T. F. Hlavac.
A Review of the Species of Scarites (Antilliscaris) (Coleoptera: Carabidae), with Notes on Their Morphology and Evolution.
Psyche 76(1):1-17, 1969.

This article at Hindawi Publishing: https://doi.org/10.1155/1969/92908
CEC's scan of this article: http://psyche.entclub.org/pdf/76/76-001.pdf, 2640K
This landing page: http://psyche.entclub.org/76/76-001.html

The following unprocessed text is extracted automatically from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

PSYCHE
Vol. 76 March, 1969 No. I
A REVIEW OF THE SPECIES OF SCARITES
(ANTILLISCARIS), (COLEOPTERA : CARABIDAE) WITH NOTES ON
THEIR MORPHOLOGY AND EVOLUTION
BY T. F. HLAVAC~
Museum of Comparative Zoology, Harvard University Recently, through the courtesy of Prof. R. A. Howard, Director of the Arnold Arboretum at Harvard University, the Museum of Comparative Zoology received a shipment of Coleoptera from the Luquillo Mountains of Puerto Rico, collected on "W. Peak" by R. A, McClain. Included in this material was a single specimen of the largest carabid in the Greater Antilles, representing a new species of Scarites (Antilliscaris), 35 mm in length. This behemoth is superficially different from but shares a unique set of characters with two known species of the subgenus from the mountains of Puerto Rico. The presence in the depauperate Puerto Rican montane cara- bid fauna of three species of Scarites raises questions about the ecology and evolution of these beetles. These questions have been approached through study of the comparative functional morphology of S. (Antit- liscaris) and other Scarites subgenera and through preliminary ob- servations and experiments on one adult collected during a recent trip to Puerto Rico.
Banninger (1949, p. 146) proposed the subgenus Antilliscaris for Scarites darlingtoni Banninger, S. mutchleri Banninger (both orig- 'I wish to thank Dr. P. J. Darlington, Jr., for suggesting this study, for advice and encouragement during its progress and for reading the manu- script. I am grateful to Dr. John F. Lawrence for reading the manuscript and for many helpful suggestions. Field facilities at the El Yunque Biological Station of the University of Puerto Rico were provided through the courtesy of Dr. Eric G. Matthews.
Field work on Puerto Rico was supported by NSF Grant GB 7346 to the Committee on Evolutionary Biology of the Department of Biology at Har- vard University.
Published with the aid of a grant from the Museum of Comparative Zoology.




================================================================================

2 Psyche [March
inally considered to be of uncertain systematic position, 1939, p. 148) and for S. danforthi Darlington. Banninger incorporated the new subgenus into his earlier key to the subgenera of Scarites (1937, pp. 123-128) and provided a key to separate the species. Scarites subgenus Antilliscaris Banninger I 949, p. 136 Diagnosis: head: large, 37-59% of length of elytra. Man- dibles: median carnassal regions large, basal molar area small with complex of interdigitating cusps, dorsal surface smooth. Eyes: small, height 30-50% of height of mandible. Antennae: long to very long for a Scarites, segment 4 similar to apical segments, dorsal and ventral setose areas with rough microsculpture and pair of narrow lateral glabrous regions without setae, sides flattened but not as distinctly as in segments 5-11. Segment 3 circular in cross section with apical ring of setae as well as a number on the apical 2/3 of the dorsal surface and a few setae, including some very large ones on the ventral sur- face (Fig. I 3 ) . Frontal plate grooved. Pterothorax: metasternum and metepisternum short, L/W of episternum 0.87-1.3, elytra fused together and to mesanotum, wings reduced, shorter than metanotum in mutchleri.
Lateral edge of mesotibia with one large spine. Mete- pimeron: slender, pleural suture very faint externally; epimeron-elytra interlocking mechanism absent.
Distribution : montane rain forests of Puerto Rico. Key to the species of S. (Antilliscaris) and to S. ( ?) darlingtoni Banninger
....
Mandibles with series of striations on dorsal surface (Fig. 7) .................................................................................... darlingtoni Mandibles : dorsal surface smooth (Figs. 6, 10, I I ) ................ 2 Elytral intervals I, 3, 5 and 7 with setiferous punctures. Dorsal surface of head with 15 setiferous punctures. Antennae short, ex- tending about to posterior border of prothorax, L/W of segments ........................................... 6-10 less than 1.1 (Fig. 8)
danforthi
Only elytral interval 3 with setiferous punctures. Dorsal surface of head with 6 or fewer setiferous punctures ............................ 3 Antennae extending to mesocoxae, L/W of segments 6-10 between I .3 and I .6 (Fig. 9). Slight impression at posterior end of scrobe. Lateral margin of pronotum with 5-6 pairs of setiferous punc- tures, L/W 0.80. Total length 12-18 mm ................ mutchleri Antennae extending to first abdominal sternite, L/W of segments 6-10 between 1.8 and 2.3
(Figs. 11, 12). Deep pit at posterior
end of scrobe. Lateral margin of pronotum with 3 pairs of seti-



================================================================================

Figs. 1, 2 -S. (Antilliscaris) megacephalus n. sp. Fig. 3 - S. (A.) danforthi.
Fig. 4 - S. (A.) mutchleri.




