Gryllus armatus Scudder

Armed Field Cricket

Figs 87, 96–98, 100, 104–112, Table 1

1902 Gryllus armatus Scudder. Psyche 9: p. 293. Lectotype male here designated (Fig. 105, photos courtesy of J. Weintraub) “Palmer’s assorting no. 1072. Ari. Gr. armatus Scudder’s type, 1901. Red label, Type no. 14067.” Labeled cotype #7 by D.C. Rentz, 1979. Scudder’s lectotype and type series at ANSP, now consists of four males and one female (Scudder’s original series listed as 6 males and 1 female) representing at least two and possibility three species. While the lectotype is missing both hind legs and may be distorted due to past alcohol preservation, it clearly agrees with Scudder’s 1902 description as follows: (1) locality of Arizona (either Ehrenberg or Fort Whipple [the latter site near Prescott], Arizona); (2) head and pronotum shining black; (3) contrasting two toned black pronotum and beige tegmina; and (4) long hind wings. Lectotype measurements: body length 17.5, PW 5.9, PL 3.5. Right tegmen removed: 120 teeth, file length 3.1, TL 14.1, TW 4.15. Currently, Cigliano et al. (2019) list a male lectotype from “ Arizona, Ehrenberg or Ft Whipple”, repeating an erroneous 2013 Orthoptera Species File iteration that discussed how Weissman et al. (1980) designated a lectotype for G. armatus . In fact, we made no such formal designation in 1980 but only discussed, on p. 345, Scudder’s original type series.

Gryllus integer (in part) of Weissman et al. (1980). ‘ Gryllus #6’ of DBW notebooks.

Distribution. Southwest and southcentral US.

Recognition characters and song. A small to medium sized cricket with long, narrow tegmina, always long hind-winged (except in certain dealate individuals), probably 2 generations/year with the second summer generation frequently of contrasting solid black head and usually black pronotum coupled with beige/straw colored tegmina (see Fig. 105) and legs. Morphologically indistinguishable from sister species G. integer but separated by habitat and geography, slight song differences, and consistent DNA differences, as follows: (1) G. armatus is from hotter, more southern desert US locations (see Fig. 106); (2) Most, but not all G. armatus, have an “introductory trill” of 2–8 evenly spaced pulses at the beginning of each burst of chirps (see Figs 107, 108, S 15-58, R15-177), usually followed by 2 (range 1–3) p/c until that burst is over. In contrast, G. integer usually with no introductory trill and 3 p/c, although some males sing with 2–4p/c); and (3) there is a congruent, and consistent association between those males with 2p/c and parallel differences in the ITS2 gene sequences when compared with G. integer (see Fig. 98). In distribution map Fig. 106, note a possible zone of hybridization in the Four Corners area of southeastern Utah, west- central Colorado, and north-eastern Arizona where males sing more like G. armatus but possess G. integer DNA. In this paper, we arbitrarily give priority to the DNA and consider such populations to be G. integer . Nevertheless, it is the extensive DNA/song congruity that ultimately convinced us to consider these two taxa as different and retain both of Scudder’s names.

Scudder (1902) notes that, in G. armatus, the “Pronotum rather more than half as broad again as long…” While this condition exists in some individuals of G. armatus, it is not universal nor diagnostic of any Gryllus species.

Song easily distinguished from other Southwestern crickets, except G. integer, because none have a combination of 2–3p/c, 700–1000 c/m, and a PR> 70. Most easily separated from G. integer, as discussed above, and by the distribution map (Fig. 106). Because different males within the same population have variable songs, we emphasize the significance of, and need for, multiple male song analysis. A couple of examples of song variation in G. armatus, within the same population, are here provided: (1) All 7 males from Deming, NM (S07-79) had introductory trills and 2p/c. (2) In contrast to the “clean and uniform” situation at Deming, of 7 males collected in Baker, CA (S05- 110), 5 males had introductory trills and then 2 (occasionally 3) p/c; 1 male had no introductory trill but 2 (rarely 3) p/c; and the 7 th male sang with an irregular trill (Fig. 109, R05-43). The latter male also had a red pronotum (see Fig. 111b) and a low PR of 38.5, yet his ITS2 DNA mapped with the other 6 males from this stop (Fig. 98), and consistent with G. armatus . Morphologically, this long-winged male agreed with typical G. armatus . (3) In Big Bend National Park, TX (S91-43 & S16-12), 3 of 6 males had introductory trills followed by 2p/c while the other 3 males had no introductory trills and 2p/c. (4) At Shoshone, CA (S92-60), some males sang with an introductory trill and then 2p/c while other males sang with no trill and 3p/c. (5) The only male heard in Bisbee, AZ (Fig. 110, R 13-52, S13-18), had an introductory trill and then groups of 2 to 10 pulses/chirp.

