Click here for the original journal page (in Acrobat pdf format).

The text below is grayed out because it is not intended to be read. It is a necessarily imperfect OCR of the original and is only used by a search engine.

Journal of

HE LEPIDOPTERISTS' SoCIETI

Volume 52                                      1998                                      Number 3

Journal of the Lepidopterists' Society 52(3), 1998, 229-261

BUTTERFLIES OF THE UPPER FRIO-SABINAL REGION,

CENTRAL TEXAS, AND DISTRIBUTION OF FAUNAL

ELEMENTS ACROSS THE EDWARDS PLATEAU

David E. Gaskin1

Department of Zoology, University of Guelph, Guelph, Ontario NIG 2W1, Canada

ABSTRACT. A survey of the butterfly fauna (1988-96) of the upper Frio-Sabinal region of the southern Edwards Plateau, Texas, is presented. Butterflies were observed along transects at five study sites and during repeated opportunistic transects at 12 secondary localities. A total of 28,035 specimens, comprising 100 species was recorded. Another 51 species were reported by other lepidopterists working 15—25 km south of this region during the same period; most were collected from the vicinity of Concan, in north-central Uvalde Co. Twenty-seven species recorded in the upper Frio-Sabinal region were represented by 1—5 specimens only. Simple approximation models were used to estimate the proportion of the theoretical total species collected to date at the upper Frio-Sabinal site. No statistically significant differences were found between the geographical components at the Frio-Upper Sabinal site and two other well-worked sites; Barton Creek (Travis Co.) and Concan (N. Uvalde Co.). Composition and distribution patterns of the butterfly fauna across the Edwards Plateau were examined by analyzing data from 16 reasonably well-collected counties. Geographically, the butterfly fauna across the Edwards Plateau has a strong W/SW trans-Pecos component in Brewster and neighboring counties, which is only weakly represented in the north and east. A S/SE element is significant only along the Balcones fault region from Uvalde to Travis counties, while a NE/E element is numerically important but decreases sharply west of Real and Uvalde counties. Both are associated with the riparian corridors of the southeast. A N/NW element is widespread, but only weakly represented in all counties. Disturbed habitats were dominated by Pieri-dae (64%). Intergrading dry, subtropical habitats, dry montane woodland areas, coastal woodlands and southern tropical woodlands were dominated by Hesperiidae (45-75%). Richest ecological zones were the south tropical woodlands (49 species), dry subtropical forest and scrub (32 species) and Great Plains savanna habitats (16 species). Other ecological zones were characterized by 12 species or less. The Hesperiidae was the best represented family (97 species) and the Pyrginae the most abundant subfamily (53 species). There is little endemism in the butterfly fauna of Edwards Plateau. The relative species richness (227, with 35 more in Brewster Co. only) can be attributed largely to its strategic geographical position. It is difficult to specify any one element on the Plateau which could truly be said to be characteristic of the butterfly fauna, because so many of these species reach a range limit at some point on the Plateau.

Additional key words: Papilionioidea, Hesperioidea.

Deceased

230

Journal of the Lepidopterists' Society

Texas has a diverse butterfly fauna commensurate with its size, wide latitudinal range, topographic variety and complex vegetational zones. The southernmost counties of Texas in particular attract entomologists seeking Neotropical species rarely encountered elsewhere in North America (Opler 1993). Information on the butterflies of Texas is scattered throughout books (e.g., Holland 1930, Klots 1951, Howe 1975, Pyle 1981, Opler & Krizek 1984, Scott 1986, Opler & Malikul 1992, Stanford & Opler 1993, Neck 1996), journal articles (e.g., Freeman 1951, Kendall 1964, 1976, Durden 1982, McGuire 1982, McGuire & Rickard 1974), and annual summary reports of the Lepidopterists' Society. However, only the survey by Durden (1982) treats the south-central region in detail. His decade-long survey concentrated on the butterflies found in the ten counties surrounding Austin, with particular focus on Barton Creek Canyon, a locality at the eastern boundary of the Balcones Fault Zone, and in the transition area of the Edwards Plateau and South Central Vegetational Zones (sensu Gould 1969, Ajilvsgi 1984, Amos & Gehlbach 1988, Enquist 1987).

The present paper describes and quantifies the butterfly fauna of the upper Frio-Sabinal region, which encompasses the confluence of the East and West Frio Rivers, in the area where Real, Bandera and Uvalde counties meet. Results are compared with data from two other intensively worked sites, Concan (E. Knudsen 1996, in lift.) in central Uvalde Co., 15—25 km south of my Frio-Sabinal sites, and Barton Creek Canyon in the Austin region, 160 km to the northeast (Durden 1982).

Relative distributions of the butterfly fauna across 16 counties of the Edwards Plateau are analyzed by geographic components (5), taxonomic groups (15), habitat types (18) and range position or limit (12), the last two following designations by Durden (1982). Comparisons of species-richness in the counties across the plateau are made with caution. While these can reveal real differences attributable to habitat diversity or differences in magnitude of unit area, they can also be a function of unequal sampling effort. Two approximation models were used to identify under-collecting, and assess the probable level of completeness of the total species counts, for 16 counties across the Edwards Plateau for which reasonable data are available.

Materials and Methods

Thirteen expeditions were made to the upper Frio-Sabinal Rivers area of the Edwards Plateau, popularly known as "The Hill Country", between March 1988 and April 1996. The study covered localities in adjacent parts of Real, Bandera and Uvalde counties bounded by 29°25-55'N and 99°20-50/W. Observations were made over 224 h during 160 days at five main study sites (Fig. 1), from 243 linear transects on 15

Volume 52, Number 3

231

Fig. 1. The upper Frio and Sabinal River region on the south-central Edwards Plateau, Texas. Dotted lines show county boundaries; solid lines state roads (with numbers) and county roads. Study sites have a county abbreviation. The five primary transect sites are in bold.

measured track lines, totaling some 285 km, and an additional 60 h of opportunistic transects at 12 secondary sites (Fig. 1). In both cases, transects were designed, based on initial surveys, to sample all major habitats at each site. Most quantitative observations were collected at sites RCl-3 near the town of Leakey and UCl-2 on the Blanket Creek west of Utopia (see below). Transect methodology followed Pollard (1977, 1991), Pollard and Yates (1993), Thomas (1983) and Warren et al! (1984), with minor modifications appropriate to local circumstances.

To facilitate data comparisons, particularly with Durden's detailed butterfly records from Barton Creek, I used species rather than subspecies. Although Durden (1969) argued that "the subspecies is the primary occupant of the niche", the value and definability of either concept is debatable (e.g., Collins 1991, Ehrlich & Murphy 1981a, 1981b, 1982, Hammond 1985, 1990, Kudrna 1986, Miller & Brown 1981, Shapiro 1982), and some of the subspecies listed by Durden have either undergone status changes or are no longer recognized.

232

Journal of the Lepidopterists' Society

Relative abundance of each species was assessed by four categories similar to those used by Durden (1982): A, abundant—a series of five to ten specimens can be taken in an hour or so; C, common—such a series can be taken in a day's collecting; U, uncommon—takes several seasons to collect such a series; S, scarce—only one to five specimens encountered during the entire study. Durden s category A incorporates both my categories A and C. In my modification, category A is reserved only for the handful of species for which well over 1000 specimens were recorded during the survey

Unlike Durden's semi-quantitative records, results from the time and distance-based transects used in the Frio-Sabinal survey also permit data to be expressed in other ways, such as the logarithmic scaling of numbers per unit search time (see Results), as proposed by Clench (1979).