================================================================================

4 Psyche
[March
............
ferou,s punctures, L/W 0.54. Total length 30-35 mm. ......................................................................... megacephalus n. sp. Scarites (Antilliscaris) mutchleri Banninger Scarites mutchleri Banninger 1939, p. 149, replacement name for montanus Mutchler 1934, not montanus Banninger, 1932. Scarltes montanus Mutchler 1934, p. 1.
Type :
American Museum of Natural History, seen ; type locality: Puerto Rico, Luquillo Mtns., El Yunque, 1800 ft. Description (diagnostic characters only). With the characteristics of the subgenus Antilliscaris.
Forms as in Fig. 4. Head: antennae
long, extending to about middle of peduncle, apical segments inter- mediate between danforthi and megacephalus, L/W of segments 4-10 between 1.3 and 1.9 (Fig. 9). Vertex: 3 pair of setae, clypeal, supra- orbital and a pair halfway between the two and closer together than either. Lateroclypeus and antero-lateral halves of vertex with a series of irregular striations. Without a large depression behind the eye and a deep pit at the posterior end of scrobe. Prothorax: lateral grooves narrow but distinct with 5 or 7 pairs of setiferous punctures, posterior angles small. Sternum : 6-8 setif erous punctures clustered in front of the intercoxal spine, the latter with 6-8 setae. Elytra:
surface dull, anterior margin with dense band of coarse tubercles in addition to the tubercles associated with setiferous punctures. Lateral groove: dense concentration of tubercles, 7-8 setiferous punctures on third interval. Striae lightly impressed. Pterothorax: mesasternum, 2 setiferous punctures; metasternum 6-7 setiferous punctures. Ab-
domen: 1st and 2nd intersegmental sutures normal. Measurements.
Head: length 2.7-3. I mm, width 3.6-4.0, length/ width 0.75-0.77. Mandible: length 2.3-3.0. Prothorax: length 3.1- 3.5, width 4.0-4.6, length/width 0.77-0.76. Elytra: length 6.9-7.9, width 4.2-4.8, length/width I .64-I .64. Total length (including peduncle) I 5 .4-I 8.0 mm. Measurements reported: extremes in total length of the four specimens measured. Specimens examined: 11010- type and three others (M.C.Z.). A specimen, currently alive, about 12 mm in total length, has not been examined critically. Distribution: El Yunque, 1800 ft. to summit (3493 ft.). Larva: Three presumably second instar Scarites were collected on El Yunque one in a decayed log in cloud forest c. 2200 ft., and two others under rocks along a forest trail c. 3200 ft. Based on size (head capsule of preserved larve 1.9 mm wide) these larvae are probably mutchleri.
Scarites (Antilliscaris) megacephalus n. sp. Holotype: a male M.C.Z. No. 31710. Type locality: Puerto



================================================================================

Fig. 5 -Mandibles, S. (s. i.) suhfriuius. Fig. 6- Mandibles, S. (^.) mufcMeri. Fig. 7 - Mandibles, 5. ( ?) dartiwgtoni. Fig. 8 - Antenna1 apex, S. (A.) dunforthi. Fig. 9-Antenna1 apex, S. [A.) mittchleri.



