Nesoressa crawfordi, Shelley, Rowland M. & Medrano, Michael F., 2006

Shelley, Rowland M. & Medrano, Michael F., 2006, Nesoressa crawfordi, n. gen., n. sp., a montane island milliped in New Mexico, USA; proposal of the new tribe Nesoressini and a preliminary cladogram of the lineage " Aniulina " (Julida: Parajulidae), Zootaxa 1285, pp. 31-50 : 40-48

Nesoressa crawfordi , new species

Figs. 1 , 5–22

Type specimens. ɗ holotype and 11ɗ and 18Ψ paratypes ( NCSM) collected by R. M. Shelley & M. F. Medrano, 24 July 2004, at the summit of Mt. Withington (9,837–9,881 ft. [2,951–2,964 m]), San Mateo Mts., ca. 22 mi (35.2 km) SW Magdalena, Socorro Co., New Mexico; coordinates are 33°52'59.1" N, 107°29'19.7" W. Three ɗ, 3Ψ paratypes deposited in FSCA; ɗ, 2Ψ paratypes deposited in CAS, MSB, NMNH, UCD, VMNH. Three ɗ, 3Ψ, and 3 juv. paratypes ( MSB, NCSM) taken at same locality on 28 July 1991 by C. S. Crawford.

Diagnosis. With the characters of the genus and tribe.

Color in life. Pleurotergites generally dark mottled brown ( Fig. 1 ), prozonites lighter imparting faint banded appearance, legs also lighter but epiproct and paraprocts darker. Defensive glands black, barely distinguishable through base color, occasional individuals lighter (perhaps more recently molted) with defensive glands distinctly visible. Epicranium uniformly dark brown anteriorly, becoming mottled at level of ocellaria and continuing caudad to margin with collum, fading in interantennal region and becoming progressively more lightly mottled to level of labrum. Antennae with 7 articles, becoming progressively darker through 6th antennomere, ultimate article pale; articles generally darker on distal halves.

Holotype. Specimen stiff and inflexible, broadly curved near midlength with head tucked tightly beneath 1st legs, impossible to straighten for length measurements without severe fragmentation. Body relatively long, slender, and parallel sided, resembling a largebodied nemasomatid more than a parajulid.

Head smooth, polished. Ocellaria forming inverted triangle with ca. 33 ocelli arranged in 8 rows: 1+2+3+4+4+5+6+8. Facial setae as follows: epicranial 1­1, clypeal 5­5, labral not detected. Antennae (Fig. 5) with 7 articles, reaching back to midlength of pleurotergite 3, becoming more setose distad, 1st article subglobose, 2­5 strongly clavate, 5 with ventrodistal lobe, 6 subparallel sided, 7 short and truncate, with 4 conical, sensory cones, no other sensory structures apparent; relative lengths of antennomeres 2>3>4>5>6>1>7.

FIGURES 5–12. Anatomical features of N. crawfordi . 5, right antenna, medial view. 6, epiproct and left paraproct, lateral view. 7, left 1st leg of male paratype, caudal view. 8–9, two midbody legs on left side, caudal views, coxae not shown. 10, penis, anterior view. 11, the same, caudal view. 12, the same, lateral view. Scale line = 0.25 mm for figs. 7–12, 0.75 mm for fig. 6, 1.0 0 mm for fig. 5.

Dorsum smooth, glabrous, and glossy, wholly without setae excepting epiproct. Collum ( Fig. 1 ) moderately enlarged, slightly overlapping epicranium, lateral margins sublinear to lightly rounded. Pleuroterga virtually identical excepting shorter ones at caudal extremity and 7th, which is slightly swollen. Ventrolateral striae conspicuous on anterior pleurotergites ( Fig. 1 ), extending roughly halfway up lateral surfaces, fading and becoming indistinct around segment 15. Defensive glands arising on pleurotergite 6, continuing to antepenultimate segment. Epiproct (Fig. 6) short and blunt, overhanging caudal margin of paraprocts, with 4 long apical and subapical setae, 2 lateral marginal setae, and one ventral seta per side. Paraprocts with margins slightly thickened, with 2 or 3 long marginal setae. Hypoproct short and semilunar, with 2 long marginal setae.