Butterflies were observed during the following periods: 1988: 16—23 Mar.; 1989: 11-16 May; 1990: 16-21 Sept.; 1991: 6-10 Jul.; 1992: 5-28 May; 5-12 Oct.; 1993: 4-19 May; 1994: 14-24 May; 6-16 Jun.; 8-24 Sep.; 1-7 Nov.; 1995: 12-24 Apr.; 1996: 10-24 Apr. Butterflies were flying on all the days indicated, although in small numbers on two days in March 1988 and two in May 1994.

Several study sites within each of three counties were surveyed (Fig. 1).

NORTHERN UVALDE COUNTY: UC1. 547 m. Blanket Creek, on the Indian Blanket Ranch (primary study site subject to periodic minor flooding once or twice each year); UC2. 550-580 m. Adjacent section of Ranch Road (RR) 1050 verges (subject to periodic minor flooding); UC3. 663 m. Edge of escarpment on RR 1050 about 6 km from Utopia; UC4. 540 m. Section of RR 2748, about 1 km from RR 1050 junction.

WESTERN & CENTRAL BANDERA COUNTY: BC1. 450 m. Verges of RR 337 about 2 km from junction with HW 187; BC2. 520 m. Gun Mountain Ranch near Tarpley; BC3. 470 m. Canyon Creek region of Lost Maples State Park (primary study site); BC4. 510 m. Verges of RR 470 about 3 km from junction with HW 187; BC5. 430 m. On RR 470 about 7 km from Highway (HW) 187.

EASTERN REAL COUNTY: RC1. 493 m. "Rio Lindo" property about 3 km N of Leakey, on RR 336 (primary study site subjected to one major flood episode in Dec. 1990); RC2. 487 m. West bank of the Rio Frio just E of Rio Lindo (primary study site subject to two major flood episodes, Dec. 1990 & Apr. 1994); RC3. 472 m. Twin Forks Estate, about 3 km E of Leakey on RR 337 (primary study site subject to two major flood episodes, Dec. 1990 & Apr. 1994); RC4. 450 m. Verges of RR 337 about 5 km E of Leakey; RC5. 550 m. Verges of 337 about 5 km W of Leakey on RR 337, towards Camp Wood; RC6. 527 m. Verges of RR 336 each side of crossing with Cedar Creek, about 7 km N Leakey (primary study site subject to periodic minor flooding).

Volume 52, Number 3

233

FIG. 2. Schematic map of the southern and central greater Edwards Plateau (after Amos & Gehlbach 1988), showing the 16 counties analyzed. Dotted lines indicate escarpments along the southern edge of the Plateau.

In addition to transect data, records of butterfly species from the 16 counties of Edwards Plateau which have been reasonably well-collected (Fig. 2), were assembled from data points of the maps of Stanford and Opler (1993), their unpublished supplements for 1994 and 1995, annual summaries of the Lepidopterists' Society and unpublished data kindly provided by Mr. E. Knudson, the Zone 6 coordinator for the Society. The data from these two sources showed 227 butterfly species recorded from this area. Each species was categorized by taxon, geographic element, ecological habitat and range limit. Thirty-seven more species were not recorded east of Brewster Co. and are included in comparative totals and calculations only when appropriate.

To compare the species composition of different geographic areas, I used cluster analysis (a program adapted by Dr. Gao Anli of the University Guelph) to calculate dissimilarity coefficients (D) for each pair of taxa. Dissimilarity coefficients (D) are calculated based on the formula D = 1 - n/N, where n is the number of areas in which both species occur and N those where either or both occur (Holloway & Jardine 1968, Holloway 1977, 1979). Relationships between all pairs can be displayed as a dendrogram, or in this case, as a compact area dendrogram summarizing average linkage between area groups.

I also assessed the relative completeness of these county species inventories. In such inventories it is important to distinguish real differences among regional faunas from artifacts of unequal collecting effort. Amos and Rowell (1988) faced a similar difficulty during assessments of relatively common species of plants in this region. Three approximation methods were applied to assess the scale of the problem:

234

Journal of the Lepidopterists' Society

1.  Records were compared for the 27—33 species of cosmopolitan butterflies with wide ranges through the central and southern United States, most of which should be present in all the 16 counties being assessed on the Edwards Plateau. If less than 24-25 of these had been recorded, then the county was considered as significantly under-collected (see Results).

2.  The theoretical maximum number of species of a fauna in a region of known size can be estimated by fitting appropriate field data to one or more asymptotic models. Early examples were given by Fisher et al. (1943), Preston (1948, 1962) and Simpson (1994). Preston (1962) suggested that the numbers of species was an approximate function of the square root of the area in comparable ecological situations. Later models were applied to regional fauna and flora, especially on islands, in various regions of the world by MacArthur (1965, 1972), MacArthur and Wilson (1967), Raven (1967), Robinson (1975), and Holloway (1977, 1979). Whittaker (1972), Poole (1974) and Pielou (1975) provided useful reviews of these techniques.

Preston (1962) and Robinson (1975) demonstrated a linear relationship between plots of log. numbers against log. surface area, for birds of the West Indies, and the Lepidoptera of several South Pacific islands, respectively. This method is applied to the 16 sample counties of the Edwards Plateau in this study (see Results).

3.  Clench (1979) found that the number of butterfly species recorded over a period of years at a study site in Pennsylvania followed an asymptotic curve when plotted arithmetically against hours (xlOO) spent collecting/observing. He used sequential fits of the constant in the logistic equation to predict the asymptotic value. A similar relationship is shown when the logarithm of species number is plotted against the logarithm of the number of individuals in large samples (Holloway 1979). Frio-Sabinal data were assessed using both these methods (see Results).

Results

Species recorded in the Frio-Sabinal region, measurements of abundance, and new county records. A total of 28,035 specimens representing 100 species was recorded and identified in the upper Frio-Sabinal region during 1988-96 (Table 1). These included eight new records for Bandera County, 10 for Uvalde County and 47 for Real County (indicated by bold initials B, R or U in Table 1). In this context "new county record" is defined as a species not listed for Real, Bandera or Uvalde Counties by Stanford and Opler (1993), or in the 1994 Supplement and Addenda appended to that volume.

Records of 51 more species were accumulated by other lepidopterists during the same period (Table 2). Almost all were from the vicinity of

Volume 52, Number 3

235

Concan (387 m) on the south-central edge of the canyon zone of the Plateau, in north-central Uvalde County, about 20 km south of the sites in the present study. Twenty-six of these species have been plotted by Stanford and Opler (1993). The other 25, indicated by bold initials in Table 2, are from annual summaries of the Lepidopterists' Society News (1988-95) and unpublished data (Mr. E. C. Knudson, in lift.). A few of these were also briefly noted by Stanford and Opler in the unpublished 1994 supplemental list to their 1993 Atlas.

Species totals for the three upper Frio-Sabinal counties are now: Bandera 86 (77), Real 98 (49) and Uvalde 145 (109). The parentheses show counts given by Stanford and Opler (1993). Numbers for Real and Uvalde also significantly exceed those projected (70 and 110 respectively) by Stanford and Opler in their unpublished 1995 supplemental list to the 1993 Atlas.

Relative abundance and absences in the upper Frio-Sabinal region, 1988—96. The five most abundant butterflies at Frio-Sabinal were Nathalis iole Bdv. (7613; 27.15%), Battus philenor (L.) (6397; 22.82%), Colias eurytheme Bdv. (1654; 5.90%), Eurema nicippe (Cram.) (1516; 5.41%) and Strymon melinus Hbn. (1062; 3.79%). At the other extreme, 29 species were represented by only 1—5 specimens. A plot of the abundance of each species for the Frio-Sabinal fauna shows a typical monotonically decreasing profile (Fig. 3a), exaggerated by the large proportion comprised of N. iole and B. philenor. Figure 3b illustrates the profile of the abundance curve when the exaggerated influence of those two species is removed.