================================================================================

6 Psyche [March
Rico, Luquillo Mountains, top of West Peak, 3447 ft., May or June 1968, W. C. McClain Coll. Paratype: M.C.Z.; Sex ?, El Yunque c. 3200 ft., Feb. 1969, L. H. Herman Jr., B. C. and T. F. Hlavac collectors. This fragmented specimen, missing the antenna1 flagella, maxillae, several legs and the apical abdominal segments, was collected dead under a stone along a forest trail. Description (diagnostic characters only). With the characteristics of the subgenus Antilliscaris (form as in Figs. I, 2). Head: very large as compared to the relatively squat pronotum and elytra. An- tennae very long, extend to first abdominal sternite, length/width of segments 4-10 between 1.8-2.3 (Fig. 12). Vertex: pair of clypeal and supraorbital setiferous punctures. With a deep, broad depression behind the eye and a deep puncture at the posterior end of the scrobe. Lateroclypeus and antero-lateral halves of vertex with series of ir- regular situations. Prothorax: lateral grooves narrow but distinct with 3 pairs of setiferous punctures (4 on left side of paratype). Sternum without setiferous punctures, intercoxal spine with I 2. Ely- tra: surface dull. Anterior: margin with small number of minute tubercles about 1/3 the width of those associated with punctures; I pair of setiferous punctures in third interval, striae strongly im- pressed. Mesasternum with 2 setiferous punctures, metasternum with 4. Abdomen: lateral portions of 1st and 2nd intersegmental sutures broadened and deepened to form a large fossa, sutures separating other segments slightly enlarged laterally.
Measurments of holotype and paratype: Plead: length 7.5-7.1 mm, width 9.5-7.2 length/width 0.79-0.98 ; Mandible: length 7.0-5.7 ; Prothorax: length 5.5-5.0, width 10.1-8. I, length/width 0.54-0.62 ; Elytra: length I 3.0-12.5, width 9.3-7.9, length/width I .4-I .58 ; Total length (including peduncle) : 35-30 mm.
Distribution: known only from the type series. Scarites (Antilliscaris) danf orthi Darlington Scarites danforthi Darlington 1939, p. 80. Type:
M.C.Z. No. 23,501 seen; type locality: Puerto Rico, Maricao forest, c. 3000 ft.
Description (diagnostic characters only) . With the characteristics of the subgenus Antilliscaris, form as in Fig. 3. 'Head: antennae, relatively short, extend to posterior margin of prothorax, apical seg- ments squat, L/W of segments 4-10 less than 1.1 (Fig. 8). Vertex with I 5 setiferous punctures including 2 pair of supraorbitals. Later- oclypeus and anterolateral halves of vertex without series of irregular situations. Without a large depression behind eye and deep pit at posterior end of scrobe. Prothorax: lateral groove indistinct near



================================================================================

Fig. 10 -Mandibles, S. (A.) danforthi
FIR. 11 -Mandibles, 5. (A.) megacephlus a, sp. Fig. 12 -Antenna1 apex, S. (A.) rneyucephattts n, sp. Kg. 13-Antenna1 segments 3, 4, 5, 5. <A.) megacephalus n. sp.



================================================================================

8 Psyche
[March
middle, with 5 pairs of setiferous punctures, posterior angles very small. Sternum : about 20 setiferous punctures clustered about the midline, intercoxal spine with 10-12 setiferous punctures. Elytra: surface shiny. Anterior margin : small number of minute tubercles 1/3 the width of the tubercles associated with setiferous punctures, 4 or 5 setiferous punctures in first interval, 7 or 8 in third, 6, 7, or 8 in fifth, and 5, 6, or 7 in seventh, striae lightly impressed. Ptero- thorax: mesasternum 8 setiferous punctures, metasternum 6 setiferous punctures.
Abdomen : lateral parts of I st and 2nd intersegmental sutures broadened and deepened to form a large fossa, other sutures enlarged laterally.
Measurements of holotype and paratype. Head: length 2.5-3.0 mm, width 3.1-3.5, lengthlwidth 0.81-0.86. Mandibles: 2.2-2.6. Prothorax: length 2.7-3.0, width 3.9-4.0, length/width 0.69-0.67. Elytra: length 6.7-6.7, width 4.0-4.0, lengthiwidth 1.68-1.68. Total length (including peduncle) I 4.6-1 6.0. Distribution: known only from the type series. Scarites (subgenus ?) ddingtoni Banninger Scar'ites dadngtoni Bann. 1935, p. 159.
Scarites (Antilliscaris) darlingtoni Bann. 1949, p. 137. Type: M.C.Z. No. 21,797, seen. Type locality: N. Haiti, Mt. Basil, c. 4700 ft.
Description (diagnostic characters only). Head: large, 43% of the length of the elytra. Mandibles (Fig. 7) heavily worn in type, with striate dorsal surface and relatively small molar area. Eyes: 5070 of the height of the mandible. Antennae: absent in type. Latero- clypeus and antero-lateral halves of vertex with series of irregular striations.
Vertex: pair of clypeal and supraorbital setiferous punc- tures. Without depression behind eye and large puncture at posterior end of scrobe.
Prothorax: lateral grooves distinct with 7 or 8 seti- ferous punctures.
Sternum: without setiferous punctures but with 17 on intercoxal spine. Elytra: surface polished. Anterior margin with small number of tubercles about 1/3 the width of those asso- ciated with punctures. Lateral groove rugose. 8 setiferous punctures in third interval, striae impressed. Pterothorax: reduced, elytra fused together and to mesanotum, wings shorter than the metanotum. Met- episternum and metepimeron absent in type, L/W about I. Mesa- sternum without setae and metasternum with 6. Abdomen: inter- segmental sutures normal.
Measurements of holotype: Head: length 4.0 mm, width 5.1, length/width 0.78. Mandible: length 3.0. Prothorax: length 4.0, width 5.6, length/width 0.72. Elytra: length 9.5, width 6.3, length/



