IH^Ip

HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

UNIVERSITY OF KANSAS ^'mmL^i^ATfoM

MUSEUM OF NATURAL HISTORY No 52

MUS.

APf? 1 4 1975

Harvard

A Demographic Study of the
Ringneck Snake (Diadophis
punctatus) in Kansas

By

Henry S. Fitch

UNIVERSITY OF KANSAS

LAWRENCE 1975 April 3, 1975

UNIVERSITY OF KANSAS PUBLICATIONS
MUSEUM OF NATURAL HISTORY

The University of Kansas Publications, Museum of Natural History, beginning
with volume 1 in 1946, was discontinued with volume 20 in 1971. Shorter research
papers formerly published in the above series are now published as Occasional
Papers, Museum of Natural History. The Miscellaneous Publications, Museum of
Natural History, began with nimiber 1 in 1946. Longer research papers are pub-
lished in that series. Monographs of the Museum of Natural History were initiated
in 1970. All manuscripts are subjected to critical review by intra- and extramural
speciaHsts; final acceptance is at the discretion of the publications committee.

Institutional libraries interested in exchanging pubhcations may obtain the Occa-
sional Papers and Miscellaneous Publications by addressing the Exchange Librarian,!
University of Kansas Library, Lawrence, Kansas 66045. Individuals may purchase
separate numbers of all series. Prices may be obtained upon request addressed to
Publications Secretary, Museum of Natural History, University of Kansas, Law-
rence, Kansas 66045.

i

The University of Kansas
Museum of Natural History

Miscellaneous Publication No. 62
April 3, 1975

A Demographic Study o£ the Ringneck Snake
(Diadophis punctatus) in Kansas

by

Henry S. Fitch

Department of Systematics and Ecology

Natural History Reservation

The University of Kansas

The University of Kansas

Lawrence

1975

University of Kansas Publications, Museum of Natural History

Editor: Richard F. Jolinston

Miscellaneous Publication No. 62

pp. 1-53; 19 figures

Published April 3, 1975

Museum of Natural History

The University of Kansas

Lawrence, Kansas 66045

U.S.A.

Printed by

University of Kansas Printing Service

Law^rence, Kansas

CONTENTS

ABSTRACT 4

INTRODUCTION ............ 4

Acknowledgements 5

Materials and Methods 5

Morphology 8

HABITAT 11

BEHAVIOR 12

TEMPERATURE RELATIONSHIPS 18

SEASONAL SCHEDULE 19

ASSOCIATED ANIMALS 20

FOOD HABITS 22

SPATIAL RELATIONSHIPS 24

REPRODUCTION 27

Aggregation and Mating 27

Reproductive Cycles 28

Eggs 29

Size of Clutch 29

Reproductive Effort 31

Nests 32

GROWTH 33

POPULATION DENSITY 39

COMPOSITION OF THE POPULATION 43

Age Structure 43

Sex-ratio 45

PREDATION 47

DISCUSSION 50

LITERATURE CITED 52

ABSTRACT

Field records of more dian 14,000 individuals of Diadophis piinctatus were assembled over
a 26-year period on several small study areas at the University of Kansas Natural History
Reser\ation, mostly within a half-mile radius, with emphasis on marking and recapture.
Emergence from hibernation begins in late March and continues into early April, ovulation
occurs in the latter half of May, and the elongate, sausage-shaped eggs, averaging approxi-
mately 3.9 per clutch, are laid in late June or early July. Hatching occurs at the end of August
or in September. Females become sexually mature after their third hibernation, typically at a
snout-vent (S-V) length of approximately 235 mm, but males mature a year earlier, at a min-
imum length (S-V) of 166 mm. Growth continues, but at a reduced rate, after attainment of
sexual maturity. Snakes already above average adult size typically gain from one to three per
cent in length annually. Mortality of adults is not age-specific, and loss is approximately 25
per cent annually. A small percentage of the snakes survives to an age of 15 or more years.
Young seem more secretive than adults and are less well represented in samples.

Sex-ratio in samples is subject to drastic seasonal fluctuation; females range from 24 per
cent in early fall to 71 per cent in midsummer, but in late March, April and early May, when
the snakes are most in evidence, males are always more numerous. The snakes are active over
a wide range of temperatures but tend to maintain body temperatures between 25 and 29.5
degrees C. Their surface activity is chiefly at air temperature witliin or below this range, and
at higher air temperatures they retreat underground. Body temperature is maintained by con-
tact with sun-warmed objects that the snakes use for shelter. Occasionally, at relatively low
air temperatures, the snakes venture into the open and bask in sunshine.

Earthworms, especially AHoIohophora caliginosa, provide most of the food, and it is esti-
mated that each snake consumes on the average approximately three times its body weight in
earthworms annually. Seven calculations of population density (based upon recapture ratios)
varied from 719 to 1849 per ha and averaged 1266. Eight kinds of predators on the Reser-
vation were found to eat ringneck snakes, but their toll would account for only part of popu-
lation regulation. The copperhead, racer and Red-tailed Hawk were the predators in most
recorded instances, but moles, mice and birds other than raptors may also be among the im-
portant predators. Marked snakes were recaptured at distances of zero to 520 m from the
original capture site; however, one-fourth were within 10 m, even after intervals of several
years, showing a strong tendency to remain within a home area. Home ranges were often
elongate, with maximum axes of approximately 140 m. Ranges tend to be progressively altered
through time.