A number of species expected for this region were never encountered, despite careful examination of sightings and samples of superficially similar taxa. These include:

1.  Heraclides thoas (L..). I examined the valvae of 43 males netted and released in April—June and September to November. All were H. cres-phontes Cram, and of these, 15 had the forewing markings similar to those often cited as diagnostic of thoas; spot shape seems to be an unreliable criterion for cresphontes in this region. The absence of this species in the southern Hill Country is supported by E. K. Knudson (in lift.) who had never encountered it during intensive searches and collecting in Uvalde County.

2.  Despite the large numbers of small yellows examined, no Eurema nise (Cramer) or E. daira (Godart) were found, only E. lisa in various forms and sizes and the seasonal forms of N. iole.

3.  The presence of the widespread Phycioides tharos (Drury) was anticipated, but not recorded; only P. phaon (Edw.) and P. vesta (Edw.) were moderately common.

4.  Pieris rapae (L.) was absent from all sites in all years, despite

TABLE 1. Table 1. Butterfly records from the Frio-Sabinal region of the Edwards Plat Durden, 1982 (see text for elaboration): A = abundant, C = common, U = uncommon, county records: B—Bandera; R—Real, U—Uvalde. Asterisks for Pyrgus indicate approxim ined in each period. Percentages do not add to 100 due to rounding.

Hesperioidea Hesperiidae

1.  Chiodes zilpa (Butler) R

2. Achalarus toxeus (Plotz) R

3.  Thorybes hathyllus (Smith) U

4.  Thorybes pylades (Scd.) B

5.  Gorgythion begga (Kby) R

6.  Systasea pulverulenta (Fid.) R

7.  Gesta gesta Evans R

8.  Erynnis juvenalis (F.)

9.  Erynnis horatius (Scd. & Bg.)

10.  Erynnis funeralis (Scd. & Bg.) R

11.  Pyrgus communis (Grote)

12.  Pyrgus albescens (Plotz)

13.  Pyrgus philetas Edw.

14.  Celotes nessus (Edw.)

15.  Pholisora catullus (F)

16.  Nostra julia (Freeman)

17.  Lerema accius (Smith) R

18.  Copaeodes aurantiacus (Hew.)

19.  Copaeodes minimus (Edw.)

20.  Hylephila phyleus (Drury)

21.  Hesperia viridis (Edw.)

22.  Wallengrenia otho (Smith)

23. Atalopedes campestris (Bdv.) B

Table 1. Continued.

Years        1988         1989

Species Mar.         May

L990 Sep.

1991 Jul.

1992 May

L992 Oct.

1993

1994

May

1 J

24. Anatrytone logan (Edw.) R

25.  Euphyes vestris (Bdv.) R

26. Amblyscirtes nysa Edw.

27. Amblyscirtes celia Skin. R

28.  Panoquina ocola (Edw.) R

29.  Lerodea eufala (Edw.) Papilionoidea

Papilionidae

30.  Battus philenor (L.)

31.  Battus polydamas (L.) U

32.  Papilio polyxenes F. R

33.  Heraclides cresphontes Cram.

34.  Heraclides ornythion Bdv. U

35.  Pterourus glaucus L. B,R>~U

36.  Pterourus multicaudatus Kby.

37.  Pterourus troilus L. R

38. Appias drusilla (Cram.) R,U

39.  Pontia protodice (Bdv. & Let.)

40. Ascia monuste (L.) R

41.  Colias eurytheme Bdv. R

42.  Colias philodice Godt. R

43.  Zerene cesonia (Stoll.) R

44.  Pheobis sennae (L.) R

45.  Phoebis philea (Johansn.) R

46.  Phoebis agarithe (Bdv.) R

47.  Kricogonia lyside (Godt.) R

48.  Eurema lisa Bdv. & Let.

49.  Eurema nicippe (Cram.)

50.  Eurema mexicanum (Bdv.) R

U

TABLE 1. Continued.

51. Nathalis iole Bdv.                                   — CA!UACUA Lycaenidae

52. Atlides halesus (Cram.) B                      ____S— UU

53.  Phaeostrymon alcestis (Edw.) R            — — — — — — S —

54.  Satyrium calanus (Hbn.) R,U               _ _ _ _ U — C S

55.  Calycopis isoheon (Btl. & Drc.)             ____C — U —

56. Mitoura grynea (Hbn.)                      UU — UC — U —

57.  Fixseniafavonius (Smith) R                  _ _ _ _ C — C S

58.  Strymon melinus Hbn.                           — CCUAUCC

59.  Brephidium exile (Bdv.)                         _____ S S _

60.  Leptotes manna (Reak.) B                     ________

61. Hemiargus ceraunus (F.) R                    — — — — — U — —

62.  Hemiargus isola (Reak.)                         —ACCCUCC

63. Everes comyntas (Godt.) B,R                _____U — — Riodinidae

64.  Calephilis nemesis (Edw.)                      — S — — — — — —

65.  Calephilis rawsoni McAlp.                     — — — — — S — — Libytheidae

66.  Lihytheana carinenta (Cram.)               — C— — U — U— Nymphalidae

67. Agraulis vanillae (L.).                             UCCCCCUC

68.  Dryas julia (F.) U                                   — — — S — — — —

69.  Heliconius charitonius (L.).                   — — — S — — — —

70.  Polygonia interrogations (F.)R            — S U — — — — U

71.  Polygonia comma (Harr.) R,U               — — S — — — — —

72.  Nymphalis antiopa (L.) B,R                  — S — — S — S S

73.  Vanessa virginiensis (Dry)                      UC — — CUCC

74.  Vanessa cardui (L.) R                            — UU — UUS—

Table 1. Continued.

Years        1988

Species Mar.

1989

1990 Sep.

1991 Jul.

1992 May

1992 Oct.

1993 May

1994

May

c c

u c

c c

s

A

C

A C

u c

c u

u u

u u

c u

s

c c s c c c c

75.  Vanessa atalanta (L.)                            U       C

76. Junonia coenia (Hbn.) B,R                  —      —

77. Junonia nigrosuffusa B. & MD. R —      —

78.  Euptoieta claudia (Cram.) R                U       A

79. Anthanassa texana (Edw.) R                 S        A

80.  Phyciodes phaon (Edw.)                       —       C

81.  Phyciodes vesta (Edw.)                         —       A

82.  Charidryas nycteis (Dbld.) R              —      — — — — — — —

83.  Chlosyne lacinia (Gey.)                        —       C C — U C —! S

84.  Chlosyne janais (Drury) R                   —      — — — SS — U

85.  Thessalia theona (Men.)                       —      — — — U — — —

86.  Texola elada (Hew.) R                          —      —                                      S

87.  Basilarchia archippus (Crm.) R           —      —

88.  Basilarchia astynax (F.)                       —      —

89. Adelpha bredowii (Gey.)                      —      —

90.  Mestra amymone (Men)                       —      —

91. Anaea andria Scd. R                            —      —

92. Asterocampa celtis (Bd. & Lt.) R         —       S

93. Asterocampa leila (Edw.) R                 —      —

94. Asterocampa clyton (Bd. & Lt.) R —      — Satyridae

95.  Megisto cymela (Cram.) R, U              —      —

96.  Megisto rubricata (Edw.)                     —       C

97.  Cercyonis pegala (F.) R                       —       C

98.  Hermeuptychia hermes (F.) U             —      — — — — — —

Danaidae

99.  Danaus plexippus (L.)                          —       UU — C C U 100. Danaus gilippus (Cram.)                      —       C C U C C C

C C

C A

U

c

u u

u c

c c c

s

u

s u

c c

s

u c c

u c

Table 2. Records of butterfly species in the Frio-Sabinal region from sources other t county records since 1993, following Table 1. Abbreviations: CB: Charles Bordelon, Jr.; EC Karges.