================================================================================

19691 H h c - scarites
9
width I 40. Total length (including peduncle) 2 I .8. Specimen ex- amined: holotype.
Comments: known from the appendageless fragments of two speci- mens. I regard it as of uncertain subgeneric position for three reasons. The structures containing the major diagnostic characters of Antit- liscaris, antennae and metepimeron, are absent in the available ma- terial. A major similarity between Antilliscaris and darlingtoni is the reduction of the flight mechanism. Fligh tlessness and, in extenso, large-scale reorganization of the flight mechanism occur with high frequency among montane carabids (Darlington 1943), which argues against uniting darlingtoni with Antilliscaris in the absence of other evidence. Furthermore, there is a. gap between Antilliscaris and dar- Zingtoni in the dorsal surface of the mandibles (Figs. 5, 6, 7, 10, I I ) . DISCUSSION
Several
character states in Antilliscarls range from atypical to unique for a Scarites s. I. Many of these differences are found in the locomotory system or are peripheral to it. This raises questions about associated behavioral and ecological differences. In an attempt to list some of the ~ossible answers to these questions, particular configurations of AntiZliscaris are contrasted with 8. (s. s.) sub- striatus Haldemann and subterraneus Fabricius. This is not done because I feel that these 2 species are closely related to Antilliscaris, but because both fly and burrow (Hlavac, 1967) and can be used as a base line in interpreting the morphology of Antilliscaris. ANTENNAE: The antennae of Antilliscaris are unusual in three respects: in length, in the setation of the 4th segment and in the relative narrowness of the lateral glabrous areas. An increase in the size of the antennae can be documented in terms of elongation of individual segments, as expressed by length/width, and in terms of increase relative to body size. The ranges of length/width of seg- ments 6-10 is 0.9-1. I in danforthi, I .35-I .55 in substriatus, I .3-I .6 in mutchleri, and I 3-2.3 in megacephalus. The antennae reach the procoxae in substriatus, the posterior margin of the prothorax in dan- forthi, and middle of the peduncle in mutchleri and the first abdominal sternite in megacephalus. As compared to substriatus, the antennae of danforthi are longer and the individual apical segments are much wider. Assuming that there are no major differences between the rel- ative widths of the first segment in the above species, then a measure of width of the antenna1 apex can be expressed as width of an apical segment/width of segment I. Using width of the 8th segment in this ratio, the figures are 0.72 for substriatus and megacephalus, 0.93 for mutchleri and 1.1 for danforthi.




================================================================================

10 Psyche [March
To summarize, the antennae of all Antilliscaris are longer than substriatus. The apical segments are broad in danforthi, elongate in megacephalus, with mutchleri and substriatus intermediate. In substriatus, 28% of the antennae extends beyond the mandible. When burrowing the antennae are held agaist the head wall causing the apical segments to curve in a broad U. In megacephalus 50% of the antennae extends beyond the mandible. If this species burrows the antennae would be doubled over while the animal is working. Short antennae are characteristic of the many diverse taxa of burrow- ing Coleoptera which have independently evolved a pedunculate body form. The long antennae of Antilliscaris and particularly megace- phalus strongly suggests that these beetles do not burrow. Captive s,ubstriatus and subterraneus and many geophilious carabids apparently do not react visually to prey; olfactory and/or tactile reactions have been observed upon physical contact. While at rolling a burrow system, the antennae of subterraneus are held parallel to the long axis; the maximum sweep width is limited by the diameter of the burrow. On the surface, they are held approximately 45O from the long axis. If only short range olfactory and tactile stimuli are used to recognize prey, then finding it will be a function of the distance travelled and the width of the band through which the sensory equipment move. On the surface, long antennae are clearly adaptive in increasing the width of the search band, and may par- tially explain the condition in Antilliscaris. There is a gap in the setation and microsculpture of the first 4 segments as contrasted to segments 5-1 I in all subgenera except An- tilliscaris and Typhloscaris from the mountains of E. Africa. In most subgenera, segments 1-4 have few or no setae and are circular in cross section, while segments 5-1 I are flattened laterally and have a dense covering of setae and rough microsculpture except for a glabrous, asetose median stripe. Segment 4 of Antilliscaris is inter- mediate in shape between segments 3 and 5 and has the surface organization of the apical segments.
The surface of the apical segments is divided into two distinct structural and, probably, functional areas. As seen in lateral view,
a pair of setose bands with rough microsculpture extend along the dorsum and venter of the apical segments; they are united near the distal end of the I I th segment (Figs. 8, 9, 12). Between the setose
bands is an asetose median band with smooth microsculpture. Mem-
bers of the genus Scarites differ widely in the distribution of these two surface types; in S. (Antilliscaris) the median band is 25-30% of the width, but 55% in substriatus.