INTRODUCTION

The ringneck snake (Diadophis piinc-
tatus) is a highly successful transconti-
nental North American species. Its local
populations are adapted to the swamp-
land of subtropical southern Florida, the
boreal forest of heavily glaciated areas
in northern Michigan and the desert
grassland of the southwestern United
States. Partly because of retiring or se-
cretive habits, the species is considered
rare throughout most of its extensive ge-
ographic range. However, in a central
area including northeastern Kansas,
where my study was made, it is abun-
dant and ecologically prominent.

Soon after field work began on the
newly-created University of Kansas Nat-
ural History Reservation, the ringneck
snake attracted attention as one of the
common animals meriting intensive study
because of its probable ecological roles

as a consumer of invertebrates and as an
abundant food source for larger preda-
tors. In the ensuing 26 years the species
has been allotted more time and effort
than any other kind of animal in the
many autecological studies carried out
on the Reservation.

This small snake has proven to be a
difficult subject for field study, despite
its great abundance. Individuals ordi-
narily cannot be observed readily under
natural conditions. When uncovered or
discovered in the open, they are glimpsed
only momentarily, escaping underground
or into dense vegetation. The snakes
cannot be found for most of the year,
including, of course, their hibernation
season. But they are also difficult to find
in summer, especially in hot and dry
weather.

The present study had its inception

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

in the spring of 1949 when I began tak-
ing field notes on ringneck snakes and
individually marking those that were
captured. In the 1950s the snakes were
observed or marked only occasionally,
as part of a comprehensive program to
study the vertebrates of the Reservation.
Several students at different times under-
took ecological studies of Diadophis in
the area: John Hawken (1951-1952),
Robert M. Hedrick (1956-1957) and
Kenneth Shain (1958-1959). Hawken's
and Hedrick's records were mostly from
the Rat Ledge area, whereas Shain's
were from the vicinity of the Reservation
headquarters. Although each of these
projects was of short duration, their
combined data contributed substantially
to my study.

In the spring of 1964 large-scale cap-
ture and marking of Diadophis was un-
dertaken at Rat Ledge, in the field near
Reservation headquarters "House Field,"
and in "Quarry Field" along the northern
edge of the Reservation ( Fig. 1 ) . On 24
September 1965 a system was initiated
for recording the pattern of pigmentation
on the posterior ventrals to supplement
the clipping of scales as a means of posi-
tive individual identification. In 1966,
operations at Rat Ledge were abandoned
and field work was concentrated in the
House Field and Quarry Field areas,
which were close enough that snakes
often shifted from one to the other.

The marking program was suspended
at the end of May, 1968, but in the
spring of 1969 all first-year young cap-
tured were group-marked and those
considered yearlings were individually
marked in an eff^ort to gain more infor-
mation about the growth of young. In
the fall of 1969, and in 1970 through
1974, efforts to capture the snakes were
continued; those taken were not marked
but were measured, weighed, sexed, and
carefully examined for marks made in
the period 1958 through 1969. Hence,
although procedures and areas have var-
ied somewhat from year to year, my in-
tensive field study extended over an

eleven-year period, and was supple-
mented by 15 years' accumulation of
earlier data. Expenditure of so much
time and energy on a local population of
one species may appear prodigal; yet,
the recapture records that were the key
to an understanding of ecological rela-
tionships and population dynamics to-
talled only a few hundred, and when
these were subdivided into sexes and
ages, some samples were inadequately
small.

Acknowledgments

Members of my family, Alice Fitch
Echelle, Chester W. Fitch, John H.
Fitch and Virginia R. Fitch, helped me
from time to time, both with the field
work and with various stages in the
analysis of data. Many persons helped
in the search and capture of the snakes;
Robert R. Fleet's contribution of many
large series was especially outstanding.
John Hawken, the late Robert M. Hed-
rick, and Kenneth Shain each contrib-
uted data accumulated from one or more
seasons of field work. Donald R. Clark,
Robert W. Henderson, Richard L. Lattis,
Julian Lee, Michael V. Plummer, Eric
Rundquist, and Arnold K. Smith were
employed as research assistants at differ-
ent times, and all contributed substan-
tially. In 1971, 1972, and 1973, while
engaged in his own field research on the
Reservation, Lattis captured and con-
tributed many ringneck snakes marked
in 1969 or before; these long term sur-
vivors provided some of the most reveal-
ing records of longevity, growth, and
population structure. Financial support
from the National Science Foundation
(6-17084) and from the University of
Kansas is gratefully acknowledged.

Materials and Methods

In spring ringneck snakes were found
in large numbers beneath flat rocks, es-
pecially in open woodland or at wood-
land edge, at hilltop edges including
"Skink Ledge," "Rat Ledge," and an old

I

6

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

500 m

Fig. 1. — Contour map of northwestern part of The University of Kansas Natural
History Resei-vation showing study areas of Diadophis punctatus, 1949 to 1974. A.
"Quarry Field," a flat hilltop with tall grass interspersed with brush. B. Abandoned
limestone quarry at south exposure of hilltop edge. B 1. Hibernation site for snakes
from both hilltop and valley, limestone outcrop with southwest exposure. B 2. Egg-
laying and "nursery" area at east end of quarry. C. "House Field," at head of small
valley. D. "Rat Ledge," limestone outcrops at hilltop edge with south exposure.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

rock quarry, where suitable flat rocks
were numerous. In 1966 and 1967 funnel
traps were used on a large scale. These
were adaptations of the cylindrical wire
funnel traps used earlier in population
studies of larger kinds of snakes at the
Reservation (Fitch, 1951; Clark, 1966).
A glass quart far served as the trap body;
a funnel of 1/8 inch mesh screen was sol-
dered to the screw-top of the jar. These
traps were used in combination with
drift fences consisting of strips of sheet
metal 2 to 6 m long and 25 to 38 cm
wide, buried to a depth of at least 5 cm,
so that snakes could not readily pass be-
neath. A funnel trap was set at each end
of the barrier, and traps were generally
situated in clumps of grass or other dense
vegetation, so that captured animals
would not be exposed to direct sunlight.

