family Stenasellidae in North America, which was previously known from Oaxaca, Mexico to southern Texas, USA. This is the first stenasellid reported swimming in the water column of a cave, a curious finding for an isopod that belongs to a group that is generally more adapted to inhabiting interstitial spaces. Mexistenasellus floridensis occurs syntopically with the subterranean asellid Caecidotea putea. Both species have been observed swimming in the water column, although neither has apparent anatomical adaptations for swimming as compared to the natatory pereopods of Remasellus parvus, an asellid that occurs in caves in Alachua, Madison and Wakulla counties,


Introduction
The northern edge of the range of stenasellid isopods in North America was previously the Balcones Fault Zone in southcentral Texas, with the eight known species distributed from approximately 30° north latitude in Texas southward along eastern Mexico to Oaxaca at about 17° (Bowman 1982).The extraordinary discovery of a stenasellid in Florida nearly doubles the Nearctic range of the family (Figure 1), which is otherwise known from across the Palaearctic (Magniez 1999).The zoogeographic implications of this discovery are immense as there now is an area stretching over a thousand kilometers between Texas and Florida where other stenasellids might be discovered, particularly given the penchant of these isopods for inhabiting the interstices of unconsolidated sediments.Needless to say, this also opens the door to the possibility of new discoveries in Florida, where the karst is replete with large water-filled cave systems.
The history of the discovery of stenasellids in the Nearctic is relatively brief.The first species described from the New World was Mexistenasellus coahuila Cole & Minckley (1972) discovered along with other crustaceans in the remarkable Cuatro Ciénegas basin of northern Mexico (Cole 1984).Magniez (1972) described M. wilkensi and M. parzefalli from Cueva del Huisache, San Luis Potosi, the latter from a single female specimen.Males of M. parzefalli were subsequently collected and described (Magniez 1973).Argano (1974) described M. magniezi from Veracruz.A second genus of Nearctic stenasellid, Etlastenasellus, was erected by Argano (1977) to receive the new species E. mixtecus from Oaxaca.Bowman (1982) described a second species in this genus, E. confinis, from Cueva del Guayabo, in Oaxaca, as well as Mexistenasellus colei from Cueva de Infiernillo, Tamaulipas, and M. nulemex from a mine below Cueva de la Boca, in Nuevo León.By 1982 eight Nearctic species of stenasellid isopods, divided among the two genera, were known.Mexistenasellus floridensis is the first new Nearctic stenasellid to be described since that time.Bowman (1982) alluded to the presence of stenasellids in the United States in reference to an unidentified collection in Texas.This was confirmed by Bowman (1992) as a range extension of Mexistenasellus coahuila in springs and artesian wells in Bexar County and a hyporheic habitat in Medina County, Texas (Figure 1).Little karst and few caves are present in the states between the Texas and Florida populations.However, evidence of the presence of groundwater habitats in the intervening area exist in records of the subterranean asellid Caecidotea paurotrigonus from a drain tile in Mississippi (Fleming 1972), a hole dug in soil in Louisiana (Lewis 2000), and the possibly conspecific Caecidotea dauphina from a drain tile outlet on Dauphin Island, Alabama (Modlin 1986).With the description of Caecidotea phreatica from saturated soil habits in coastal Virginia, Lewis and Holsinger (1985) discussed other species of Caecidotea occurring in similar habitats in the Gulf Coastal Plain physiographic province.These included C. hobbsi from crayfish burrows in Florida (Steeves 1966) and C. beattyi from drain tiles in southern Illinois (Lewis and Bowman 1981).There clearly remains a large geographic area where stenasellids might occur in the southeastern US.
Description.Eyeless, unpigmented, longest male approximately 9.0 mm, female 9.4 mm.Body slender, linear, about 4.3× as long as wide.Head about 1.4× as wide as long, rostrum and lateral incisions absent.Coxae not visible in dorsal view.Body becoming more dorsally spinose on pereonites 6-7 and pleon.Pleotelson about 1.65× as long as wide, caudomedial lobe moderately produced, broadly rounded.
Antenna 1 of 20 articles, distal 6 articles each with one esthete, then alternate articles with one esthete (8 esthetes total).Antenna 2 broken or detached in most specimens, flagellum 57-58 merous, esthetes absent.Mandibles with 4-cuspate incisors and lacinia; spine row with 5 spines on left, 4 spines on right, both with simple stout spines adjacent to incisors that resemble the cusps of the incisors and spines distad from incisors with complex plumosity; left molar with row of 15 plumose setae, right molar with row of 12 plumose setae.Mandibular palp 3-merous, with plumose setae on distal articles.Maxilla 1, inner lobe with 4 apical stout plumose spines; outer lobe with 12 dentate apical spines.Maxilliped without retinaculae.
Etymology.Named for the state of Florida, in recognition of the first stenasellid discovered in the United States east of the Mississippi.Suggested vernacular name is the Florida cave isopod.
Relationships.Magniez (1981) believed that the North American stenasellidae would prove to be multi-generic, although the paucity of collections has so far supported the partitioning of the New World taxa into only two genera (Magniez 1999;2008).Within the genus Mexistenasellus, M. floridensis seems to most closely resemble M. coahuila and M. colei.The male pleopod 2 endopodite of all three species is an elongate, sub-rectangular structure terminating in a short stylet that is presumably the sperm transfer cannula, surrounded by the rounded terminal lobes of the endopodial groove.The exopod of each is divided by a suture into two sub-equal parts, with sparse setation along the distal margin.The first maxilla of all three species also share an inner lobe with 4 setae and outer lobe with 12 apical spines.