Species no.

Family & species

115.

116. 117. 118. 119. 120. 121.

122.

Pyrgus oileus (L.)

Heliopetes macaira (Reak.) U Heliopetes lavianus (Hewitson) U Ancyloxypha numitor (F.) Ancyloxypha arene (Edw.) Polites vibex (Gey.) Ambylscirtes aenus Edw. U

Ambylscirtes oslari (Skin.)

Uvalde. Uvalde.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Bandera.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Uvalde.

Real, Bandera.

Uvalde Co.

Bandera, Uvalde.

Bandera, Uvalde.

Bandera, Uvalde.

Table 2. Continued.

Species no.

Family & species

County

Ambylscirtes eos (Edw.)

Calpodes ethlius (Stoll.) U Megathymus yuccae (Bdv. & Let.

Paramidea midea (Hbn.)

Anteos clorinde (Godt.) U Anteos maerula (F.) U Eureraa proterpia (F.) U

Chlorostrymon simaethis (Drury) U Ministrymon clytie (Edwards) U

Ministrymon azia (Hew.) R, U

Xamia xami (Reak.) U Incisalia henrici (G. & R.)

Parrhasius m-album (Bdv. & Let.) U

Strymon alea (Godm. & Sal v.) U Strymon columella (F.)

123.

124. 125.

Pieridae

126.

127. 128. 129.

Lycaenidae

130. 131.

132.

133. 134.

135.

136. 137.

Bandera, Uvalde.

Uvalde Co. Concan. Bander, Real, Uvalde.

Uvalde.

Uvalde. Uvalde. Uvalde.

Uvalde. Uvalde.

Real, Uvalde.

Uvalde.

Bandera, Real, Uvalde

Uvalde.

Uvalde. Uvalde.

Table 2. Continued.

Species no.

Family & species

County

138.

Riodinidae

139. 140.

Nymphalidae

141. 142. 143.

144. 145. 146.

147. 148.

149. 150.

Danaidae

151.

Zizula cyna (Edw.)

Calephelis perditalis B & McD. U Caria ino (Godm. & Salv.) U

Dione moneta Hbn. U Euptoieta hegesia (Cram.) U Phycoides tharos (Drury) U

Phyciodes tulcis (Bates) U Anartia jatrophae (Johansn.) Siproeta stelenes (L.)

Eunica monima (Stoll.) U Dynamine dyonis Gey. U

Biblis hyperia (Cram.) U Anaea aidea (Guer.-Men.) U

Danaus eresimus (Cram.)

Uvalde.

Uvalde. Uvalde.

Uvalde. Uvalde. Bandera, Real, Uvalde

Uvalde Co. Uvalde Co. Bandera, Uvalde.

Uvalde. Uvalde.

Uvalde. Uvalde.

Uvalde.

Volume 52, Number 3                                                                             243

8000 T

7000

6000

CO

c

j= 5000

a

CO

■g 4000

ai

E 3000

3

z

2000

1000

0

r-mo)Cor^T-ioa>eor--'>-ioo)cor^-i-io05cor^T-tno>cor-i-

T-i-CMOJCMCOC3'<tTl--*if)lO<DCDCDr^-r-OOC00005a)0

Number of species

Fig. 3. Abundance versus species of butterflies recorded on transects from the upper Frio-Sabinal region 1988-1996, showing a typical monotonically decreasing profile (3a). 3b (inset) shows the profile with the two commonest species, N. iole and b. philenor removed.

scrutiny of all white butterflies netted or observed at each locality, even in the cultivated areas around Utopia and Leakey townships.

5.  Among the Hesperioidea, specimens of Erynnis Shrank and Pyrgus Hbn. were netted and examined whenever possible to minimize the chance of overlooking similar species. P. philetas Edw. was found to occur in the hotter, later part of the year at some sites, but all Erynnis examined proved to be E. horatius (Scudd. and Burg.) or E. funeralis (Scudd. and Burg.), with the exception of a single E.juvenalis (F.) taken at the Blanket Creek in March 1988.

6.  The apparent absence of Polites vibex (Gey.) was also noteworthy because it has been recorded previously in Bandera and Uvalde counties, as well as Kerr and Medina (Stanford and Opler 1993). Occasionally, males may have been confused with Hylephila phyleus (Drury) on transects, or females overlooked among clusters of the more common Atalopedes campestris (Bdv), even though we were actively looking for P. vibex.

Assessment of relative completeness of county inventories. Records of butterfly species for counties on the Edwards Plateau range from 170 in Bexar, to only two in Midland (Stanford and Opler 1995 un-

3a

Table 3. Counts of butterfly species from well-collected counties of the Edwards Pla assigning butterfly species to a particular range. *"Tom Green" is a conglomerate based o (providing a provisional estimate for the central northern region of the Plateau).

Volume 52, Number 3

245

11} 2 \

Real 1996 W Bandera

"Tom Green" Conglomorate

2.5          2.6          2.7          2.8          2.9

Log. of relative county unit area

FlG. 4. Log-log relationship between total species recorded in each county versus area in square kilometers, assuming similar ecological parameters (after Preston 1962). Values on or near the line are presumed to be close to the "maximum". Counties falling below the line are probably under-collected, but may have less diverse habitat.

publ. supp.). Of course Midland is an extreme case, but unequal collecting effort may still be an important, but overlooked, factor when relatively high species totals are being compared. The cosmopolitan butterfly species were summed for each of the 16 sampled counties of the Edwards Plateau. Thirteen had records of 27—33 such species and three had 17-25, with a mean of 28.8 (Table 3). In the case of Real Co., 49 species in total had been recorded by 1993, of which 23 were cosmopolitan. These might seem to be appropriate numbers, because Real is one of the smallest counties on the Plateau, but Stanford and Opler's (1993) data base showed Vanessa cardui, V. virginiensis, /. coenia, A. texana, F.favonius and E. vestris still unrecorded, so Real was obviously under-collected. During the present survey, the species total for Real has risen to 98, with a cosmopolitan element of 29.

When the data on number of species per county from the Edwards Plateau counties are log-transformed and plotted against log(county area) (Fig. 4), the well-collected counties such as Bexar, Brewster, Comal, Kerr, Travis and Uvalde fall close to the line, while most of the others counties fall well below it. The recent data from the Frio-Sabinal surveys have shifted the values for Bandera and more markedly, Real Co., closer to the line in Fig. 4. This data indicate that the theoretical "maximum" species total for both these counties could be 160-165 species, if ecological parameters and migratory patterns are approximately comparable across the Plateau (see Discussion).

246

Journal of the Lepidopterists' Society

£ 90

eo co

£ 80 •a

i 70

o

d>

: 6o + «

o

a 50

10

! 40 JS

| 30

c

0)

£ 2 0 I 10 -f

,•■1991 ■1990

■1995

*1996

■1992

0                                50                             100                            150                            200                            250                            30(

Cumulative hours of recording In Frio-Sabinal region 1988-1996

FlG. 5. Clench (1979) extimated "maximum" number of species for a by plotting cumulative number of species against cumulative sampling time. The curve for Frio-Sabinal seems asymptotic after 8—9 years of recording.

When the Edwards Plateau data are plotted following the models of Clench (Fig. 5), and Holloway (Fig. 6) however, the asymptotic curves indicate that the "maximum" number of species in Bandera and Real counties might be only 110—120, not 160—165.