Strips of sheet metal that were left
lying in grassy places from time to time
were found to have groups of ringneck
snakes congregated beneath them. A se-
ries of such experiences brought belated
realization that formerly pastured grassy
fields harbored high populations of ring-
neck snakes with densities comparable to
those in the rocky upper-slope and hill-
top habitats. Subsequently, large num-
bers of sheet metal strips and weathered
boards were distributed over study areas
selected in House Field and Quarry
Field, and collectively these objects
served as highly effective traps. Success-
ful use of these devices as traps de-
pended upon proper timing of field work.
The snakes were attracted to the metal
chiefly in cool weather, when the metal
was warmed by sunshine but air and
substrate temperatures were lower than
the snake's preferenda. However, at
most times the metal was uncomfortably
hot or cold to the snakes and was
avoided. The boards provided better in-
sulated hiding places and were used at
certain times when the metal strips were
unattractive. The boards and metal
strips proved to be so effective for cap-
ture of the snakes that use of funnel
traps was discontinued after 1967.

The ringneck snakes captured were
processed in the field during the early
years of the study, but during most of
the study period snakes were brought to
the laboratory. Processing consisted of
probing for the invaginated hemipenes
in the tail base to determine sex and
measuring the snout- vent (S-V) length
and tail length to the nearest millimeter.
For the snout-vent measurement the
snake was held against a rigid ruler and
gently pulled out straight for several
seconds until muscle fatigue caused it to
relax momentarily, permitting stretching
to the maximum length. The accuracy of
measurement was checked in a random
sample of 200 adults recaptured within
a two-week period — too short an interval
for the snake to make measurable growth.
Mean deviation from the original meas-
urement was 1.52 mm, approximately 0.6
per cent.

Snakes captured were individually
marked by scale-clipping, in a modifica-
tion of the Blanchard and Finster ( 1933 )
system. Three scales were clipped on
each snake to attain a unique combina-
tion; the left or right end of one of the
posterior ventrals was clipped out, also
one subcaudal on the left side and one
on the right were partly or wholly ex-
cised. The subcaudals clipped included
numbers 2 to 20 on each side, counting
posteriorly from the vent. On each side
the count began with the first scale that
made contact on or near the midline
with a counterpart on the opposite side;
smaller scales just behind the vent,
which did not extend to the midline,
were not included.

Snakes have some capacity to regen-
erate lost or damaged scales (Conant,
1948) and this capacity seems especially
well developed in Diadophis. The scars
left by clipped scales often became ob-
scure after several years. Regeneration
of the scale surface was sometimes so
complete that partial fusion along a su-
ture or a break in the sequence of pig-
mented dots on the subcaudals provided
the only clues. Examination of the scars

8

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

under a dissecting microscope often was
necessary before identification was made.
Natural scars on the tail were common
and were a potential source of confusion.
Bites of small predators such as Blarina
brevicauda caused indentations that
sometimes were mistaken for those made
by clipping scales with scissors.

Because of these potential sources of
confusion or error, it was necessary to
supplement scale-clipping with other
means of identification. Such means was
provided by the variable pigmentation
of the ventral surface. Markings on the
posterior ventrals (numbers one to ten
anterior to the anal plate) were classi-
fied as "bars," "blotches," "spots," and
"dots" according to their size and shape.
The number and position of such marks
on each posterior ventral was recorded.
In many instances positive identification
was based on the ventral markings when
it could not have been derived from the
scars of clipped scales alone.

Weight was recorded to the nearest
one-hundredth of a gram. Laboratory
balances were used until 1969 when
Oskar Ludi spring scales became avail-
able and greatly expedited weighing.
Obvious bulges in snakes were palped
up to the gullet for identification; and
when intact they were removed from the
snake for weighing and/or measurement.

Reproductive condition was noted.
In females of adolescent size, thickening
of the cloaca indicative of sexual matu-
rity was recorded. A drop of normal
saline was pipetted into the cloaca, and
a smear was taken and examined for
motile sperm under magnification x 60.
Mature snakes of both sexes often had
great c^uantities of sperm in the cloaca.
In the male its presence was considered
indicative of breeding condition, and in
the female it was considered indicative
of recent copulation.

Morphology

Characters of the local population. —
The head scutellation is typically colu-
brine (Fig. 2), with nasal divided, loreal

small and rhomboidal, 2 preoculars and
2 postoculars on each side, one anterior
and one posterior temporal, and 7 supra-
labials, the third and fourth in contact
with the eye. There are 15 or 17 scale
rows, each body scale with an apical pit,
anal plate divided; eye with round pupil
and golden-brown iris. The following
are ratios to snout-vent length in a typ-
ical adult male: tail length, 24.50%; head
length, 4.38%; body diameter, 2.14%; di-
ameter of eye, .63%; length of longest
ventrals, .67%; length of longest dorsals,
.56%. The teeth are slender and conical,
with the following numbers most fre-
quent on the bones that are dentigerous:
dentary 16 (or 17), maxillary (Fig. 4)
13 (or 14), pterygoid 10 (or 9), palatine
20 (or 21). In the dentary series the
anteriormost tooth is relatively small and
slender, the second less so and the larg-
est are those a little anterior to the mid-
dle; farther posteriorly the teeth are
progressively smaller and more recurved.