1st leg ( Figs. 1 , 7) with 5 podomeres, short and slightly incrassate, tucked tightly beneath head in situ and relatively inconspicuous, relative lengths of podomeres 4>5>2>3>1, 5th narrowing distad, claw bent abruptly ventrad, apically blunt and rounded. 2nd legs strongly reduced, consisting of swollen coxosternum on which penis and greatly reduced telopodites are situated. Subsequent legs with 7 articles, pregonopodal legs not noticeably modified; midbody legs (Figs. 8–9) becoming progressively more setose distad, tarsal claws long and gently curved, apically acute. Sternum of segment 8 not modified. Penis (Figs. 10–12) short and blunt, with shallow apical indentation, sides expanding at midlength, indenting then expanding again, with short, blunt projection arising from midwidth of anterior surface and directed anteroventrad, also with minute, terminal, podomere protrusions (telopodal remnants of 2nd legs) on each side giving rise to one seta and a conical structure resembling a sensory antennal cone.

7th pleurotergite ( Figs. 1 , 13–14 ) with ventral corners overlapping and extending ventrad to form elevated shield on caudal margin of gonopodal aperture, angling dorsad and curving around sides of opening. Anterior gonopod coxal lobes protruding through aperture, angling caudoventrad and nearly or actually contacting elevated caudal margin thus effectively or actually closing opening; posterior gonopods barely visible between/ beneath anterior gonopods and caudal shield in situ.

Gonopods in situ oriented generally along body axis and lying with ags ventral to pgs. Anterior gonopod ( Figs. 15–16 ) comprising very large coxal lobe and short, inconspicuous telopodite; sternum broad, equivalent in breadth throughout length; coxal lobe long and broad, extending well beyond level of distal extremity of telopodite, overlying and obscuring medial half of latter in anterior view, contiguous or closely proximate to opposite member basally and covering entire breadth of gonopod, medial margin angling smoothly and subcontinuously ventrolaterad, lateral margin curvilinear, apices broad, flattened, gently rounded, or slanted mediad; telopodite approximately ½ to 2/3 as long as coxal lobe, comprising two segments, basal one short and truncate, with small, rounded, distomedial lobe, distal segment longer and subovoid, with small field of apical setae. Posterior gonopods ( Figs. 17–19 ) joined by sclerotized sternum, structurally complex with three elongate projections each; sternum with prominent, subtriangular medial prolongation, indented slightly apically; coxa broad, globose laterad; coxal process broad basally and rounded laterad, angling mediad and extending ventrad, narrowly segregated from opposite member in midline; telopodite comprising long, closely appressed, and nearly identically configured prefemoral process and solenomere, parallel and subequal in length, curving slightly anteriad distally and appearing as one in lateral view, former lateral to latter; prefemoral process expanding at 2/3 length into small lateral lobe, curving broadly anterodorsad distally, expanding then narrowing into blunt tip; solenomere slightly separated from prefemoral process basally but otherwise closely appressed, lateral surface expanding at midlength and overlapping medial margin of prefemoral process, continuing distad and curving broadly anteriad apically to bifurcate tip in parallel configuration to prefemoral process; process "C" a subtriangular, spiniform projection closely appressed to caudomedial surface of solenomere, ½ to 2/3 as long as latter, apically acuminate. Prostatic groove running along anterior surface of solenomere to apical opening, visible distad in lateral view through translucent prefemoral process.

Female paratypes. Agreeing closely with males in color and somatic features. Cyphopodal aperture ( Fig. 20 ) open, not covered or overlapped by pleurotergal folds or extensions, synoperculum (terminology of Brolemann (1922)) visible in opening, overlying and completely obscuring valves. Cyphopods ( Figs. 21–22 ) with two incompletely segregated valves, divided and detached ventrally but coalesced dorsally, located dorsad to synoperculum and obscured by latter in situ, lateral valve with four setae on ventrolateral margins; synoperculum comprising fused opercula for most of lengths, diverging distad, apically with sides extending into lightly hirsute medial and lateral lobes, former slightly longer.