Linkages across the Edwards plateau. Species totals for Edwards (47 species), Pecos (52), Terrell (68), Kinney (73) and Kimble counties

.-■•*

10                                              100                                            1000

Log. of numbers of specimens/ species in samples

Fig. 6. Holloway (1979) plotted log species number against log number of specimens in large samples. The asympototic curve produced from the Frio-Sabinal data is similar to Fig. 5.

Volume 52, Number 3

247

COUNTIES & SITES*

Rescaled cluster distance ___1__________l__________I___

UVALDE

Concan

BEXAR

TRAVIS

Barton Ck

COMAL

KERR

REAL

Frio-Sabinal

BANDERA

MEDINA

VAL VERDE

KINNEY

KIMBLE

PECOS

TERRELL

EDWARDS

BREWSTER

.50

.60

.70

Fig. 7. Dendrogram showing provisional interrelationships of the butterflies of the Edwards Plateau. Dissimilarity coefficients were calculated for each pair of taxa of the total species sample. The butterfly fauna in the southeastern cluster of counties is more cohesive than those in the west. The relative isolation of Brewster County was anticipated, but that of Edwards may be a function of under-collecting.

(62), are likely to be too low, although all but the last range from relatively dry to arid (Riskind and Diamond 1988). The estimate for Tom Green county (85) is a composite (see Table 3 caption) and should be interpreted with special caution. Totals for other counties (78 to 170 species), i.e., Bandera, Bexar, Brewster, Comal, Kerr, Real, Travis, and Uvalde counties are not likely to change by more than 5—20% in the near future. Projecting totals by weighted averages is risky without good information.

The dendrogram (Fig. 7) shows the dissimilarity coefficients for pairs of counties or groups of counties. Lower numbers (a lower dissimilarity coefficient) indicate a greater degree of overlap in the butterfly fauna of those regions. This dendrogram demonstrates the relative distinctiveness of Brewster Co., even from Pecos and Terrell counties, despite their having some faunal components in common. Edwards appears almost as isolated as Brewster. The general faunal cohesion of the southeastern block of counties, Travis, Comal, Bexar, Kerr and Uvalde is confirmed, as is the anticipated similarity of Real and Bandera counties. The

248

Journal of the Lepidopterists' Society

FIG. 8. N-NW species as a proportion of the butterfly fauna of 16 sampled counties on the Edwards Plateau.

Barton Creek site is shown to be usefully representative of the fauna of Travis, as are Concan of Uvalde Co., and Frio-Sabinal of Bandera and Real counties. Remaining counties also appear relatively isolated but interpretation is best reserved until more information is available.

A frequency count revealed that only 37 species were known to occur in all but one to three of the counties, while 114 of the 261 had been recorded in only one to three counties. Among families, restricted distribution was most apparent in the Hesperiidae; 54% and 57% of all Pyrginae and Hesperiinae, respectively were restricted to three or fewer of the sample counties, generally in the western sector of the Plateau.

Comparative distributions of butterfly species on the Edwards plateau. The data matrix was used to show all species classified by: A) Geographical components by counties (Table 3), B) Species in ecological habitats by counties (Table 4) and taxon (subfamilies) (Table 5); C) Relative position by county with respect to overall range (Table 6). All data are given as numbers, not percentages.

Comparison of geographical components. Butterfly species were allocated to one of five groups based on their geographical distribution relative to the Edwards Plateau (Table 3): 1) N/NW-species which have predominantly California-Rockies-Great Plains distributions (Fig. 8); 2) NE/E-prairie-eastern coastal woodlands (Fig. 9); 3) S-tropical coastal forest (Fig. 10); 4) SW/W (Sonoran-Coahuilan) (Fig. 11), and 5) COSM-cosmopolitan species, or ranging at least through Mexico- Texas-central Gulf states (not figured). Most attributions are from Durden (1982); species not recorded by him are by my allocation. Data were extracted from Stanford and Opler (1993), and their unpublished supplements of 1994, 1995. Accounts by Howe (1975), Pyle (1981), Scott (1986), Opler

TABLE 4. Numbers of butterfly species from 16 selected counties across the Edwards Plateau (1982). Parenthetical values for Brewster County include 37 species not yet recorded east of this

Table 5. Numbers of butterfly subfamilies from 16 selected counties across the Edw Durden (1982). The 37 species recorded from Brewster or counties further west are exclud

Habitat vs. taxon                      Pyg-           Hesp.         Megath.         Papil.          Pierid.         Liphyr.         Thecl.         Polvom.         Riod.

A.  Disturbed sites            000070000

B.  Great Plains flora         150100110

C.  Dry warm temp.

—subtropical              10         61           130212

D.  Subtropical

thorn forest                  233010221

E.  Subtrop. thorn

scrub/desert                  410000300

F.  Oak/Jun. Sonor.

grass/woodland             010000100

G.  Broad temp.—

trop. woodland             210110000

H. Eastern forest             000000010 I. Gulf Coast

trop. woodland             100000000 J. Subtropical &

Gt. Plain brush             001010200

K. Subtrop. mont.

woodland                       450            100            100

L. Eastern decid.

woodland                       390            100            200

woodland                       350            110            000

woodland                     16          10            480            120

TOTALS                         53 38         6          12 25         1          18         8          7

M. Appalach.-Miss.

Basin woodland             400            011            110

N. Southern coast

O. Western decid.

woodland                       200            120            000

P. Endemics of S.

Madre/Balcones            000            000            102

Q. Endemics of C.

Texas/Coahuila              011            000            102

R. South tropical

Table 6. Numbers of butterfly species in 16 selected counties across the Edwards Plate Durden (1982).

252

Journal of the Lepidopterists' Society

FlG. 9. N-NE species as a proportion of the butterfly fauna of 16 sampled counties on the Edwards Plateau.

and Krizek (1984), Opler and Malikul (1992) and Neck (1996) were consulted for clarification of habitat and eastward range, when necessary.

The distribution of each regional geographical component is presented as the proportion of total species recorded for each county. Paired t-tests indicated that the variances of samples were not significantly different from those corrected for relative area (p = 0.37-0.51). Significant differences between the geographical groups were confirmed by AN OVA (p < 0.01). The N/NW element (0.06-0.17) is uniformly weak across the Plateau (Fig. 8). NE/E species account for 0.25-0.33 of each total in most of the eastern and central region, falling to 0.11-0.13

FlG. 10. S species as a proportion of the butterfly fauna of 16 sampled counties on the Edwards Plateau.

Volume 52, Number 3

253

FIG. 11. W-SW species as a proportion of the butterfly fauna of 16 sampled counties on the Edwards Plateau.

in the west (Fig. 9). The S (tropical) element is strongly represented only in the southeast region of the Plateau (Fig. 10), while the W/SW (Sonoran-Coahuilan) element (Fig. 11) is predominant (0.21-0.47) in the western and southwestern counties, but 0.16 or less in the central and eastern regions.

Comparison by habitat zones. Butterfly data from the 16 counties were tabulated by subfamily against 17 habitat types designated by Dur-den (1982) by county (Table 4) and by taxon (Table 5). Some interesting patterns were noted, which do not require statistical analysis.

Type A: Disturbed habitats were typified by 11 species but dominated by Pieridae (7 species-64%).

Type B: Great Plains grassland habitat supported 16 species on the Plateau; Hesperiinae (5) and Satyrinae (4), together comprising 56% of the total.

Types C-E: The intergrading dry, warm subtropical habitats, Arid woodland (C), Thorn forest (D) and Thorn scrub (E) together accounted for 58 of the total species, dominated by Hesperiidae (29), comprised of Pyrginae (16), Hesperiinae (10) and Megathyminae (3). Together with Theclinae (7) and Melitaeinae (8), these taxa accounted for 75% of the total.