Fig. 2. — Head of adult Diadophis piinctattis

armji showing scutellation and pattern; top,

lateral; center, dorsal; bottom, ventral.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

9

Table 1. — Tail Length as Percentage of Snout- Vent Length in Male and Female
RiNGNECK Snakes of Different Size Classes. For Each of the Eight Samples N = 25.

Size Class Males Females

Large adults 22.1 ±: .315 (25.4-20.8) 17.9 ± .201 (19.7-15.3)

(males 298-272 mm,
females 347-300 mm)

Typical adults 23.2 ih .323 (25.9-20.7) 17.9 ± .217 (19.8-15.9)

(males 249-240 tum,
females 289-270 mm)

2nd year young 21.9 ± .069 (23.0-20.9) 17.3 ± .154 (19.1-15.6)

(189-170 mm)

1st year young 20.8 ± .345 (25.2-17.8) 17.1 ± .248 (18.6-15.4)

(128-111 mm)

with the smallest only about one-third
the size of the largest. In the maxillary
series the anterior teeth are relatively
long, slender and erect, and those far-
ther back are stouter and directed more
posteriorly. The last two of the series
are almost side by side and are stouter
and more backward pointed than the
others, and they are separated from the
next anterior tooth by a space about
equal to the length of that tooth. The
palatine teeth are small, about half the
size of the maxillary and dentaiy teeth,
with marked decrease in size posteriorly.

Relative tail length is subject to con-
siderable individual variation and also is
subject to sexual difference and onto-
genetic difference. In 200 randomly se-
lected individuals with intact tails, tail
length varied from 15.3 to 25.9% of snout-
vent length ( Table 1 ) . At all ages males
have relatively longer tails, and adults of
both sexes have relatively longer tails

than do young. In males this ontogenetic
change in relative tail length is three
times as great as in females. In the larg-
est males the trend is reversed and rela-
tive tail length is a little less than in
smaller individuals. There is overlapping
between the sexes in relative tail length
only in the first-year young ( 12% ) .

Trends in size, as indicated by snout-
vent length and weight, are shown in
Table 2. In this table the figures for the
largest adult and the smallest sexually
mature snake are based upon single in-
dividuals, and hatchling size is based
upon 83 young of 21 clutches hatched in
the laboratory. All other figures are
based upon larger samples, with 100 or
more snakes in each.

The body color is dark bluish olive
dorsally, the head darker than the body.
Just behind the head is a pale buffy or
dull orange band VA scales wide, edged
with black both anteriorly and posteri-

Table 2. — Sizes of Male and Female Ringneck Snakes of Different Age Classes in

A Local Population

Male

Female

Length S-V
(mm)

Largest adult 315.0

Average adult 253.6

Young adult (3rd year) 223.0

Smallest sexually mature 168.0

Average 2nd year young (in spring) 184.6

Average 1st year young (in spring) 122.6

Hatchling 109.3

Weight
( grams )

Length S-V Weight

(mm)

9.0

5.4

3.9

2.0

2.3
.97
.74

391.0
281.7
243.8
210.0
194.1
127.0
114.1

( grams )

7.0
5.0
3.5
2.8
1.05
.81

10

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

orly. On the side of the head the sepa-
ration of the darker dorsal color from the
paler ventral color is accentuated by an
irregular black band along the upper
parts of the supralabials. The chin is
cream-colored or buffy, speckled with
black, and gradually brightening to yel-
low on the neck and more posteriorly on
the ventral surface. The ventral yellow
has some orange suffusion which be-
comes much more prominent on the tail.
Black spots are prominent on the ventral
surface, and may be arranged in a fairly
regular mid-ventral row or may be
paired or irregular. Ventrals usually have
black areas in their lateral corners. Fig.
3 shows six types of common ventral pat-
terns, but intermediate stages occur, and
usually two or more of the types shown
are present on the same snake. In a ran-
dom sample of 200 snakes mid-ventral
spots ( type A ) were most common, with
44.0%, followed by paired spots ( types B
and C) with 40.7%, blotches and bars
(type F) with 6.9%, irregular marks (type
D) with 6.0%, and unmarked (type E)
2.4%. The subcaudals have black in the
corners and lack other marks.

The hemipenis (Fig. 4) consists of a
shaft, and a capitate part which com-
prises the distal one-fourth. The capitate
part is slightly constricted at its base,
slightly expanded and shallowly bilobate,
with many large and irregular pitlike
depressions (calyces). The shaft, except
for the part near the sulcus, is covered
with spines which are variable in size
and distribution and are thick and Heshy
except at the tips. The sulcus spemia-
ticus is bifurcate distally.

Geographic variation. — In the popu-
lation studied, much individual variation
was found in characters that have been
reported to vary geographically and
which are the bases for recognition of
the many subspecies of Diadophis punc-
tatus. Characters that are notably vari-
able are: body size, dorsal color, ventral
color, size and shape and distribution of
black markings on ventral surface, pres-
ence and width of neck ring, black mark-

FiG. 3. — Six types of ventral patterns common

in ringneck snakes of The University of Kansas

Natural History Reservation.

on facial region, number, of dorsal
scale rows, number of ventrals and sub-

mg

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

11

caudals and number of supralabials. The
population studied is centrally located in
the geographic range of the species, and
is intermediate relative to other popula-
tions in some of its characters but is near
the extreme in some others.