================================================================================

19691 Hlavac -- Scar ites I I
The elongate and/or broad apical segments and the state of the 4th segment in Antilliscaris represent an increase in surface area of antenna1 apex over substriatus. Due to differences in distribution of surface types, there is a large increase in setose areas, which can be interpreted as an intensification of the functions served by this region. EYES: The eyes of AntilZiscaris are small, 30~50% the height of the mandible, vs. 100% in substriatus, and are nearly circular except in danforthi where the posterior border is nearly straight. Smith (1964, p. I 16) stated that the eyes of flightless carabids are not reduced, except for cave inhabiting species; this is clearly not the case for montane Scarites or for the montane carabid fauna of New Guinea (P. J. Darlington Jr., personal communication). Eye re- duction in Scarites seems to have come about either from a reduction in diameter (AntiZZiscaris) or from diameter reduction and shortening the anterior posterior axis, giving the eyes of 8. (Ty$hZoscaris) and S. (T'aenioZobus) cubanae Banninger a peculiar slit-like appearance. MANDIBLES: The mandibles of scaritine carabids are structurally complex and can be divided into three functional regions, an apical and 2 medial carnassal areas and a basal molar part. Antilliscaris and Scarites s. s. represent extremes in the relative sizes of the latter 2 of these areas (Figs. 14, I 5). In an unworn specimen of sub- striatus (Fig. s) the apical carnassal is 34%) the median carnassals 26% and the molar 40% of the total length (perpendicular distance between apex and base) ; while in Antilliscaris (Figs. 6, 10, I I ) the apical carnassal ranges from 24-34% the median carnassals from Fig. 14. - Mandibles, 8. (AntiUiscaris). Fig. 15 - Mandibles, 8. (3, s.). Medial carnassal areas shaded.




================================================================================

12 Psyche [March
45-59% and the basal molar part ranges from 17-21% of the total length. In Antilliscaris then, the median caranassals are twice as large and the molar area half the relative size of the corresponding areas of substriatus.
If it can be assumed that these structural differences correlate with behavioral differences) then the condition in Anti1Ziscari.r may reflect 3 types of such differences. The mandibles of substriatus and sub- terraneus are used to loosen material from the head wall of the bur- row, in mating and, of course) in feeding. Some substrate loosened by the protibia is compressed further and compacted into a part of the stable burrow system by the vertex of the head and pronotum (Hlavac, 1967, and unpublished observations). The smaller molar area in Antilliscaris comuld indicate either that these animals burrow through a less dense) more easily compacted) substrate than Scarites s. s. or that they do not burrow at all. Secondly) the differences in mandible morphology may reflect differences in feeding behavior (see comments on body form) below) or mating. FLIGHT MECHANISM: Antilliscaris represents an extreme point in the atrophy of the flight mechanism. The fused elytra are immovably joined to the mesathorax at two points, the posterior part of the mesatergum is tucked around and closely pressed against the anterior contours of the elytral cavity and secondly) the elytral hinge mech- anism is solidly fused to the anterior part of the mesatergum. The volume enclosed by the metathorax is reduced; externally, this is seen most clearly by comparing the relative size of the lateral portion of the sternum and the L/W ratio of the episternum in flying and wingless species. In s,ubstriatus the narrowest lateral part of the sternum is about half the mid-line length and the L/W of the episternum is about 4 whereas in mutchleri the lateral portion is 20% of the mid-line length and the L,/W of the episternum is about I. The central part of the sternum is also shortened in Anti/- liscaris but less dramatically; the mid-line in substriatus is equal to the length of tthe first 3 112 abdominal sternites while in nzutchZeri it is equal to the length of the first 2 2/3 abdominal segments. The vestigial wings of mutchleri are very small (.375 X -100 mm) and circular in cross section in a specimen examined in fluid. In the Scaritini there are usually 3 lateral elytra-thorax and abdo- men interlocking mechanisms,
(Hlavac) unpublished observations.)
The anterior lateral edges of the elytra fit into grooves on the dorsal rim of the mesepisternum, mesepimeron and metepisternum (Fig. I 7). The dorsal part of the metepimeron slopes inwards; the elytra curve slightly outwards at this point and rest on top of the flat surface of