Distinguished from G. texensis, with which it overlaps in the eastern part of G. armatus’ Texas distribution, by G. armatus’ generally higher number of file teeth vs. shorter hind femur length, and always long hind wings. Some 22% of G. texensis have short hind wings and are never as clearly two-toned as seen in G. armatus and G. integer . Very warm (>35°C) field-singing individuals of G. vocalis can sound like G. armatus but are clearly identified when they sing in the laboratory, around 25°C, because of their slower chirp and pulse rates.

Derivation of name. “arma” is Latin meaning weapons, and “-atus” is Latin meaning provided with, having the nature of, pertaining to, apparently in reference to the increased number of spines on the outer side of the hind tibia (seven to eight) when compared with the three other Gryllus species that Scudder (1902) believed to occur in the “southern Rocky Mountain region (Colorado, Utah, New Mexico and Arizona).” Scudder (1902) notes that in the southern US, both G. firmus and G. rubens have 7–8 spines on the hind tibia. In fact, since the type series of G. armatus is comprised of at least two, and possibly three species, his hind tibia spine distinction is useless. Plus, a quick comparison of our extensive, song-associated series, shows variation of between 5 to 7 spines in both G. armatus and G. integer within many populations.

Geographic range. (Fig. 106). Southwestern US deserts including the eastern Mohave, Sonoran, and Chihuahuan Deserts with other non-desert populations in southwestern Utah on the edge of the Great Basin Desert, northeastern Arizona, northern New Mexico, eastern Colorado east of the Rocky Mts., southern Texas , and western Kansas.

Habitat. From near sea level to 2190m (El Malpais National Monument, NM, S96-68). Often in towns and associated with a variety of microhabitats: watered lawns, garbage dump areas, cracks in concrete and dry ground, around florescent lights at gas stations, soil cracks in dry pond and lake beds (El Malpais National Monument, NM, S96-68).

Life cycle and seasonal occurrence. No egg diapause: Cottonwood Cove, NV (S90-44); Arizona Western College (S90-54); Van Horn, TX (S91-48); near Goffs, CA (S16-32). Probably two generations/year but not sure if drought conditions at certain times of year, can limit populations to one generation/year.

Variation. Color: Second generation adults frequently contrasting two toned with black head and (usually) black pronotum but beige/straw tegmina and legs. Male G513 from Baker, CA (S05-110), with a reddish pronotum (Fig. 111b). A male from Cornudas, TX (Fig. 112, S01-65, R01-177) darker than those Baker, California specimens. Hind wing length: Always with long hind wings although these are occasionally shed in field collected adult males (Weissman et al. 1980). Of 268 song-confirmed G. armatus males, all but 4 had intact long hind wings. Those without long hind wings were, in fact, dealate, a condition discussed elsewhere for Acheta domesticus (Weissman & Rentz 1977b, Walker 1977). Ninety-nine field collected females all had long hind wings. Song: See discussion above under Recognition characters and song.

