Imantodes cenchoa (Linnaeus 1758)

Imantodes cenchoa

NATURAL HISTORY

Imantodes cenchoa is a slender, elongate (snout-vent length to 901 mm but adults usually less than 800 mm), arboreal snake. It occurs in primary and secondary forests mainly below 1500 m but up to 2000 m. Frequenting low vegetation, it is commonly found in bromeliads (Stuart, 1948; Taylor, 1951) and coffee trees (Slevin, 1939; Landy, et al., 1966) during the day. In a simulated natural environment, captive specimens spent 90 percent of daylight hours coiled in bromeliads (Henderson and Nickerson, 1976). One adaption for arboreal living, an enlarged middorsal scale row, provides rigidity while spanning gaps between branches (Schmidt and Inger, 1957) as much as one-half its body length (Gans, 1974). It forages for food at night, often feeding on sleeping anoles (Henderson and Nickerson, 1976). Small arboreal lizards (mainly Anolis sp.) appear to be the principal prey (Henderson and Nickerson, 1976; Landy, et al., 1966; Stuart, 1948, 1958; Wehekind, 1955), although reptile eggs have also been reported as stomach contents (Landy, et al., 1966). Beebe (1946) reported an individual pursuing a tree frog (Ololygon rubra). In captivity, frogs (Colostethus and Eleutherodactylus) are accepted as food (Test, et al., 1966).

BODY SIZE

HATCHLINGS.—The largest snake with a yolk-sac scar was 327 mm (snout-vent length). Out of 17 individuals of this length or less, five (232, 255, 259, 272, and 327 mm long) possessed yolk-sac scars. The average snout-vent length of hatchlings is 279.7 ± 27.2 mm; range 232–327.

GROWTH AND SEXUAL MATURITY.—Sexually mature females of Imantodes (all localities) had a mean snout-vent length of 715.9 ± 67.7 mm and a range of 621–901 mm (N = 33). Females (5) with oviducal ova and class V follicles have an average snout-vent length of 789.4 ± 96.7 mm (637–901). The sexually mature females from Chiapas (2) have a mean length of 708.5 (693–724), from Honduras (3) 640.7 ± 21.7 mm (621–664), from Costa Rica (1) 639.0 mm, from Panama (7) 734.3 ± 83.4 mm (662–901), from Ecuador (15) 742.5 ± 59.6 mm (641–818), and from Peru (5) 674.0 ± 47.6 mm (628–732). From these data, we infer that female Imantodes reach sexual maturity at about 620 mm. Again, as in both Coniophanes and Dipsas, we can only speculate that males mature at about the same size or less.

A size-month matrix for Imantodes (Figure 11) exhibits weak segregation of size classes. We propose that most hatchlings appear from March through August. After one year, they have grown to an average of 450 mm, and after two years of growth, their average body length approaches 640 mm. This suggests that most of the snakes have attained sexual maturity by the end of two years. From these speculations, an approximate growth rate of 3.5 mm/week (first two years) can be estimated.

SEXUAL DIMORPHISM

BODY AND TAIL LENGTH.—In the more northern samples (Chiapas; Yoro, Honduras; Costa Rica) adult males on the average have longer tails than equal-sized females (Figure 12) but only slightly so. Snakes from southern localities (Panama, Ecuador, and Peru) do not show sexual dimorphism in tail length (Figure 12). The Ulúa River, Honduras, sample (N = 20) females possess longer tails.

GONAD POSITION.—The relative position of the gonads in Imantodes shows little sexual dimorphism. F values show borderline significance for our two larger samples; 8.02, df 1/27 for Ecuador (Napo), and 1.78, df 1/13 for Peru (Loreto). The gonads are situated 6 to 8 percent of snout-vent length from the vent.

TAIL BREAKAGE.—Tail break frequency in Imantodes is low but substantial. Combining localities, the incidence is 6 percent in females and 10 percent in males of all sizes, 9 percent in mature females, and 14 percent in mature males. Males of all sizes have a higher incidence of tail breaks in five of nine local samples, females in only one of nine; mature males have a higher incidence of breaks in four of the nine samples, females in two of nine. These differences are not significant (x2 = 0.95 for all sizes of all localities; x2 = 0.40 for adults of all localities); hence there is no evidence of differential prédation of the sexes.