Var ia t io n. Aside from minor differences in the distal expansions of the pg prefemoral processes and solenomeres, and the lengths of the medial extensions of the coxal processes, the gonopods of N. crawfordi are remarkably constant with negligible variation among the populations.

As N. crawfordi is a rather stiff, inflexible parajulid and tends to lie in a natural curl, individuals fragment severely with all but the gentlest handling. Every individual must therefore be placed in a separate microvial, or it would be impossible to determine which pieces go together, and even with a single individual, it is impossible to determine the sequence of the fragments. To avoid this extreme breakage and leave primarily intact specimens for future workers, we opted for estimations of length and width measurements, which admittedly contain error. For lengths, we measured three or four sections of the curled animals with digital calipers and then added these partial figures; for widths, we gently placed specimens on our fingers and measured with the calipers adjacent to the millipeds without squeezing them. The measurements in the following table therefore contain a small amount of error, which we preferred over reducing the samples to shards and fragments. To ensure accuracy, segment counts were obtained before the measurements and potential breakage, and include the epiproct.

Length(mm) Width(mm) Segment nos. (including epiprocts)

We believe the data collected on widths and segment counts was performed in a sufficiently standardized manner to conduct meaningful statistical analyses. The results of two­sample t­tests demonstrated that females were significantly broader than males (p<0.0001); however, segment count differences between the sexes were insignificant (p=0.128).

Ecology. In boreal forests on mountain summits, N. crawfordi is found on or in moist soil and litter under rocks, fir logs, small twigs, and branches & leaves. A few type specimens that we collected on Mt. Withington were under large logs, primarily in meadows, but most were encountered in and under moist aspen leaves, fir needles, chips of bark, and other debris in a shallow ditch beside the gravel road that climbs the mountain. Labels in vials with non­typical samples provide the following habitat information: "spruce forest," "alpine meadow," "under rocks and in fir litter in deep forest of fir," "under bark of logs, Douglas fir ( Pseudotsuga menziesii var. glauca ) forest," "under bark of logs, mixed conifer forest," and "Berlese sampling of litter from aspen­spruce­fir area." We suspect that the specimens collected on the "alpine meadow" were actually on moist substrate under a log or rock and not in the open, which was our experience on Mts. Withington and Taylor, as all the individuals we saw were resting quietly. While millipeds in general tend to be poorly vagile, our observations suggest that N. crawfordi is especially so, perhaps even rather immobile; we therefore doubt that individuals will be found openly wandering, particularly in daylight, as happens with other parajulids. No habitat information is provided on the label with the individuals taken in caves (Cibola Co.), and while the exact environment and collecting site cannot be recalled today, the milliped may have come from cool, moist plant litter in the sinks (S. B. Peck, pers. comm. to RMS).

Distribution. Known from six localized, allopatric populations in Cibola and Socorro cos., one subterranean and five on high­elevation mountain summits ( Fig. 3 ); the aniulinine, Aniulis paiutus (Chamberlin, 1925) , occurs sympatrically and syntopically on the summits ( Shelley 2001). All localities lie west of the Rio Grande and east of the Continental Divide; they are also west of highway I–25, and all but Mt. Taylor are south of I–40. The montane populations occur along a generally linear, north­south axis, and the cave population is displaced about 60 mi (96 km) westward; it and the Mt. Taylor population, the two northernmost and the only ones in Cibola Co., are approximately 60 mi (96 km) apart and are segregated by some 150 mi (240 km) from the four southern ones in Socorro Co., which occur along a 125 mi (200 km) axis. The overall distribution is approximately 210 mi (336 km) long. These mountains are small inselbergs; N. crawfordi has not been encountered in extensive ranges like the Blacks and Mogollons, which lie to the south & west and west of the Continental Divide, or the southern ranges of the Rockies to the north. It seems safe to say that N. crawfordi is absent from the Sandia Mountains, in Bernalillo Co., which are readily accessible by paved road and cable car and lie east of both Albuquerque & the Rio Grande, as repeated investigations by a number of collectors have not revealed it, whereas A. paiutus has been taken there ( Shelley 2001). We also investigated the Manzano and Sierra Blanca ranges, in Torrance and Lincoln cos., which lie east of the Rio Grande opposite known populations to the west, and did not find N. crawfordi . Additionally, C. S. Crawford has sampled at comparably high elevations in the Capitan Mountains, Lincoln Co., without encountering the species, and the Sacramento Mountains, to the south in Otero Co., have been investigated by numerous arthropod biologists; therefore, we do not believe that N. crawfordi occurs east of the Rio Grande. Few additional ranges exist between the Rio Grande and Continental Divide, so we suspect that few if any additional populations exist; if so, they can only be on remote mountains that are either inaccessible by road or are located on Native American reservations, where visitation requires costly special permits if allowed at all. In addition to the types, the following material was examined:

NEW MEXICO: Cibola Co, ca. 15 mi (24 km) ENE Grants, Mt. Taylor (elevation not provided), 2ɗ, 2Ψ, 20 July 1953, C. C. Hoff (FSCA); W of La Mosca Peak, Mt. Taylor, 9,580 ft. (2,921 m), 2ɗ, 20 July 1953, C. C. Hoff, Gorham, Joseph (FSCA); nr. top of Mt. Taylor (11,150 ft., 3,345 m), ɗ, 21 July 1953, C. C. Hoff, Gorham, Joseph (FSCA); 11,000 ft (3,300 m), 6ɗ, Ψ, juv., 6–12 August 1977, S. B. Peck (VMNH); and 10,614 ft. (3,049 m), ɗ, 5Ψ, 31 July 2004, R. M. Shelley, M. F. Medrano (NCSM); and 20 mi (32 km) SW Grants (7,300–7,800 ft., 2,190–2,340 m), "caves nr. ice caves," 3ɗ, Ψ, 23 June 1979, S. B. & J. Peck (VMNH). Socorro Co., ca. 18 mi (50 km) W Socorro, S. Baldy Peak (10,400 ft., 3,120 m), Magdalena Mts., 6ɗ, 9Ψ, juv., 20 June–7 July 1979, S. B. & J. Peck (VMNH) and top of Magdalena Mts. N of Langmuir Lab., 10,000 ft. (3,000 m), 5ɗ, 2Ψ, 12 August 1990, C. S. Crawford (MSB, NCSM); and ca. 40 mi (64 km) SW San Antonio, San Mateo Peak (10,000 ft., 3,000 m), San Mateo Mts., 4ɗ, 2Ψ, 27 July 1991, C. S. Crawford (MSB, NCSM).

Etymology. We are pleased to name this endemic New Mexico species for our friend, colleague, and advisor, C. S. Crawford, Emeritus Professor of Biology at the University of New Mexico. A specialist in arid ecosystems, he is a long­time student of the state's natural history.

Remarks. The sample from "caves nr. ice caves," Cibola Co., is believed to have come from "Ice Caves Resort," a site that existed in 1979 and was located ca. 2 mi (3.2 km) SE of today's "Bandera Ice Caves." The "Resort" was located in an area called "Lava Crater," which contained collapsed lava tube caves, one being "Trucket Guano Cave," now called "Bat Cave." Other caves and sinkholes exist northeast of El Calderon Volcanic Cone, but according to S. B. Peck (pers. comm. to RMS), they were dry with no special fauna even at wet areas around melting ice. Based on a specimen of unreported sex collected in 1979, Peck (1982) reported the nemasomatid " Nemasoma " uta Chamberlin, 1912, from this location, the determination being by our colleague, W. A. Shear. We can imagine their consternation at finding a nemasomatid­like julidan in an area so remote from where this slender form was then known. Their identification as, now, Orinisobates utus , surely derived from its being the most proximate julidan with this body form, even though its southern limit is the Wasatch Mountains in northern Utah ( Enghoff 1985), some 400 mi (640 km) to the northwest.