Specimens examined. Arizona: Cochise Co., Bisbee, 5500’, 1-vi-2013 (S13-18). 2m E Bowie, 30-vii-1981, (S81-39). Coconino Co., Tuba City area, 6-viii-1991, 4800’ (S91-82). Hwy 99 6.85m NW intersection with I40. 8-viii-1991 4900’ (S91-89). Gila Co., Coolidge Dam, 30-vii-1981, 2400’ (S81-43). Globe, 30-vii-1981, 3548’ (S81- 44); 10-vi-2012, (S12-21). Graham Co., Safford, 30-vii-1981, 2920’ (S81-41); 16-vi-1990 (S90-51); 28-vii-2015 (S15-103). Prison Hwy 366 near junction Hwy 191, 3333’, 28-vii-2015 (S15-102). La Paz Co., Ehrenberg, 3-ii- 2008, 190’ (S08-4). Quartzsite, 1000’, 26-vi-1980 (S80-46); 27-vii-1981 (S81-33); 26-viii-1982 (S82-104). Wenden, 14-ix-2011, 1800’ (S11-87). Maricopa Co., Aquila, 23-vii-1990, 2100’ (S90-71). Buckeye, 18-ix-2011, 840’ (S11-102). Goodyear, 31-vii-1981 (S81-46). Mohave Co., Bullhead City, 14-vi-1990, 500’ (S90-47). Davis Dam, 14-vi-1990, 900’ (S90-46). Navajo Co., Holbrook, 8-ix-1999, 5080’ (S99-116); 2002-032, 9-viii-2002. Hwy 77 0-12m N I40, 7-viii-1991, 5400’ (S91-86). Pima Co., Catalina, 18-viii-1998, 2940’ (S98-65). Tucson, 28-vii-1981 (S81-35); 17-viii-1998, 1920’ (S98-63). Yavapai Co., Camp Verde, 21-viii-2012, 3151’ (S12-107). Yuma Co., Arizona Western College, 10-viii-1988, 200’ (S88-89); 18-vi-1990 (S90-54). Telegraph Pass, 15-ix-2011, 676’ (S11- 92). Yuma, 31-vii-1981 (S81-48). California: Inyo Co., Death Valley National Park, Furnace Creek, 0’, 23-vi-1980 (S80-32); 5-vi-1983 (S83-60); 25-vi-1992 (S92-61); Stovepipe Wells, 5’, 5-vi-1983 (S83-59); 25-vi-1992 (S92- 62). Shoshone, 25-vi-1992, 1560’ (S92-60). Riverside Co., Blythe, 400’, 26-vi-1980 (S80-47); 14-ix-2011, 240’ (S11-90). Indio, 6-viii-1988 (S88-74). Palm Springs, 2-iv-1989 (S89-8). San Bernardino Co., Baker, 4-viii-1991, 1000’ (S91-72); 30-viii-2005, (S05-110). Barstow, 16-viii-1998, 2420’ (S98-58); 28-vi-2003 (S03-73); 22-vii-2016 (S16-28). Havasu Lake, 6-vi-1983 (S83-62); 13-ix-2011, 460’ (S11-84). Ludlow, 2060’, 16-viii-1998 (S98-60); 23-vii-2016 (S16-30). Newberry Springs, 16-viii-1998, 2160’ (S98-59). Truck stop at Goffs exit off I40, 2103’, 23-vii-2016 (S16-32). San Diego Co., Borrego Springs, 8-viii-1988 (S88-83). Colorado: Crowley Co., Manzanola, 26-viii-1989, 4200’ (S89-65). Otero Co., La Junta, 26-viii-1989 (S89-66). Pueblo Co., Baxter, 26-viii-1989 (S89- 63). Boone, 26-viii-1989, 4350’ (S89-64). Kansas: Ford Co., Dodge City, 27-viii-1989, 2400’ (S89-71). Nevada: Clark Co., Cottonwood Cove, 800’, 24-vi-1980 (S80-36); 26-vii-1981 (S81-31); 14-vi-1990 (S90-44). Searchlight, 3512’, 26-vii-1981 (S81-30); 23-vii-2016 (S16-32). New Mexico: Bernalillo Co., Albuquerque, 22-viii-1982 (S82- 89); 1-vii-1994 (S94-33). Catron Co., Reserve, 3-vii-1994, 5770’ (S94-50). Cibola Co., El Malpais National Monu- ment, Hwy 117, 31m S Hwy 40, 14-vi-1996, 7200’ (S96-68). Dona Ana Co., Las Cruces, 23-viii-1982 (S82-99); 26-vi-1983 (S83-103); 16-vi-1986 (S86-40); 5-vii-1986 (S86-78); 26-iv-1990 (S90-35); 26-ix-1992 (S92-127); 1- vi-2004 (S04-37). Eddy Co., Artesia, 3428’, 30-vi-2015 (S15-58). Hope, 4095’, 30-vi-2015 (S15-57). Hidalgo Co., Roadforks, 29-vii-1981, 4195’ (S81-38); 19-viii-2012 (S12-104). Lea Co., Eunice, 6-ix-2010, 3420’ (S10-62). Luna Co., Deming, 21-vii-2007, (S07-79). Socorro Co., Socorro, 4420’, 29-vi-2015 (S15-54). Hwy 25 1m S La Joya, 15-viii-1993, 4850’ (S93-68). Socorro, 13-vi-2007, 4460’ (S07-50). Texas: Brewster Co., Alpine, 5-vi-1991, 4200’ (S91-44); 12-vi-2007 (S07-41); 2-vii-2015 (S15-73). Big Bend National Park, Rio Grande Village, 5-vi-1991, 1860’ (S91-43); 28-v-2016 (S16-12). Hwy 118 near Terlingua, 2566’, 2-vii-2015 (S15-72). Cameron Co., Rio Hondo, 8’, 10-vii-2013 (S13-44); Hwy 4 3.1m W Boca Chica State Park, 10-vi-2007, 10’ (S07-26). Culberson Co., Van Horn, 6-vi-1991, 4100’ (S91-48). Hidalgo Co., Bentsen-Rio Grande State Park, 110’, 3-viii-2002 (S02-34), 10-vi-2007 (S07-27). Howard Co., Big Springs VA hospital grounds, 30-vi-2009, (S09-72). Hudspeth Co., Cornudas, 13-vii- 2001, 4420’ (S01-65). Jeff Davis Co., Ft. Davis, 4852’, 1-vii-2015 (S15-67). Kinney Co., Brackettville, 4-vi-1991, 1100’ (S91-40); 11-vi-2007 (S07-32); 7-ix-2010 (S10-63). Presidio Co., Presidio, 2580’, 27-v-2016 (S16-5). Hwy 67 just S Shafter, 3920’, 27-v-2016 (S16-3). Val Verde Co., Del Rio, 11-vi-2007, 1000’ (S07-33). Ward Co., Monahans, 2-vii-1986 (S86-69). Willacy Co., Raymondville, 10-vii-2013 (S13-47); Farm Road 1420 3.1 m N Farm Road 1018, 10-vii-2013, 15’ (S13-46). Utah: Washington Co., La Verkin, 11-ix-2004, 3420’ (S04-121). 1m NE La Verkin, 9-viii-1991, 3700’ (S91-95).