================================================================================

19691 ~HZavac - Scarites I3
the epimeron. An internal eIytra1 carina extends from just behind the level of the epimeron to the elytral apex and is enlarged pos- teriorly forming the internal plica, The carina rests against the edge of the infolded abdominal sternites. AntiZZiscaris differs from all other subgenera of Sc,arites examined, including many flightless species, in that the dorsal part of the metepi- meson is in the same plane as the episternum and the dorsal groove extends along the epimeron, i. e. the elytra-epimeron interlocking mechanism is absent, becoming part of the elytra-episternum device (Fig. 18).
Fig. 16 - Elytra-metathorax interlocking mechanism S. (s. s.) substriatus Fig. 17 - Elytra-metathorax interlocking mechanism, S. (Antilliscaris) (EC, internal carina of elytra ; Epm3, metepimeron; Eps3, metepisternum ; G, metapleural groove)
In burrowing and substrate dwelling Coleoptera, the elytral in- terlocking mechanisms maintain the structural integrity of the elytra- body joint against the forces generated in moving through a dense material. Lack of an integral portion of this mechanism in AntiZ- Ziscaris can be explained in two, non-mutually exclusive manners. If the ancestral stock were not a highly adapted burrower (see below), then loss of epimeron-elytra joint could be due to neutral selection pressure for structural integrity. Or, secondly, after becoming flight- less, the primary interlocking mechanism becomes the fusion of the elytra to each other and to the mesothorax; loss of epimeron-elytra joint represents post-adaptation21 adjustment; there is no need to assume a change in selection pressure for structural integrity. BODY FORM: S. (A.) megacephaZus and S. (s. s.) substriatus rep- resent extremes in body proportions with mutchleri and dunforthi intermediate. The outline of nzegacephaZus can be derived from that of substriatus by shortening the pterothorax and abdomen (Elytra L/TV from 1.8 to 1.4 in S. and A. respectively) and the prothorax



================================================================================

I4 Psyche [March
(L/W from 0.7 to 0.5) and by increasing the size of the head until it is 1.36 the length of the prothorax (0.82 in substriatus) and until the length of head and mandibles is 1.1 the length of the elytra (0.6 in subst&ztus).
A partial explanation of the shortening of the elytra is the atrophy of the flight mechanism. Differences in size between the head and prothorax can be looked at in terms of differences in muscle volume) i. e., a relative decrease in volume in the prothorax and a relative increase in the head. The prothorax houses most of the muscle vol- ume responsible for its movement as well as that which moves the head and proleg (Larsen) 1966, p. 140)) while the head encloses muscles responsible for the movement of its appendages. Ventro- dorsal rotation of the head and prothorax are principal actions in burrowing) decrease in muscle volume for performing these actions indicates a ,shift in or absence of burrowing behavior. The large head implies a large mandibular adductor muscle and this along with the large carnassal areas of the mandibles could be an adaptation for seizing and breaking open large prey objects. ECOLOGICAL SHIFT: Of the eight known montane species of cara- bids on Puerto Rico listed below, three are Scarites, and, with Oxy- drepan,ux coamensis Mutchler) scaritines make up 50% of the known carabid fauna. On Jamaica) an island of similar size (4500 sq. mi. vs. 3400 for P. R.) 17 montane restricted species are known) nine of which are Colfiodes; there are no Scarites ( Darlington) I941 ). As currently known, the montane biota of Puerto Rico is unique in the Greater Antilles in not having a large number of ecologically variable Colpodes. The size range of Puerto Rican species is 2.0 to 35 mm with the following distribution:
1-5 mm) 4. spp., Oxydrepanus coamensis Mutchler, 2 mm ; Bembidion portoricense Darlington, 3.5 mm ; Perigona microps Darlington, 2.9 mm ;
Phloeoxena portoricensis Darlington) 4.5 mm. 5-10 mm, I sp., Colpodes estriatus Darlington, 4.0-6.7 mm. 10-20 mm, 2 spp.) Antilliscaris danforthi, 14.6-1 6.0 mm ; Antilliscmis mutchleri) 12.0-1 8.0.
30+ mm,
S. (A*) megacephalus) 30-35 mm.
Antilliscaris is thus the largest carabid taxon, is at the apex of the size pyramid and is doubtless one of the top predators among ground invertebrates in this habitat.
It is useful to look at the mountains of Puerto Rico as habitat islands and to analyze the large Scarites fauna in the manner of MaArthur and Wilson (1967). In harmonic continental biotas,