SEX RATIO

In six of the nine local samples for all sizes, males outnumber females = 3.64 for combined nine samples). The same trend is observed when considering only mature snakes (x2 = 3.52). The Panamanian sample differs in having a preponderance of females (13 females: 4 males, juveniles and adults; 7:2, only adults). Only the Chiapas and Rio Bobonaza samples show an equality of sexes.

REPRODUCTION

CLUTCH SIZE.—The modal clutch size for oviducal eggs in I. cenchoa is two. Classes IV and V follicles show a modal clutch size of three in the two northern samples (Chiapas and Honduras), whereas the three southern samples (Ecuador and Peru) have modal sizes of two (Table 6). The range for all samples combined is one to three. The single instance of three oviducal eggs occurred in the largest female (snout-vent length 901 mm; Panama). Despite the inadequate sample size, this may suggest, as in Coniophanes, a positive correlation between body size and clutch size (Figure 13), although available data provide a regression slope of nearly zero. Potential clutch size in classes II and III ranges from 1–14 with 4–6 occurring most frequently in II and 1–3 most often in III.

Fitch (1970) reports a gravid female from Iquitos, Peru, containing two oviducal eggs. Egg size (oviducal) in our sample (9) ranges from 21.2 to 37.2 mm long, with a mean of 30.7 ±5.13 mm.

REPRODUCTIVE CYCLE.—In the Ecuadorian and Peruvian samples (Figure 11), the occurrence of hatchlings throughout the year probably indicates continuous reproduction. Although the northern sample (Mexico to Panama) collectively shows similar trends, individual localities lack a sufficient number of hatchlings to draw conclusions (Table 7). Owing to the seasonal nature of rainfall at some of the sample localities, e.g., Panama Canal Zone, a seasonal reproductive cycle might be found in these populations. The presence of gravid females during late April and mid-June in Guatemala (Stuart, 1948) and of hatchlings during July and August in adjoining Chiapas, Mexico, indicates an extended reproductive season. It correlates well with the long wet season of that area, May through November (Stuart, 1948).

Discussion and Summary

Our two primary questions on reproduction—regional variation in clutch size and seasonality of reproduction—remain inadequately answered. The opportunistic nature of museum collections does not currently permit us to obtain large and year-round samples from single localities. A few old and large collections of single snake species are available, but all too often they lack collecting dates or the viscera have been removed. Surprisingly and disappointingly, we have discovered that, all too frequently, snakes from recent collections have been poorly prepared and preserved and not uncommonly held in captivity before being preserved. Both practices lead to inaccuracies in the initial data gathering and the final interpretations. In spite of these difficulties, museum collections with or without recently collected specimens can contribute to an improved understanding of snake reproduction.

Coniophanes fissidens, D. catesbyi, and I. cenchoa are small or, at most, moderate-sized snakes. Their modal clutch sizes of three, two, and two eggs, respectively, are a reflection of their short and/or narrow -body cavities. Perhaps the possession of small clutch sizes makes them poor candidates for demonstrating regional variation in clutch size. We suspect, however, that life history adaptations, such as growth rate, size at sexual maturity, and average adult life span, will produce locally different reproductive patterns. Differences in clutch sizes will be as apparent in “small clutch” species as in “large clutch” species owing to the lower variance of “small clutch” species. We think our inability to recognize the presence or absence of regional variation results from our small sample sizes.

Coniophanes fissidens is the smallest of the three species examined, yet it has the largest clutch. This is accomplished by having the shortest egg length, mean of 22.4 mm. In D. catesbyi and I. cenchoa, mean egg length is 27.7 and 30.7 mm, respectively; egg length increases directly with body length. This suggested trend is provocative. How closely are egg length and volume associated with body size, intra-and interspecifically? Presumably egg (fertile) size is relatively constant within a species but increases with increasing body size in species comparisons. The larger egg will result in a larger hatchling, whose chance of survival is presumably enhanced by its larger size. Nonetheless, egg size must have an upper limit (plateau) where the advantages of large hatchling size are outweighed by the female's need to limit energy expenditure in egg production and/or to invest energy in egg number rather than size. Fitch's data (1970, figs. 12, 14) show that few subtropical or tropical snakes invest heavily in large clutch sizes (80 percent of snakes with clutch size of 12 or less) no matter what the adult body size of the species. The indication is that there is selective pressure for larger eggs or fewer eggs per clutch but more clutches per year.

