Part 3

KEY TO THE SPECIES AND SUBSPECIES

Although many juveniles differ greatly in general coloration from the adults, both the juveniles and the adults of any species or subspecies can be identified from the following key; juveniles differ from adults in extent and intensity of dark pigmentation but not in rows of scales involved.

1. Seven supralabials (3rd and 4th below orbit); 3 to 8 dark stripes along body 2

Eight supralabials (4th and 5th below orbit); unstriped or with more than 4 dark stripes along body, or dark with 2 or 4 pale stripes 3

2. Dark stripes involving no more than one longitudinal scale-row C. lineatus lineatus (part), p. 267

Dark stripes involving at least two adjacent scale-rows C. vittatus, p. 277

3. Supralabials having black borders above; head and body generally black with 2 or 4 white lines running length of body C. nevermanni, p. 272

Supralabials immaculate or having dark borders below; head and body usually pale with dark stripes, or without stripes 4

4. Lateral dark stripe through eye involving upper half of second scale-row; dark stripe on paravertebral row, at least posteriorly C. pulcher, p. 274

Lateral dark stripe becoming indistinct on body, or restricted to 4th or 3rd and 4th rows anteriorly, not involving 2nd scale-row on anterior 1/3 of body (an auxiliary lateral stripe sometimes present involving 2nd row); no paravertebral stripes 5

5. Stripes disappearing posteriorly (except for small spots of pigment on scale-row 4 or 7); 1st scale-row unpigmented C. lineatus concolor, p. 270

Stripes present posteriorly; 1st scale-row pigmented 6

6. Lateral stripes narrow on nape, restricted to 4th scale-row on body C. lineatus lineatus (part), p. 267

Lateral stripes involving 3rd and 4th rows, at least on nape C. lineatus dunni, p. 262

Analysis of Characters

Characters showing inter-specific and intra-specific variation and that have a wide range of variation were analyzed statistically, when possible, in order to determine extent of variation. One character (see table 3) was analyzed for sexual dimorphism, and for it the coefficient of difference is also given. The statistical terms and formulae have been adopted from Mayr, Linsley and Usinger (1953). Dorsal head shields varied individually and were of no taxonomic importance. Osteological and hemipeneal characters did not show enough variation to be considered here.

Scutellation

Labials, dorsals, ventrals, and subcaudals were the most useful scales.

Labials.--All species usually have eight supralabials except C. vittatus, which has seven. The only other population having a relatively high frequency of occurrence of seven supralabials is C. l. lineatus. In specimens having eight supralabials, the fourth and fifth enter the orbit; in specimens having seven supralabials, the third and fourth enter the orbit (the second and third are fused). Usually there are ten infralabials, sometimes nine or eleven; specimens having seven supralabials usually have nine infralabials, sometimes eight, rarely ten.

Dorsals.--Although there is no variation in the number of rows of dorsal scales, there is some in the method of scale reduction. There are 19 rows of dorsal scales from close behind the head to about midway on the body where two rows are lost, leaving 17 rows from there to near the base of the tail. This reduction is accomplished by fusion of the scales of the 3rd and 4th rows or sometimes by the dropping out of the 3rd row. The place at which reduction occurs in number of dorsal scales in relation to the ventral (scale) directly below is highly variable and of little taxonomic importance (table 1).

TABLE 1.--VARIATION IN THE PLACE OF DOSAL SCALE REDUCTION IN CONOPHIS.

Key to Columns ==================================== Std. Dev. = Standard Deviation Std. Err. = Standard Error Coe. Var. = Coefficient of Variation

==============+===========+========+=======+======+======+====== | Number of | | | Std. | Std. | Coe. Taxon | Specimens | Range | Mean | Dev. | Err. | Var. --------------+-----------+--------+-------+------+------+------ l. concolor | 45 | 89-114 | 102.5 | 5.57 | 0.83 | 5.43 l. dunni | 36 | 91-111 | 102.1 | 4.59 | 0.77 | 4.50 l. lineatus | 26 | 91-107 | 100.2 | 3.59 | 0.72 | 3.58 nevermanni | 6 | 84- 97 | 93.2 | 4.71 | 1.92 | 5.05 pulcher | 26 | 94-119 | 104.6 | 4.90 | 0.96 | 4.68 vittatus | 170 | 84-118 | 102.3 | 6.60 | 0.16 | 6.45 --------------+-----------+--------+-------+------+------+------

Ventrals.--The number of ventral scutes varies from 149-183, and shows no significant variation in the means (table 2).

Subcaudals.--The number of subcaudal scutes varies from 55 to 89. In some populations there is no overlap in the range of variation of males and females. The total variation and sexual dimorphism are analyzed in table 3.

Size and Proportions

Although considerable variation in size is observable, little taxonomic use is made of size since sufficient series are not available to determine age classes. The subspecies attaining the largest size is C. lineatus concolor; all others are smaller and of about the same size and proportions. The longest specimen, a male of C. l. concolor, has a body length of 893 mm., a tail length of 274 mm., and a total length of 1167 mm.

TABLE 2.--VARIATION IN THE NUMBER OF VENTRALS IN CONOPHIS.