In body size the snakes studied are
near the minimum for the species. Popu-
lations in the northeastern states, the far
West, and especially those in Utah and
Arizona, are much larger, up to about
twice the linear dimensions of the Kan-
sas snakes. In some regions, especially
the far western states, the snakes are
more slender than those from Kansas,
which appear relatively short and
stubby. The dark dorsal color is slaty in
some populations but may be suffused
with brown, green or blue. In some
populations the ventral color is best de-
scribed as yellow, with relatively little

E
E

Fig. 4 — Above, hemipenis of Diadophis punc-

tatus; below, lateral view of left maxillary of

Diadophis punctatus.

change from body to tail. In others, es-
pecially those in which the tail-spiralling
response is highly developed, the under-
side of the tail is coral red. Scales are
in 15 or 17 rows in the Kansas snakes
and some others. Over much of the range
15 rows are the normal number, but 17
is the normal number in some western
populations in which body size is large.
In some other western populations which
are notably slender-bodied, scale rows
vary from 15 to 13. In the population
studied, there is normally a neck ring at
least 1/2 scale rows wide, but in eastern
and some far western populations the
ring is narrower and /or broken on the
middorsal line, and in southwestern pop-
ulations it is absent. In the Kansas
snakes the red or orange ventral color is
confined to the ventrals and subcaudals,
but in most more western populations
the ventral coloration encroaches onto
the dorsal scale rows and may cover the
first two on each side. The Kansas snakes
have seven pairs of supralabials but some
other populations have eight. In the
Kansas snakes the black marks of the
ventral surface tend to be irregular but
in the southeastern populations the mark-
ings are regular and symmetrical with a
half-moon shaped central spot on each
ventral scale, and in some far western
populations the ventral markings are
normally absent.

The number of ventrals in Diadophis
punctatus arniji ranges from 142 to 185,
and, for the species as a whole, from 126
to 239 (Wright and Wright, 1957:159).
The numbers of ventrals in females ex-
ceed the numbers in males by from 10 to
20 in different populations. In general,
far western populations have more ven-
trals than D. p. arniji and eastern popu-
lations have fewer. The subcaudals num-
ber from 30 to 79 and there are usually
6 to 8 more in males than in females
within any population.

HABITAT

As a transcontinental species Dia-
dophis punctatus occurs in most of the

12

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

major biomes of North America. In the
region of my study it occurs in a variety
of habitats but is especially associated
with the ecotone between deciduous for-
est and prairie. Under original condi-
tions the transition from forest to prairie
was sharp and clearly defined. Under
modern conditions woodlots, groves, old
fields and pastures provide a spectrum
of combinations of the two main types of
vegetation with habitats that are favor-
able for the snakes.

Ringneck snakes are present on all
parts of the Natural History Reservation.
However, this area, having a variety of
topographic features, past treatments,
and plant associations, is unequally used
by them (Figs. 5, 6). The transition
from low to high numbers reveals the
combination of ecological factors re-
quired by the species. There is some
seasonal shift between habitats, and in
some situations where the snakes use
loose surface objects for shelter, they are
more readily found than where such
shelters are lacking, even though the lat-
ter situations may have more snakes.
Subject to these constraints, habitats rep-
resented on the Reservation have been
found to have ringneck snakes, in ap-
proximately the following order of abun-
dance : ( 1 ) Old hilltop pastures domi-
nated by Bromiis inennis, with thickets,
clumps and scattered small trees of Cor-
nus drummondii, Mains ioensis, Madura
pomifera, Gleditsia triacanthos, Ulmtis
americanus and others. Bottomland pas-
tures with similar plant assemblages are
slightly less favorable habitat for the
snakes, perhaps because they are on the
average cooler and damper than the hill-
tops. (2) Hilltop edges with limestone
outcrops and with loose rock strewn over
adjacent upper slopes, and with brush
and intervening open spaces. South-
facing hilltops that are relatively open
and have an abundance of flat rocks are
preferred over those that face north or
are heavily shaded with trees or brush,
lack loose ground objects providing hid-
ing places. (3) Woodland; well-drained

situations, especially south-facing with
an open canopy and little underbrush,
are favorable, whereas soggy soil and
deep shade under unbroken leaf canopy
are unfavorable. (4) Prairie, but gener-
ally in association with such features as
gullies, rock outcrops and patches of
brush, indicating that extensive areas of
level, homogeneous grassland are little
used. (5) Riparian situations along the
edges of intermittent streams or artificial
ponds. Especially in dry weather mois-
ture renders these situations attractive,
and eroded banks, exposed tree roots,
rocks, and accumulations of drift provide
hiding places. (6) Old fields, but only
after dense ground vegetation has be-
come established, providing shelter.

In summary the snakes are not lim-
ited to any one habitat, but occur in
various types of terrain and vegetation.
Some of the chief requirements are soil
that is slightly damp but not wet or
soggy, abundant shelter in the form of a
surface mat of dead vegetation and /or
loose objects such as flat rocks, boards or
trash, and screening shrubs or trees with
leaf canopies sparse enough to permit
abundant sunshine to reach the ground.

BEHAVIOR

Undisturbed ringneck snakes are
rarely seen despite their great abun-
dance, because they tend to keep under
cover. However, over the 26-year span
of my observations at the Reservation,
many have been found on a county road
that crosses the area or on the driveway
between the county road and residence
(.5 km). Such a snake, when first seen,
was usually motionless and would re-
main so even when closely approached
by a person on foot, or when an auto-
mobile passed over it. However, after a
pause, especially if actually touched, it
would make violent wriggling move-
ments to gain a place of concealment.
In a highly excited individual there was
sometimes a lateral lashing of the tail as
the snake crawled. In all instances the
snakes on the road or driveway were

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

13

Fig. 5. — Upper, summer habitat of high population of Diadophis in "House Field," with

dense, low vegetation, scattered trees and brush, July 1974. Lower, nursery area, with

sunny, open slope, loose rock, and deep crevices from former blasting, at abandoned

quarry where many females congregated for egg-laying; July 1974.