DNA. Multilocus G3439 (Goffs, CA, S16-32) and G3374 (Presidio, TX, S16-5) map together and nearest to G. integer from Tulare Co., CA (Gray et al. 2019). Fig. 98 shows the distribution of ITS2 mapping done for both G. armatus and G. integer . Sequence identity of ITS2 sequences (N = 684 aligned positions) was 98.5 ± 0.87 % within N = 29 G. armatus, 99.0 ± 0.77% within N = 34 G. integer, and 95.7 ± 0.78% between the two species.

Discussion. Scudder’s series includes 3 males (our lectotype and D.C. Rentz’s cotypes labeled #4 and #6) from “ Arizona ”, but none are further labeled as to locality. We collected an Ehrenberg, AZ (S08-4) specimen, of G. armatus, that sang (R08-04 & R08-06) with an introductory trill and 2p/c and maps with G. armatus for ITS2, confirming the ID of one AZ type locality. The other possible AZ type series locality, Fort Whipple, is located near Prescott. While we have no specimens of G. armatus from Prescott, that area is still within the geographic range of G. armatus (see Fig. 106). Other specimens in the type series include D.C. Rentz’s labeled Utah cotypes #5 (male) and #8 (female). This male and female from Utah (Beaver Dam), are both well preserved, short hind winged specimens that may represent G. vocalis . While Scudder (1902) lists the Utah locality as Beaver Dam, this is probably not the unincorporated community of Beaver Dam in Box Elder Co., Utah, some 4 km E Fielding. Rather, Richard Baumann of BYU (pers. comm. to DBW, September, 2012) believes this original Beaver Dam locality is actually the Beaver Dam Wash, Washington Co., Utah, a seasonal stream near the southwestern Utah-Nevada border that occurs upstream from the unincorporated community of Beaver Dam, Mohave Co., Arizona. According to R. Baumann, BYU has many labeled specimens that say Beaver Dam, Utah, and they are from Beaver Dam Wash in Washington Co. This area is possibly near Lytle Ranch Preserve, a field station owned by BYU since 1986 (S. Clark, collection manager BYU, pers. comm. to DBW, January, 2017). Vasco M. Tanner did much collecting there after his arrival at BYU in 1925. In any case, we have collected G. armatus from Washington Co., Utah (S91-95 & S04-121). The Beaver Dam in Box Elder Co., Utah, is near Cutler Dam. The nearest Box Elder Co., Beaver Dam appropriatelysinging cricket that we collected was from S87-51, 19 km north Tooele in Tooele Co., and the one male there usually sang with 3p/c (occasionally 2p/c) without an introductory trill. Given our absence of genetic data, he is tentatively identified as G. integer based on song.