================================================================================

the large number of carabid species occupy three ~ocornotmy adaptive zones: arboreal (several tribes), ground in the broadest sense in- cluding soil surface and leaf litter (most tribes and specis) and burrowing (primarily the Scaritini). All terrestrial zones available to a beetle are exploited by the Carabidae. Breadth of adaptahn for the family and high local diversity (see Darlington 1943 for data) is the result of adaptive radiation and species packing via competition. The energetics of food searching differs between the zone. If the energy expended by a ground species in searching a unit area is taken a base point, then both arboreal forms, which fly and climb vertical suriaces, and burrowers producing a new section of burrow expend more. In the absence of competition from related taxa, as in the mountains of Pwrto Rico for a large carabid, if ecolqical shift occurs in a new colonizxr it should be in the direction of increased &ciencY of search behavior.
The burrowing mechanism of Scaritine carabids consists of many interrelated morphological and behaviora1 adaptations. Evidence that several structures and proportions difier from a known burrow prc- ducer is presented above and can be interpreted as indicating a re- organization of the burrowing mechanm for a ground surface existence.
The steps taken by a species after colonizing a depauperate habitat island are: initial adaptation, eco1ogical shift, lw of dispersal power, speciation, and adaptive radiation ( MacArthur and Wilson 1967, Chap. 7). Speciation and 10s of dispersal power have occurred in AnMircaris; there is evidence for ecoIogicd shift. That an adaptive radiation has occurred in AtztiZliscar~ is an exciting posibiIity and ~hodd stimulate further fieId work.
BEHAVIOR : A specimen of S, [Am) mutchleri is currently being ob- served. While a fuIl report of its behavior is planned, the following observations are germane at this point, The bede has not con- structed an elaborate burrow system under conditions where sub- ~triatus and dt~rruneus would ( HIavac 1967) Daylight hours are spent under retreats or in conceaIment burrows, about 2 body lengths long, dug in arms where the sail has not been highly cmprtsed. During the night, this animal has eaten food placed on the surface and fallen into a ma11 trap in the center of the container 24 nights out of 37. Thew fragmentary data are consistant with the hypothsis of ecoIogical shift.
MCGEOGRAPHY: The factors responsible for the high frequency of wing atrophy in montane carabids have been analyzed by Darling- ton ( 1943). He pointd out ( 1939) pp. 79-80) that mountains over



================================================================================

16 Psyche [March
sooo ft. high on Cuba, Jamaica, and Hispaniola have endemic mon- tane species that had become -winged on the spot. Noting that three -winged stocks on the mountains of Puerto Rico (Antilliscaris, Oxy- drepmus and Colpodes estriatus Darl.) were closely related to forms known only from the mountains of Haiti, he concluded that each had reached Puerto Rico from Haiti after becoming flightless, and that mountains lower than 4000 ft. in the Antilles "are too low to induce wing atrophy among Carabidae." S. darlingtoni was here held to be closely related to the known species now in Antilliscaris. While I consider this species tol be incertae sedis at the subgeneric level, the evidence for so doing (see systematic treatment) is not overwhelming, and this decision mainly reflects caution. In any case, questions as to the point of origin and the site of wing atrophy remain. Long distance dispersal of montane forest-floor-inhabiting, flight- less carabids to another suitable montane environment must be a rare event. Such species have all "bad" attributes that preadapt a species to be "good" colonizers (MacArthur and Wilson 1967, p. 81). Flightless carabids are rare in
the tropical lowlands (Darlington
1943, pp. 44, 49-50) and would be severely selected against in areas subject to flooding. A +winged hydrophilic lowland species has many "good" preadaptations for colonization, such as occupation of an unstable habitat including many coastal sites with the potential for high population densities in hydric environments and, of course, the ability to fly.
With 3 montane spp. of Scarites on Puerto Rico, additional col- lecting on Haiti should turn up new species if the area species curve is remotely applicable to montane scaritine faunas (but see comments on ecological shift). When more data becomes available, I propose that the following model be tested. Major colonizing Scarites is a +winged form which burrows in a variety of shore and lowland habitats. It disperses to similar habitats on neighboring islands and subsequently invades the montane zone where -wings and flight mechanism atrophy occur along with ecological shift. Evidence for this interpretation could be obtained from detailed analysis of widely distributed +winged lowland forms.
Colpodes estriatus Darlington
and its allies probably represent a distinct genus (Darlington 1939, p. 95). Then, of the three taxa represented in the mountains of Puerto Rico and Hispaniola by a -winged species one, Oxydrepanus, is widely distributed in the lowlands of the Greater Antilles; 0. rubs (Putz.) is known from Cuba, Jamaica and Florida, and 0. micans Putz. from Hispaniola and Puerto Rico; both species