Nearctic Mexistenasellus species can be tentatively identified by their ranges (Fig. 1).From a morphological standpoint, Mexistenasellus floridensis is separated from the most similar species (M.colei, M. coahuila) by the structures of the male second pleopod, namely the more slender, elongate endopodite and the presence of five setae along the margin of the exopod, as compared to one seta in the other species.Other unique morphological aspects of M. floridensis include the absence of retinaculae on the maxilliped and the distally bifurcate endopods of pleopods 3-5.
When Cole and Minckley (1972) described the first stenasellid discovered in the Nearctic Region they noted the morphological resemblance between Mexistenasellus coahuila in northeastern Mexico and Parastenasellus from northern Africa.Magniez (1981) concurred with Cole and Minckleys' observation, and further compared the morphology of M. coahuila with the African stenasellid genera Parastenasellus, Magniezia and Metastenasellus.The male second pleopods of all of these genera are similar, excluding Metastenasellus, which possesses a spiraled-appearing endopodite that is quite dissimilar.
The hypothesis of Magniez entails colonization of groundwater by the stenasellid ancestors prior to the separation of Africa and South America during the early Cretaceous (140-150 million years before present).This idea requires the presence of stenasellids in South America that remain to be discovered, and subsequent dispersal to North America during the Cenozoic.Magniez acknowledged the flaw in the hypothesis presented by the fact that the North and South American continents were separated during the majority of the Cenozoic.This problem is avoided by pushing back the invasion of freshwater habitats by stenasellids even further, to a time prior to the rifting of Pangea during the Jurassic (about 175 million years b.p.).The molecular genetic analysis of stenasellids by Morvan et al. (2013) supports the hypothesis that stenaselllids were present on Pangaea.Their data indicates that Mexistenasellus and Magniezia are sister groups and form a group separate from Stenasellus, with divergence possibly dating to the late Paleozoic.If one assumes that the African precursors were present prior to the breakup of Pangea, the theoretical presence of stenasellids in South America is no longer necessary.On the other hand, the molecular data do not support the argument of Bowman (1982) that supposed the invasion of groundwater in Mexico by the marine ancestors of Mexistenasellus as the shallow marine embayment regressed during the early Cenozoic.
Ecology.The Hole in Wall Cave is a water-filled cave system that is a popular dive site located in Merritt's Mill Pond, east of Mariana, Florida.The cave was mapped by Exley (1978) with the current surveyed length of approximately 8.5 km and maximum depth of 42.3 m.The isopods were collected as they swam, mid-water, at an average depth of approximately 24 m.This observation of the isopods swimming is curious since the pereopods of Mexistenasellus floridensis do not exhibit the dense rows of setae present on the legs of Remasellus parvus that appear to be natatory adaptations.Regardless of how ill-prepared the isopods were for swimming that was precisely what they appeared to be doing each time specimens were collected.The asellid Caecidotea putea Lewis (2009) was also present in the water column with M. floridensis.This isopod seems no more morphologically adapted for swimming than the stenasellid.
On 17 July 2015 and 31 July 2015, dives were conducted in Hole in the Wall Cave to conduct careful behavioral observations of the isopods as well as to collect physicochemical data from the cave and surface pond.During these dives isopods were observed crawling on the floor, walls, and ceiling of the cave, as well as swimming in the water column.The swimming behavior observed in these animals may be in response to physical disturbance.This hypothesis is based on the fact that isopod swimming behavior was observed more frequently by the second diver than the lead diver as the team swam through the cave.Both Mexistenasellus floridensis and Caecidotea putea exhibited a fluid, graceful swimming motion.This fact at least suggests that although fin kicks and bubbles exhaled from SCUBA equipment may have induced their movement, it is not an unfamiliar, artificial behavior.It is easier to visually spot the isopods as they swim in the water column, where their unpigmented bodies are set against the dark cave background, than against the white limestone walls and ceilings or the silt-covered bottom.Due to this fact divers were focused on collecting animals that were swimming in the water column, and therefore likely missed many isopods that may have been crawling on surfaces within the cave.In total, these facts may explain why animals, seemingly poorly adapted for swimming, were collected solely from the mid-water column.
This swimming behavior may be an adaptive flight response to escape predation from the crayfish Cambarus cryptodytes, the Georgia cave salamander Eurycea wallacei or the trogloxene yellow bullhead catfish Ameiurus natalis.Other members of the community, a subset of the Florida subterranean fauna discussed by Franz et al. (1994), were the stygophilic amphipod Crangonyx floridanus and other (probably undescribed) stygobitic amphipods.
During the 31 July 2015 dive, physicochemical data was taken using a Hydrolab HL4 sonde.Readings of depth, temperature, pH, conductivity, and dissolved oxygen (DO) were recorded every 20 seconds continuously during the dive.Dive bottom time (time spent swimming through cave passages and not entering and exiting the cave) was 51 minutes, and 155 separate readings were taken with depth varying between 19.57 and 31.53meters.Readings were also taken in the open water of Merritt's Mill Pond.For each parameter measured, the shallowest regions of the cave most closely approximated the pond water; however, temperature, pH, and DO progressively decreased with depth and specific conductivity increased with depth (Table 1).The lowest temperature, pH and DO, and highest conductivity were measured below 31 meters.The cold, low DO, high conductivity water entered the main cave passage from deep side tunnels at approximately 600 meters from the cave entrance.One of us (TRS) has explored these cave passages for the past six years and has noted seasonal variation in cave water temperature at depths between 19 and 30 meters.In total, these data suggest that the cold water entering the primary passage comes from deep water sources, less influenced by seasonal variation.While no isopods were observed in this colder water, stygobitic amphipods were collected.

Table 1 .
Physicochemical data correlated with depth.The 2 m reading was outside of the cave in the open water of Merritt's Mill Pond.The deepest sections of the cave correlated with the lowest temperature, pH, and DO, and highest specific conductivity.