Type J: In the subtropical Great Plains brush habitat nine species were recorded; four (44%) were Melitaeinae.

Type K: In Subtropical montane woodlands 12 species were represented, Hesperiidae; (9, 75%), Pyrginae (4) and Hesperiinae (5).

Type L: Eastern deciduous woodland habitat accommodated 20 species of the total; (9, 45%) were Hesperiinae.

254

Journal of the Lepidopterists' Society

Type M: Basin woodlands were characterized by 10 species of which four (40%) were Pyrginae.

Type N: Southern coast woodlands fauna comprised 12 species, again dominated by Hesperiidae (67%-Pyrginae 3, Hesperiinae 5).

Type R: The largest faunal group of butterflies was that associated with South tropical woodlands, 49 species, mostly sporadic in their appearance in any given county, and dominated by Pyrginae (16, 33%), subtropical Pieridae (8, 16%) and subtropical Nymphalinae, Apaturinae and Heliconiinae (10, 20%).

Types F-I and O-Q (see Table 5 and Durden 1982): the remaining seven habitat types were sparsely characterized by 1—8 species (mean 5.7) scattered across the range of families and subfamilies.

Overall, the Hesperiidae was the best represented family in the 16 counties, (97, 43% of the total), while the Pyrginae was the most abundant subfamily (53, 23.5%).

Comparisons of range limits. The diversity of the butterfly fauna of the Edwards Plateau is largely a function of its strategic position (see above); but despite a respectable total of 227 species for the 16 counties, even excluding those found only in Brewster or westward, it is surprisingly difficult to point to any one category, other than the Eastern deciduous woodland group which could truly be said to be characteristic of the geographic unit.

Of the 11 cosmopolitan species associated with disturbed habitats (Durden's Category A), only one (9%), the northern straggler Colias philodice, reaches its southern limit of range limit on the Plateau (Table 6). Of the 15 species of the Great Plains fauna (B), five (33.3%) are at their southern or eastern limit, while only five (25%) of the Eastern deciduous woodland species (L) are at their S or SW limit (Table 6).

However, the number of species which reach a range limit on the Plateau comprise more than 50% in all remaining ecological habitat groups (Table 6); C (50%), D and E (73%), F (66.6%), G (86%), H and I (100%), J (55%), K (83%), M (80%), N (91.5%), O (62.5%), P (100%), Q (60%) and R (57%). The implication is that a moderate change in one or more climatic factors could have a significant effect on present species composition. Given the altitude of much of the region (300 m+), an extended period of cooler climate might eliminate many of the southern components, upwards of 60—100 species. Warming trends might seem more likely at present, but the effect would perhaps be more variable, depending whether seasonal and annual rainfall decreased or not.

Discussion

Location and environment. Much of the Edwards Plateau is a southward extension of the Great Plains Physiographic Province (Hunt

Volume 52, Number 3

255

1974), and the southern extremities of the Rocky Mountains are only some 200 km from its northwestern margin. Altitudes ranges from about 900 m in northern counties such as Tom Green, to 300 m at Concan in north-central Uvalde Co. on the southern extremity of the Balcones Escarpment. The western section of the Plateau blends into the arid, subtropical Trans-Pecos region. The major rivers and associated riparian systems generally run northwest to southeast; their ecological influence is widespread and particularly significant in the southeast and less so in the southwest. The counties of the Edwards Plateau are on hard, porous limestone (Gould 1969, Ajilvsgi 1984), except where the central, basaltic Llano Uplift breaks through. In contrast, on the eastern margin of the Plateau, counties such as Bastrop, Lee, Fayette and most of Caldwell have sandstone and shale soils, with podsols, peat bogs, and Neogene sands and clays covered with coastal-type prairie (Durden 1982).

Isohyets over western and central Texas have a consistent general north-south orientation (Riskin & Diamond 1988), with a marked clinal decrease in annual rainfall from east to west. As a consequence, Brewster, Pecos and Terrell counties have annual rainfalls of only 20—35 cm. Eastern and northeastern counties receive more than twice that (ca. 90 cm). Significant differences in rainfall occur even across the south-central region of the Plateau and along the complex southern margin of the exposed limestone scarp of the Balcones Fault. The upper Frio-Sabinal area receives about 30% less rain annually (60 cm) than Austin (85 cm).

Floristic patterns on the plateau. The geography, physical environment and climatic characteristics of the Plateau have all played roles in colonization and establishment of distribution patterns of the historical and extant flora and fauna. Amos and Rowell (1988) applied principal component analysis (PC A) to a large database of woody and endemic vascular plant records, to try to determine the ecological relationships, floristic patterns, and perhaps origins of the regional components. The analysis identified two major floristic zones across the Plateau: a numerous and diverse eastern component and a less rich and less widespread western one. Val Verde County was identified as the major transitional zone. A third, southern element of Neotropical origin was important only in the counties along the southern rim of the Plateau. Simple regional characterizations are complicated by the existence of parallel ecological habitats within each floristic zone (Gehlbach 1988). For example, the woody taxa of the Plateau tend to form tight geographical clusters (Amos and Rowell 1988), while in contrast, the vascular endemics, are clustered in a relatively few eastern counties without any such detectable associations.

Flora in adjacent zones at the same latitude across the Plateau can be surprisingly dissimilar. Using data of Ajilvsgi (1984) and Enquist (1987),

256

Journal of the Lepidopterists' Society

I examined a sample of 313 vascular plants from about 70 families (not just endemics), and compared the composition of the south-central Edwards Plateau (Real, Bandera, Uvalde and Kerr counties) with that of the South Central Vegetation Zone, the westerly section of which takes in parts of Travis, Hayes, Comal and Bexar counties on the curved margin of the Balcones Fault between San Antonio and the Austin region. Although 148 species were common to both regions, 52.7% of the flora was not shared; 73 species in South Central are not known from the Edwards Plateau and 92 found on the Plateau are not recorded from South Central.

North-south differences in flora can be quite marked over relatively small distances as well. The upper Frio-Sabinal vegetation consists predominantly of Juniper-Oak savanna (Amos and Gehlbach 1988). The northern, eastern and western margins do not include any transitional areas, but intergrade with Mesquite Savanna in the north-central parts of Uvalde and Medina counties along the southern margin of the Balcones Fault (Simpson 1994).

The eastern counties have much more complex vegetation patterns. Llano County and the western part of Burent lie in the Central Texas Mesquite-Oak Savanna; Hays, Travis and Williamson straddle the southern tongue of the Blackland Prairie; while the eastern counties of Lee, Bastrop, Caldwell and Fayette are in the Post Oak Savanna zone (Simpson 1994).

The origin of the small but widespread eastern deciduous element on the Plateau remains an enigma. Amos and Rowell (1988) conclude that the following explanation remains the most likely; i.e., that it represents an isolated remnant of an earlier cool climate forest which extended into and through the Gulf states. The distributions of groups of endemic plants, often taxonomically related within each cluster, seems to imply that the Plateau had several long periods of climatic and ecological stability during the Tertiary.

These clusters of floristic endemics are not matched by corresponding distribution patterns among butterfly taxa, which show little endemism in the region (see Durden 1982). As a general rule, winged insects would seem to have greater potential for mobility and rapid colonization than most plants with distributions restricted to certain ecological regimes.