We are puzzled by the structure on the pg telopodite that we label process "C"; its presence was confirmed by mounting and examining cleared gonopods from all angles under varying light intensities and magnifications under a compound microscope. Nothing similar exists in eastern North American parajulids, so for now we consider it autapomorphic for Nesoressini. We label it branch "C" because it is overshadowed by the longer and more prominent prefemoral process and solenomere.

According to Causey (1974), the 2nd male legs in parajulids are reduced and incorporated into the bilobed penial apparatus, a characteristic (= autapomorphy) for the family, and in Nesoressini, these legs are only "minute rudiments." There are indeed minute podomere­like protrusions from the penial apex that we think represent the telopodites of the legs; each gives rise to a seta and a conical structure resembling an antennal cone. Under light microscopy, these protrusions could seemingly be "hyaline, membranous tubes" as in the Julidae ( Enghoff 1996) , and the cones, if not also the apical setae, appear to be sensory structures, so SEM is needed for higher resolution. Longer leglike appendages arise from the apical penial margin in Bollmaniulus furcifer (Harger, 1872) (Bollmaniulini) ( Brölemann 1902, figs. 3–4), but in the Mexican species Parajulus amulensis & P. aztecus (Parajulini) , both authored by Pocock, 1903, the appendages are lateral to the penis ( Pocock 1903, figs. 9a, 10), which negates Causey's generalization for the Parajulidae as a whole. In N. crawfordi , the penis possesses a short, anteroventrally directed projection from around midlength/midwidth of the anterior surface, which is also evident in B. furcifer though located farther distad. The penes in these parajulids differ significantly from those in Aprosphylosoma darcenae Hoffman, 1961 (Aprosphylosomatidae) and apparently also the Paeromopodidae , in which the structure is situated caudal to the 2nd legs and is deeply divided with two separate projections that join basally ( Hoffman 1961, Enghoff 1996). Enghoff (1996) characterized the parajulid penis as double (with separate gonopores) and without setae; this is the case in these parajulids, although the separate gonopores coexist in a single, coalesced structure, and the penis in N. crawfordi possesses apical cones and a couple of minute setae. As in the Julidae ( Enghoff 1996) , penial characters warrant greater attention in the Parajulidae , which necessarily requires the greater magnification and resolution obtainable through SEM.

Figures

FIGURE 1. Nesoressa crawfordi. Head and pleurotergites 1 – 9 of a male paratype, setation omitted. Scale line = 1. 0 0 mm.

FIGURE 3. Distribution of the clade " Aniulina. " 1, Aniulini; 2, Gosiulini; 3, Parajulini; dots, Nesoressini. The dashed line denotes the uncertain boundary of both Aniulini and " Aniulina " through Manitoba and western & central Ontario.

FIGURES 13 – 19. Gonopodal features of male paratype of N. crawfordi. 13, segment 7, lateral view of left side. 14, the same, ventral view. 15, anterior gonopods, anterior view. 16, the same, caudal view. 17, posterior gonopods, anterior view. 18, the same, lateral view of left side, prostatic groove omitted on stem of solenomere to simplify the drawing; the labeled process " C " is that on the right telopodite. 19, right posterior gonopod, caudal view. ACL, anterior gonopod coxal lobe; C, posterior gonopod telopodal process " C "; CO, coxa; PCL, posterior gonopod coxal lobe; PFP, prefemoral process; S, solenomere; ST, sternum; T, anterior gonopod telopodite. Scale line = 0.25 mm for figs. 15 – 19, 2. 4 mm for figs. 13 – 14.

FIGURES 20 – 22. Cyphopodal features of female paratype of N. crawfordi. 20, cyphopodal aperture and cyphopods in situ, ventral view. 21, cyphopods, dorsal view. 22, the same / synoperculum, ventral view. SO, synoperculum; V, valve. Scale line = 0. 42 mm for fig. 20, 0. 50 mm for figs. 21 – 22.