14

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

i ,r. V:'

>^:.

V

l^'iG. 6. — Upper, liibeniation area at abandoned quarry, with sunny, rock-strewn slope and
fissured limestone outcrop in background; metal strip in foregroimd was attractive to
snakes when warmed by sunshine in cool weather; November, 1973. Lower, "Quarry
Field," providing optimum summer habitat for the snakes, with dense herbaceous vege-
tation, interspersed with trees and clinups of brush.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

15

found by day, sometimes in bright sun-
liglit, leading to the conclusion that they
are mainly diurnal; but, because of small
size and dark coloration, they would not
have been readily detected at night.
Ernst ( 1962 ) observed nocturnal activity
in Maryland.

Behavioral thermoregulation usually
involves contacting sunshine-warmed ob-
jects rather than exposure to direct sun-
light. The snakes use flat rocks, and in
altered habitats use boards and metal
strips. In a horizontal flat coil, the snake
has its dorsal surface in contact with the
undersurface of the shelter, and may be
warmed to above ambient air temper-
ature. When body temperature is suffi-
ciently raised, the snake withdraws from
contact with the sheltering object and
nestles in a depression, retreats into a
hole or crevice, or leaves the vicinity.

The combination of a dark, almost
uniform, somberly colored dorsal surface
and a brightly colored ventrum com-
bines the advantage of cryptic coloration
with the capacity for flashing a startling
display, or making a show of sematic
colors. Musk with semi-liquid feces and
uric acid is voided by a snake that is
captured, frightened or injured. A snake
struggling to escape writhes vigorously,
smearing its captor and its own body
with malodorous voided substances. Si-
multaneously it coils the tail in a tight
spiral, elevated so that the coral red
undersurface is conspicuously displayed
(Figs. 7, 8).

Many genera of snakes, both venom-
ous and non-venomous, have tail displays
(Greene, 1973) that may serve to warn
natural enemies of venomous or distaste-
ful qualities. However, tail-spiralling is
peculiar to Diadophis. The display varies
in the position of the tail, the number of
coils in the spiral, and the kind and de-
gree of stimulation required. Myers (1965)
has shown that the response varies geo-
graphically, and that it is correlated with
intensity of ventral coloring. Those pop-
ulations in which the ventral surface is
all yellow, with no contrast between tail

and body, show no spiralling response,
or have it only feebly developed. Those
populations that have yellow pigmenta-
tion of the body grading into bright
coral red on the tail have the spiralling
response well developed. My own ex-
perience has been mostly with Diadophis
punctatus arnyi of northeastern Kansas,
in which the display is relatively well
developed, but my limited experience
with the subspecies amahilis and occi-
dentalis in California and Oregon has
indicated that in these the response is
stronger and more easily triggered than
in the Kansas snakes.

At the Reservation in October, 1973,
one hundred recently captured ringneck
snakes were tested in the laboratory for
tail spiralling response. Each in turn was
grasped at mid-body with metal tongs,
removed from the container where it
had been burrowing in decaying wood,
and released in a large enamel pan.
Grasping the snakes and releasing them
in an open place excited them and often
evoked spiralling, but if the response did
not occur after about five seconds, the
snake was tapped on the tip of the snout
as further inducement. If the spiralling
still did not occur after several taps, the
snake was held down against the sub-
strate. If the spiralling did not occur
even after several seconds of such re-
straint, a final stimulus was applied in
the form of a light pinch on the rear of
the body with sharp-pointed forceps.

When released 12 of the snakes coiled
their tails with no further stimulus (in
four of these the coil was loose, forming
a loop rather than a spiral). Two more
coiled their tails when tapped; 46 others
did so when held down, but in six of
them the tail was merely looped, and in
12 others the spiral was imperfect or
loose. Fourteen coiled their tails when
pinched (eight of these formed a loop).
The remaining 26 snakes did not coil
their tails in response to any of the
stimuli. However, responses were un-
doubtedly conditioned somewhat by
handling of the snakes in the field at the

16

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Fig. 7. — Upper, adult female ringneck snake and one of her
brood of four, illustrating relatively large size of hatchlings;
August 1974. Middle, adult female and her clutch of four eggs;
July 1974. Lower, clutch of three eggs, one normal (left), one
unusually elongate (liehind), and one with right angle bend,

July 1974.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

17

Fig. 8. — Upper, adult female ringneck snake after capture, coiling tail and

exposing brightly-colored undersurface in "warning" display. Lower, adult

male after handling and release, with tail elevated and coiled in tight spiral

displaying bright ventral colors. Photo by Joseph T. Collins.

18

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

time of capture, and by repeated stimuli
from contact with other snakes when
many were in the same container.

In April 1974, 100 more snakes were
tested in the field when they were first
uncovered, by holding each down mo-
mentarily and noting its responses.
Ninety-five spiralled their tails, another
held its tail in a loop, and only four
failed to show any tail coiling. The
snakes tested were presumably a random
sample and nearly half (48) were adult
males. However, the spiralling response
seems to be equally well developed in
adults and in first year young, and in
both sexes.

Tail spiralling is likely to have
evolved as a defense against diurnal
predators having color vision (especially
birds), and the musk is sufficient deter-
rent to reinforce the display. It may be
speculated that in parts of the geographic
range where tail spiralling is not de-
veloped, other kinds of predators that
are nocturnal, or lack color vision, or are
not deterred by the musk, are most im-
portant.