On the other hand, there is marked correspondence between components in the butterfly fauna and large-scale geographical groupings in the regional floras. In the butterflies, a western-southwestern element is strong in Brewster, Pecos and Terrell and Kinney counties, but declines by nearly 50% to the north and east (Fig. 11). This equates with the Far West vegetation zone of Gould (1969) and Ajilvsgi (1984). The subtrop-

Volume 52, Number 3

257

ical south-southeastern element makes up 25—30% of the butterfly species in some southeastern counties (Fig. 10), but is almost insignificant in the northern counties of higher altitude and in the western counties, most of which have limited riparian habitat. This component parallels the mixed South Central and Coastal Texas vegetation zones. The northeast-eastern butterfly component contributes about 25-33% to the total species in the eastern and northern counties (Fig. 9), and can be equated to the South Central vegetational zone. Finally, the butterfly element of largely northwestern distribution, relatively weak in all sampled counties (6—17% of total species, Fig. 8), is not just equivalent to the Panhandle vegetation zone of Gould (1969) and Ajilvsgi (1984) but has its origins beyond the Plateau, as part of the large western Woodlands and Chaparral zone. Amos and Rowell (1988) also stressed that the NW region of the Edwards Plateau was floristically allied with the High Plains, not the rest of the Plateau.

For much of the year, central Texas is usually hot and dry, conditions to which many Hesperiidae are well adapted, but not so many other butterfly taxa. The relative species richness of the south-central and eastern Edwards Plateau is supported by the extensive rivers that flow south and southeast from the high country to South Texas, and the variable riparian habitat associated with them. The importance of healthy riparian systems to butterfly diversity and conservation in two other semi-arid warm temperate regions has been previously noted by Yela (1992: Central Spain) and Gaskin (1996: Greece). This important habitat also facilitates the frequent penetration of southern tropical and subtropical species of Lepidoptera into the relatively northern and eastern counties of the Plateau (Fig. 12), even when wind directions are not particularly favorable.

Geographical range, climatic zone and general adaptation to a particular ecological regime are useful characteristics by which we can classify faunal components such as butterflies. In most cases, however, a significant fraction of any county record consists of a long "tail" of occasional visitors, non-resident species, and the updated Frio-Sabinal list is no exception (Tables 1, 2). There is a natural tendency to focus on the new and unusual, but this can divert attention from the species which are important, consistent components of local ecosystems. Each individual species, sub-species and often, population, needs a suite of natural resources to survive; a favorable temperature range, suitable food plants for oviposition, nutrition for larvae and adults, relatively safe perches for adults at night, and sites for pupation and whichever instar over-winters. Each stage of the life history is vulnerable to a range of density dependent and independent events over which individuals have relatively little control.

258                                                     Journal of the Lepidopterists' Society

Fig. 12. Major landform features of the greater Edwards Plateau (sensu Gould 1969) and associated structures in the trans-Pecos region. Decline in altitude to the S and SE, together with the relative abundance of rivers and riparian habitat in that region probably facilitates butterfly dispersal and seasonal migration northwards.

We need to understand the potential importance of rather subtle factors. Durden (1982) for example, noted that the tropical elements of the butterfly fauna of south Texas and northern Mexico are most commonly encountered in the Austin region when the daily photoperiod is shortest. Not all his rarities found around Austin were dated, but at least 11 unique or scarce records of southern species were reported between late September and November (Durden 1982).

Consistent differences in local climatic conditions can exert effects important over relatively short distances. For example, there is a general similarity between the three intensively worked sites in this article; Barton Creek, Concan and upper Frio-Sabinal. Yet altitude in most of the Frio-Sabinal region exceeds that of the Austin Barton Creek site studied by Durden (1982) by about 100-300 m, providing somewhat cooler, windier weather for part of the year, especially at night. Concan, in north-central Uvalde is right on the southern edge of the escarpment, and about 150 m lower in altitude than my sites, with corresponding higher humidity and somewhat different soils and drainage. The relative frequency of dispersive weather systems over the three sites may also vary because of the different isohyet regimes and topographies.

Volume 52, Number 3

259

Such variation can have important effects on microhabitats and conservation of semi-isolated butterfly populations. For example, spacing of trees and amount of understory can differ in subtle ways that do not easily yield to simple measurement. At one site in central Europe, for example, Kudrna (1993) discovered that a population of a Parnassius species was in serious decline because the degree of shading of the food plant in the understory had slowly increased to a point where the thermal regime was too cool for the larvae to complete their metamorphosis within the available growing season.

Even when resources for adult and larval butterfly species visibly occur throughout a region, spacing and density of these host plants can differ from one site to another. Such variation may be apparent to searching butterflies, but not necessarily to entomologists. There are many more mysteries on the Edwards Plateau still waiting to be solved.

Acknowledgments

This study was made possible by the tolerance, interest and assistance of Mr. R. and Mrs. B. Garrison of the Indian Blanket Ranch on RR# 1050, west of Utopia, Texas; Mr. R. and Mrs. D. Rudolph of Houston, owners of the 'Rio Lindo' property in Leakey, Texas; Mr. R. and Mrs. M. Davies of Twin Forks Estate near Leakey; and Mr. A. and Mrs. L. Sharpe, of Rio Frio Bed and Breakfast, often assisted with transport, accommodation problems, delayed mowing schedules and gave many small kindnesses. Mrs. E. A. Littler (Michigan Entomological Society) made all her data from the region available to me, and both she and Ms. L. Dueck (formerly of the University of Guelph, Ontario, Canada) gave invaluable assistance with field studies, often in oppressive weather conditions, dodging ever-hungry ticks. This study was partially supported by funding from the Natural Sciences and Engineering Council of Canada, through Operating Grant A5863.

Literature Cited

AjlLVSGl, G. 1984. Wildflowers of Texas. Shearer Publ., Fredericksburg, Texas. 414 pp.

Amos, B. B. & F. R. Gehlbacii (editors). 1988. Edwards Plateau vegetation: plant ecological studies in central Texas. Baylor University Press, Waco, Texas. 144 pp.

AMOS, B. B. & C. M. Rowell, Jr. 1988. Floristic geography of woody and endemic plants. In B.B. Amos & FR. Gehlback (eds.), Edwards Plateau vegetation: plant ecological studies in central Texas. Baylor University Press, Waco, Texas.

CLENCH, H. K. 1979. How to make regional lists of butterflies: some thoughts. J. Lepid. Soc. 33:216-231.

COLLINS, M. M. 1991. Speciation: a review of concepts and studies with special reference to the Lepidoptera. J. Res. Lepid. 30:45—81.

DuRDEN, C. J. 1969. Ecological aspects of the species concept. Armadillo Papers 1:1—14.

----------. 1982. The butterfly fauna of Barton Creek Canyon on the Balcones fault zone,

Austin, Texas, and a regional list. ]. Lepid. Soc. 36:1—17.

EllRLICH, P. R. & D. M. MURPHY. 1981a. Butterfly nomenclature: a critique. J. Res. Lepid. 20:1-12.

----------. 1981b. Nomenclature, taxonomy and evolution. J. Res. Lepid. 20: 198-204.

----------. 1982. Butterflies and biospecies. J. Res. Lepid. 21:219-225.

ENQUIST, M. 1987. Wildflowers of the Texas hill country. Lone Star Botanical, Austin, Texas. 275 pp.

Fisher, R. A., A. S. Corbet & C. B. Williams. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. J. Anim. Ecol. 12:423-458.

260

Journal of the Lepidopterists' Society

FREEMAN, H. A. 1951. Ecological and systematic study of the Hesperioidea of Texas (Lepidoptera: Hesperioidea). Southern Methodist University Studies 6:1-67.

Gaskin, D. E. 1996. Hesperioidea and Papilionoidea of the Ionian island of Kefalonia, Greece: Additional species and a review of faunal components. Nota Lepid. 18: 213—223.