================================================================================

19691 dkl/avac - scarites 17
+winged. A second montane species, 0. reichoides Darl., occurs on Hispaniola.
The sophisticated taxonomy necessary to separate monophyletic from polyphyletic origins of -wings in closely related lower cate- gories does not now exist for Scarites.
Until a more refined sys-
tematics becomes available, I think it heuristic and economical to view the biogeography of Antilliscaris in light of the model presented above. PRELIMINARY EVOLUTIONARY DIAGNOSIS: From an assumed low- land, winged and burrowing ancestor, Antilliscaris has entered the montane rain forests of Puerto Rico, become flightless, undergone an ecological shift from burrowing to ground crawling, and speciated. REFERENCES CITED
BANNINGER, M.
1935. Uber alte und neue formen der Scaritina. Ent. Blatt. 1935: 148- 160.
1937-39. Monographic der subtribus Scaritina (Col. Carab.). Pt. I- Deutsche Ent. Zeitschr. 1937: 6-160. Pt. I1 - Deutsche Ent. Zeitschr. 193 8 : 41-181. Pt. I11 - Deutsche Ent. Zeitschr. 1939 : 126-161.
1949. Uber Carabinae. Mittellungen der Miinchner. Ent. Gesellschaft. DARLINGTON, P. J., JR.
1939. West Indian Carabidae V. New forms from the Dominican Re- public and Puerto Rico. Mem. S.oc. Cubana Hist. Nat. 13: 79-101. 1941. West Indian Carabidae VI. The Jamaican species and their wings. Psyche 48 : 10-15.
1943. Carabidae of mountains and islands: data on the evolution of isolated faunas, and on atrophy of wings. Ecol. Monog. 13: 37-61.
HLAVAC, T. F.
1967. Observations on behavior in Scarites. Col. Bull. 21: 18-22. LARSEN, 0.
1966. On the morphology of a function of the locomotor organs of the Gyrinidae and other Coleoptera. Opus. Ent. Sup$. 30: 1-241. MACARTHUR, R. H., and E. 0. WILSON
1967. The theory of island biogeography. Princeton U. Press, 203 pp. MUTCHLER, A. M.
1934. New species of Carabidae from Puerto Rico. Amer. Mus. Nov. No. 686, 5 pp.
SMITH, D. M.
1964. The structure and development of flightless Coleoptera: a light and electron microscopic study of wings, thoracic exoskeleton and rudimentary flight musculature. J. Morph. 114: 107-184.



================================================================================

STUDIES ON NEOTROPICAL POMPILIDAE
(HYMENOPTERA)
V. AUSTROCHARES BANKS*
BY HOWARD E. EVANS
Museum of Comparative Zoology
In the previous paper of this series (Evans, 1968), I described several cases of dual sex-limited mimicry in the genus Chirodamus and mentioned that certain species of Austrochares also exhibit this phenomenon. Since the ,situation is somewhat different from that prevailing in Chirodamus, and since I had previously stated (1966) that Austrochares as originally conceived by Banks ( 1947) was diphyletic, it seems desirable to clarify this matter further. Austro- chares was based upon gastricus Spinola and properly includes this species and four others; this genus is closely related to Episyron, Poecilopompilus, and Sericopompilus. The remaining species placed in the genus by Banks are much more closely related to Anoplius and Pompilus and are properly assignable to the genus Dicranoplius Haupt (1950). I hope to review Dicranoplius in the next paper of this series.
The cases of dual mimicry occur in Austrochares as here restricted, and not in Dicranoplius. As in the instances of dual mimicry I described in Chirodamus, the females belong to a Mullerian mimetic complex principally involving solitary wasps, while the males appear to be Batesian elements in a large mimetic complex centering around certain social wasps. Both complexes happen to be different than in Chirodamus, but I assume that similar selection pressures have been operative. The females evidently spend most of their time on the ground hunting for spiders and have evolved a color pattern of black and rufous like that of a great many solitary wasps occurring in the same situations (e.g. Dicranoplius satanus) (Figs. 2, 4 ) while the males spend most of their time flying in low vegetation in company with the workers of various social Vespidae. In this instance the model of the males is undoubtedly Polybia occidentals Olivier and its complex of sibling species (partially re- solved by Richards, 195 I ). Bequaert ( I 944) speaks of occidental& (that is, of the complex) as "the most common social wasp of tropical America". The males are black except for conspicuous yellow spots on the ,scutellum and metanotum, a yellow margin on the pronoturn, *Manuscript received by the editor November 25, 1968 18




================================================================================