Of the three species examined, only the data for the Peruvian D. catesbyi and Ecuadorian/Peruvian I. cenchoa are sufficient to indicate continuous or aseasonal reproduction. At the other localities for these two species and all localities for C. fissidens, the data are insufficient to discriminate between continuous and seasonal reproduction; however, we intuit the Central American samples, particularly the Mexican ones, to have seasonal reproduction for Coniophanes and Imantodes.

Our bias is to assume cyclic or seasonal reproductive patterns in subtropical and tropical snakes unless data indicate otherwise. Fitch (1970) and other biologists have the opposite bias. No matter what the researcher's preference, we must all be aware of the potential diversity in reproductive cycles at a single site and the potential of a species for modifying its pattern at different locations. The first point is demonstrated by the recognition of six reproductive patterns in Cambodian snakes (Saint Girons and Pfeffer, 1971): (1) polyestrous with aseasonal reproduction, (2) polyestrous with midwet season reproduction, (3) monoestrous with spring reproduction, (4) monoestrous with early summer reproduction, (5) monoestrous with early fall reproduction, and (6) double period of reproduction (data can also be interpreted as monoestrus with winter or dry season reproduction). The snakes of the Iquitos region, Peru, show both continuous and seasonal reproduction (Dixon and Soini, 1977, table 1); data are insufficient for finer subdivisions. The modification of scheduling pattern by a tropical colubrid at different localities is indicated by our data and those of others, although as yet not convincingly so.

Neill (1962) postulated seasonal reproduction for Belize snakes by estimating the age of posthatchlings on the basis of yolk-sac scar condition. His evidence suggested an August to October period for hatching and birth and a May to July period for fertilization and egg laying. The synchronization mechanism was assumed to be the cooler temperatures of December through February and the resulting period of inactivity. Henderson and Hoevers (1977) agreed with the reproductive schedule proposed by Neill but disagreed with the mechanism and its periodicity. They suggested the February to May dry season as the mechanism. Unwittingly, Henderson had presented data earlier (1974, table 1) that demonstrates that the period of inactivity is December through March, but he did not analyze these data. A comparison of the growth rates of Belize Oxybelis aeneus shows that the rate is 0.12 ± 0.15 mm/day during the dry season (December to March) and 0.62 ± 0.56 mm/day during the wet season (April to November)—a strong indication of reduced activity during the dry season.

The annual growth rate for Oxybelis aeneus (calculated from Henderson, 1974, table 1) is 0.87 ± 0.47 and 1.18 ± 0.75 mm/day in immature (less than 700 mm snout-vent length) females and males, respectively. These rates are considerably faster than our estimates of 1 mm/week for Coniophanes fissidens, 1 mm/week Dipsas catesbyi, and 3.5 mm/week Imantodes cenchoa. With the possible exception of I. cenchoa, our estimates of growth may be underestimates, for aside from Carphophis vermis (Clark, 1970), other snakes have a growth rate greater than 0.5 mm/day, e.g., Tropidoclonion (Blanchard and Force, 1930), Australian elapids (Shine, 1978).

Underestimates of growth will result in overestimates of age at sexual maturity. Our estimates are two to three years for these tropical snakes. Estimates for temperate-zone colubrids are usually no longer than these and often shorter. Three Australian elapids, Unechis gouldii, Hemiaspis daemelii, and H. signata, reach sexual maturity in the surprisingly short time of one year (Shine, 1978). It seems doubtful that any tropical species will mature faster than these temperate-zone species.

The available evidence on reproduction in tropical snakes suggests that they are not very different from their temperate-zone conspecifics and congenerics. Only the possibility of continuous reproduction and two or more clutches per year is available to tropical colubrids, but not all tropical localities will permit even these reproductive adaptations.