Gehlbach, F. R. 1988. Forests and woodlands of the northeastern Balcones escarpment, pp. 57-77. In B.B. Amos & F.R. Gehlback (eds.), Edwards Plateau vegetation: plant ecological studies in central Texas. Baylor University Press, Waco, Texas.

GOULD, F. W. 1969. Texas plants—a checklist and ecological summary. Texas Agricultural Experiment Station, MP-585.

HAMMOND, P. C. 1985. A rebuttal to the Arnold classification of Speyeria callippe (Nymphalidae) and defense of the subspecies concept. J. Res. Lepid. 24:197-208.

----------. 1990. Patterns of geographic variation and evolution in polytypic butterflies. J.

Res. Lepid. 29:54-76.

HOLLAND, W. J. 1930. The butterfly book. Doubleday & Company, Garden City, New York. 2nd edition. 424 pp.

HOLLO way, J. D. 1977. The Lepidoptera of Norfolk Island, their biogeography and ecology. Series Entomologica 13, vi + 291 pp., Dr. W. Junk B.V. Publishers, The Hague, Netherlands.

----------. 1979. A survey of the Lepidoptera, biogeography and ecology of New Caledonia. Series Entomologica 15, xii + 588 pp., Dr. W. junk B.V. Publishers, The Hague, The Netherlands.

HOLLOWAY, ]. D., & N. Jardine 1968. Two approaches to zoogeography: a study based on the distributions of butterflies, birds and bats in the Indo-Australian area. Proc. Linn. Soc. Lond. 179:153-188.

Howe, W. H. (Ed.). 1975. The butterflies of North America. Doubleday & Company, Garden City, New York. 633 pp.

HUNT, C. B. 1974. Natural regions of the United States and Canada. W. H. Freeman and Co., San Francisco, California. 725 pp.

Kendall, R. O. 1964. Larval foodplants for twenty-six species of Rhopalocera (Papilionoidea) from Texas. J. Lepid. Soc. 18:129—157.

----------. 1976. Larval foodplants and life history notes for some metalmarks (Lepidoptera: Riodinidae) from Mexico and Texas. Bull. Allyn Mus. 32:1-12.

KLOTS, A. B. 1951. A field guide to the butterflies of North America, east of the great plains. The Peterson Field Guide Series, Houghton Mifflin Company, Boston. 349 pp.

KUDRNA, O. 1986. Butterflies of Europe. Vol. 8. Aspects of the conservation of butterflies in Europe. AULA-Verlag, Wiesbaden, Germany. 323 pp.

----------. 1993. Conservation of butterflies in central Europe. A report to the British Butterfly Conservation Society at "Ecology and Conservation of Butterflies, an International Symposium". 10-12 September 1993, Keele University, Staffs, England.

MACARTHUR, R. H. 1965. Patterns of species diversity. Biol. Rev. 40:510-533.

----------. 1972. Geographical ecology: patterns in the distributions of species. Harper, and

Row, New York, New York. 269 pp.

MACARTHUR, R. H. & E. O. WILSON. 1967. The theory of island biogeography. Princeton University Press, Princeton, New Jersey. 203 pp.

McGuiRE, W. W. 1982. Notes on the genus Hesperia in Texas: temporal and spatial relationships. Bull. Allyn Mus. 73:1-21.

McGuiRE, W. C. & M. A. RiCKARD. 1974. An annotated checklist of the butterflies of Bentsen-Rio Grande Valley State Park and vicinity. Texas Parks & Wildlife Department, Mission, Texas.

MILLER, L. D. & F. M. BROWN. 1981. Butterfly taxonomy: a reply. J. Res. Lepid. 20: 193-197.

NECK, R. W. 1996. A field guide to the butterflies of Texas. Texas Monthly Field Guide Series, Gulf Publishing Company, Houston, Texas, 323 pp.

OPLER, P. A. 1993. Conservation of butterfly diversity in North America [abstract], p. 8 In A. Pullin, (ed.), Ecology and conservation of Butterflies: an international symposium organized by Butterfly Conservation, Keele University, United Kingdom, 10-12 September 1993.

Volume 52, Number 3

261

Opler, P. A. & G. O. KRIZEK. 1984. Butterflies east of the great plains: an illustrated natural history. Johns Hopkins University Press, Baltimore, Maryland. 294 pp.

Opler, P. A. & V. Malikul. 1992. Eastern butterflies. The Peterson Field Guide Series, Houghton Mifflin Company, Boston, Massachusetts. 396 pp.

PlELOU, E. C. 1975. Ecological diversity. Wiley Press, New York, New York.

POLLARD, E. 1977. A method for assessing changes in abundance of butterflies. Biol. Conserv. 12:115-134.

----------. 1991. Monitoring butterfly numbers, pp. 87-11. In F.B. Goldsmith (ed.), Monitoring for conservation and ecology. Chapman & Hall, London.

POLLARD, E. & T. J. Yates, 1993. Monitoring butterflies for ecology and conservation. Chapman & Hall, London. 274 pp.

POOLE, R. W 1974. Quantitative ecology. McGraw Hill, New York, New York. 532 pp.

PRESTON, F. W. 1948. The commonness and rarity of species. Ecology 29:254-283.

----------. 1962. The canonical distribution of commonness and rarity: part I. Ecology

43:185-215.

----------. 1962. The canonical distribution of commonness and rarity: part II. Ecology

43:410-432.

Pyle, R. M. 1981. The Audubon society field guide to North American butterflies. Alfred A. Knopf, New York. 916 pp.

Raven, P. H. 1967. The floristics of the California Islands, pp. 57-67. In R. N. Philbrick (ed.), Proceedings of the Symposium of the biology of the California Channel Islands. Santa Barbara Botanic Garden, Santa Barbara, California.

RiSKJND, D. H. & D. D. Diamond. 1988. An introduction to environments and vegetation, pp. 1—15. In Amos, B. B. & F. R. Gehlbach (eds.), Edwards Plateau vegetation: plant ecological tudies in Central Texas Baylor University Press, Waco, Texas.

ROBINSON, G. S. 1975. Macrolepidoptera of Fiji and Rotuma, a taxonomic and geographic study. E.W. Classey Ltd., Faringdon, England.

SCOTT, J. A. 1986. The butterflies of North America. Stanford University Press. 583 pp.

SHAPIRO, A. M. 1982. Taxonomic uncertainty, the biological species concept, and the Nearctic butterflies: a reappraisal after twenty years. J. Res. Lepid. 21:212-218.

Simpson, B. 1994. Texas oak savannas. Wildflower, Autumn 1994:35-38.

SMEINS, F. E. & L. B. MERRILL. 1988. Long-term change in a semiarid grassland, pp. 101-114. In Amos, B. B. & F R. Gehlbach (eds.), Edwards Plateau vegetation: plant ecological studies in Central Texas Baylor University Press, Waco, Texas.

Stanford, R. E. & P. A. Opler. 1993. Atlas of western USA butterflies including adjacent parts of Canada and Mexico. Denver and Fort Collins, Colorado, and BioQuip Products Inc., Gardena, California. 275 pp.

THOMAS, J. A. 1983. A quick method of estimating butterfly numbers during surveys. Biol. Conserv. 27:195-211.

WARREN, M. S., E. POLLARD, & J. A. Tpiomas. 1984. The status of the heath fritillary Mellicta athalia Rott. in Britain. Biol. Conserv. 29:87-305.

Whittaker, R. H. 1972. Evolution and measurement of species diversity. Taxon 21:213-251.

YELA, J. L. 1992. Plant communities and noctuids of central Spain: some notes on their relationships (Lepidoptera). Nota Lepid. Suppl. 4:57-63.

Received for publication 3 March 1996; revised and accepted 25 November 1997.