Although trailing by olfaction is a
prominent aspect of the behavior, and
ringneck snakes tend to aggregate, social
relationships are primitive, and associ-
ations between individuals are ephem-
eral. As in most other snakes, there is
no territoriality, and aggressive behavior
is rare in intraspecific interactions. How-
ever, when many individuals are depos-
ited in the same collecting sack, attacks
may occur. In a group of several dozen
that have been bagged together, there
may be several interlocking pairs, either
young or adults and of either sex. In my
experience the contestants are always of
approximately equal size. The upper in-
dividual has the face and muzzle of the
lower one in its mouth, while the lower
one holds the lower jaw of its opponent,
and because of the sharp, recurved teeth,
the snakes are not readily disengaged.
Anderson (1965) stated that the odor of
disgorged earthworms smeared on the
snakes' bodies caused them to attack

each other, with cannibalism sometimes
ensuing. Possibly this is the sole basis
for the observed attacks.

TEMPERATURE RELATIONSHIPS

Compared with most other reptiles of
northeastern Kansas, the ringneck snake
has a relatively long season of activity,
averaging about 213 days and excluding
only that part of the year in which
freezing regularly occurs. Behavorial
thermoregulation allows it to maintain a
body temperature more nearly optimum
than ambient temperatures. Small size
and slender habitus pemiit rapid adjust-
ment of temperature with little effort on
the part of the snake.

The snakes occasionally bask in direct
sunlight, but usually remain concealed
and elevate body temperatures by con-
tact with sun-warmed objects, such as
flat rocks, logs, strips of bark, leaf litter,
and mats of dead herbaceous vegetation.
Artificial objects, including boards,
scraps of sheet metal, plastic or tarpaper,
are often preferred. The snake lies with
its dorsal surface in contact with the
overlying object.

Body temperatures averaged 26.1°
(11.7= to 34.4°) for 129 snakes found
beneath surface objects. In 109 instances
body temperature exceeded air temper-
ature and the difference ranged from
zero to 13.6°. Only nine per cent of the
snakes (11) had body temperatures that
were lower than nearby air temperatures.
For 37 snakes found in the open, body
temperature averaged 26.6° ± .62 (18.2°
to 33.3°). Air temperatures ranged from
16.2° to 30.0° at the times these snakes
were found, and for the entire group
body temperature averaged 3.7° above
air temperature. Only four of the 37
were cooler than the air.

When air temperature approaches or
exceeds their preferendum, the snakes
tend to stay under shelter, but when air
temperature is relatively low, the oppor-
tunity to bask in sunshine provides in-
centive for them to move into the open.
For instance on 20 April 1955, when air

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

19

•

/

•

/

•

•

• /

UJ
— )

?30

•

.•••

^

■

• /

•

q:

•*»

UJ
Q.

•2^

•

^y

-

UJ

'

•

/

h-

► •

/

/

>-

Q
O

QD

i'°\

/

/

•

-

/

•

1

1

20 25 30

AIR TEMPERATURE

Fig. 9. — Body temperatures of Diadophis ptinc-
tatus and adjacent air temperatures (degrees
Celsius), in snakes found in the open. Body
temperature is usually a little higher than air
temperature and ranges from 19° to 33° in
snakes that are active.

temperature was 16.3°, a basking snake
had body temperature of 30.3°. Below
16° the snakes do not venture into the
open, but still are responsive to vertical
temperature gradients, moving upward
to contact sun-warmed objects, or down-
ward to escape surface chilling. Figures
9 and 10 show body temperatures and
nearby air temperatures for snakes that
were in the open and those that were
beneath surface objects. The majority of
snakes found active on the surface were
between 25° and 27° or between 29°
and 31°. The majority of those found
under shelter had temperatures of 26° to
30°. The highest body temperature,
34.4°, was of a gravid female in June, as
were those of several other records ex-
ceeding 30°. Seemingly the females
while carrying eggs have a strong ten-
dency to bask, and perhaps their prefer-
endum is slightly above that of other
individuals.

The maximum body temperature re-
corded was 34.4°, but the snakes are able
to tolerate considerably higher temper-
atures, at least for brief periods. Experi-
mental exposures demonstrated that for

Diadophis the critical maximum is ap-
proximately 41° and the critical mini-
mum is approximately freezing (Fitch,
1956, 1965). Although the lowest body
temperature recorded was 11.7°, it is
certain that the snakes tolerate lower
temperatures during hibernation. Win-
ter temperatures in hibernacula from .3
to .8 m beneath the surface are usually
well within the range 0° to 10°.

SEASONAL SCHEDULE

On the Reservation ringneck snakes
have been found in every month except
January and February, and only one has
been observed in December. Ordinarily
the snakes become active in the third or
fourth week of March and retire to hi-
bernation in late October or early No-
vember. Some emerge later and retire
earlier. Even those that are potentially
active may be immobilized for days at
a time near the beginning and ending of
their season of activity when weather is
inclement.

10 20

AIR TEMPERATURE

Fig. 10. — Body temperature of Diadophis punc-
tatus and adjacent air temperatures (degrees
Celsius), in snakes found under shelter. Larg-
est circles each represent three records, medium
circles each represent two, and smallest circles
each represent one. Body temperatures are
concentrated at 25° and 30°.