Our initial 16S genetic data showed clear separation between G. integer and G. armatus (see Fig. 104), but given our previous documentation of multiple 16S clades, in species G. vernalis, veletis, and lightfooti, we were unsure of this significance. With our subsequent incorporation of the ITS2 gene, such separation was again confirmed. Nevertheless, several populations along the song-boundary between G. integer and G. armatus, have G. integer ITS2 DNA and a mostly G. armatus song. These populations are discussed more completely under G. integer (see p. 104). We also entertained the thought that the G. integer—G. armatus complex could represent one widespread species with a west to east cline in song pattern. Again, subsequent DNA data corresponding with song differences and temperature regimes, persuaded us that two sister taxa were a better explanation.

We suggest that a minimum of 3- 5 males per population be analyzed with sonograms, because of individual variation and because subtle song differences between G. integer, G. armatus, G. rubens, and G. texensis are not audible to most human ears. If the majority of males have an introductory trill followed by 2p/c, then they are probably G. armatus . Likewise, several males with no introductory trill and 3p/c are likely G. integer . But select G. armatus -DNA males can also have no trill and 2p/c and, rarely, no trill and 3p/c. And to complicate matters, males with G. integer DNA can have a short introductory trill of 1–3 pulses followed by 2–4p/c and, rarely, a pure trill. In the face of such variation, multiple male analysis from the same population has usually given us a definitive answer. Additionally, it appears that the farther east one goes within the distribution of G. armatus, the higher percentage of males have introductory trills followed by 2p/c, although all 3 males from Texas, Bentsen-Rio Grande State Park (S02-34 and S07-27), where G. armatus occurs microsympatric with G. texensis, had 2p/c without an introductory trill.

Interestingly, two males changed their song with time. One male recorded (R91-143) in the field from Tuba City, AZ (S91-82), had a short introductory trill at 22°C and then 2p/c, consistent with G. armatus . Five days later in the laboratory at 25°C, this male sang (R91-155) without an introductory trill and with 2–3p/c, more consistent with G. integer song. He was not teneral upon capture. No DNA could be isolated more than 15 years later. A second male (2006-241) from Cordes Junction, AZ, was initially recorded at 20.8°C, with no introductory trill and 2p/c. A subsequent recording, at 22.4°, showed an introductory trill and 2 (rarely 3) p/c.

Rare field collected adult males of G. armatus sing a pure trill song of 1p/c: Coolidge Dam (S81-43) and Buckeye (S11-102), AZ; Las Cruces, NM (S82-99); and Baker, CA (S05-110)], or predominantly so (Cottonwood Cove, NV [S90-44]; Globe, AZ [S82-103]; Las Cruces, NM [S83-103]; and Van Horn, TX [S91-48]) but, in all cases, the fast PR separates them from other trillers G. regularis and G. cohni in the Southwest, G. rubens in Texas, but not from G. texensis in Texas (see discussion in next paragraph).

G. armatus is sympatric with the Southeast fast trilling field cricket, G. texensis, at the Texas localities of Alpine (S91-44, S07-41, along with the slower triller G. regularis); Brackettville (S91-40, S07-32, S10-63); Rio Hondo (S13-44); town of Van Horn (S91-48); Big Springs (S09-72); Rio Grande Village in Big Bend National Park (S91-43, S02-34); and Bentsen-Rio Grande State Park (S02-34, S07-27); in Kansas at Dodge City (S89-71); and in New Mexico at Eunice (S10-62). When both G. armatus and G. texensis are singing adjacent to each other, and G. armatus is singing with 2 or 3p/c, differences in song, especially evenness and pitch, are apparent to “young” ears. When singing isolated from each other, we do not appreciate such song differences as easily. Because some 22% of adult G. texensis have short hind wings, the taxa can sometimes be separated in the field even though both have PR between 70-100 at 25°C

The population of crickets from Aguila, AZ, studied by Hedrick and Kortet (2006) under the name G. integer, is most likely G. armatus .

The tachinid Ormia ochracea recovered from G. armatus from single males collected at Havasu Lake, CA (S11-84) and Wenden, AZ (S11-87).