20

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Probably some individuals remain in
the areas where they have been active
and hibernate in such shelters as are
available there, including the tunnels of
voles (Microtus), moles (Scaloptis) and
probably cicada nymphs (Tihicen, Magi-
cicada). Others make trips of hundreds
of meters to hilltop rock outcrops where
there are deep crevices that provide
more secure hibernacula. Emergence
from hibernation in spring occurs over
several weeks and is controlled by tem-
perature that depends not only on
weather but on the location and depth
of the hibernaculum. Fall retirement
into hibernation also spans several weeks
for the population as a whole, judging
from the dwindling numbers of snakes
found active in late October and early
November.

Adult males seem to emerge some-
what earlier than females; anyhow they
arc much more in evidence at first. There
is a shift from rock outcrops to more
open situations during April, and some
shuffling of populations occurs as indi-
viduals that have hibernated together
scatter and mingle with individuals from
relatively remote hibernation sites.
Ovarian follicles, minute at the time of
emergence from hibernation, enlarge
rapidly in early May, and ovulation takes
place in late May or June. Sloughing
occurs in most of the population in late
April or early May. Beginning in early
May the snakes become increasingly se-
cretive and subterranean, and by mid-
summer they are rarely found. Egglay-
ing occurs in late June or early July, and
hatching in the last week of August or
in September. Also in September a com-
bination of cooling weather and heavy
rains cause the snakes to reappear in
abundance in superficial situations where
they have not been in evidence since
spring. In late September and October
there are population shifts back to the
vicinity of hibernacula.

Table 3 shows an incomplete phe-
nology for ringneck snakes on the Reser-
vation. The first tabular row shows both

mean dates for various events in the
annual cycle and mean sizes (length
S-V) of first-year young. The subse-
quent rows are for specific years and
show deviations in days from the mean
date, or deviation in millimeters from
the mean length.

ASSOCIATED ANIMALS

On the Reservation and nearby areas
in northeastern Kansas many species of
animals are associated with the ringneck
snake, and elsewhere in its geographic
ranee the snake interacts with other,
quite different communities. In the area
of my study, earthworms were the ani-
mals most important in its ecology,
providing its main food source. Allolo-
hophora caJiginosa is much more impor-
tant than all other kinds combined, in
terms both of the frequency with which
Diadophis interacts with it and the bio-
mass of food that it provides. This is a
medium-sized, stubby-bodied, tan-col-
ored earthworm that is native in Europe
(Dr. W. Murchie, pers. comm.) and in-
troduced into North America, where it is
now widespread and abundant, occur-
ring in various habitats. It is especially
abundant in damp soil under flat rocks
in open woodland or woodland edge in
the same sorts of situations that are fa-
vored by the snakes.

Other associated animals may be
grouped roughly into those that affect
the ringneck snake indirectly by exten-
sive alteration of the general habitat,
tliose that prey upon it, those that are
potential competitors for food, those that
themselves are potential food sources,
and those that are essentially neutral in
the ecology of the snake. In the first
category are livestock (mainly cattle)
and deer, which alter the vegetation and
soil by their feeding, trampling and
droppings. They may often kill snakes
by accidental trampling. Also, they may
improve the habitat by removing dense
vegetation and creating open areas
where the snakes have more opportunity
to bask. Areas that are subjected to

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

21

a
o

a
u

Plh

PS

<

u

PS

>

o

z
o

H
>

H

en
U

PS

o

H

en

Z

en
<

Z

<

O

!h

H
»-t
en

z

;^

a

H

H

<

s

c

3

b
O

O
O

>J

o

z
u

CO

M
i-l

CQ
<

M
B
3
O
>H

o

>

I

e/3

a «

►2^

■*

►J Ji

t^

e/2

(N

rH

Ml

*..s

Ot

.sj=

<U

feii

C/3

a

Ji

CO

o-

c

a

.s^

►— >

'^n,

CO

Vi

«N

4-1 'O

t« flJ 4)

S S £

^ o S

u

S OX!

CO
CO

c<i

CO

05

<N

O
CD

IC i

i ; o in --H

CO
1 (N P t- CO

1

1 ; CO -H cq
+ + +

1 CO O .-I CO

+ + 1 1

q p

I— I in

CD 00

d in

p CO -<i; tp
^ in iri i-H

00 o

t-" O O 00

+ + + + +

CD

in

copcqppppco^P'-ji-H 05
c<ic<i>-Hdc<jin^i-icocococoo(>i

+ + I + + + + + +

+

t~ 00

CO d

ooincDcD-*cq>— loooq
'^ci-HCDc^icooocvid

c>i

+ + I+ + + + + +

<£>

C^

00

+

+

+ +

^ 00

O CM l>

in <M

CD CO CO

1 +

+ + +

CO in O >-H CD 05

+ I + + + +

Tf< Tfl t^ 00 (M

I + I I +

o

+

CD

CD CD

d in

a>

o o o

1— 1

-^ — 1 t- <N l^

I— 1 i-H

+

1 + + 1 1

CJ

1— I
CO

e

^^
c .2

cs «

+

in -^ o tp i-(

I + I I +

CD l> ^ CO

I + I +

O ^ CO CO

1—1

+ I +

cq,— iin-^in'— icDi— iTt<co
I I + + + + I + + I

CDTfC320^t^OC0005
11+ + + + I + I

cqcoTfiincot^ooasOi— iMco-^incDt^ooosOr-^tsico
inininininininincDcocDcDCDCDcococDCDi^t^i^t;-

05C75O505O5C71C350505C35O5C35050505C35020205O5C3505

pH
II

bO

a

00

05
«

II

bC

a

o

II

a

Si:

eo

o

II

M
B

01

Ml
C

I

IS
K

22

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

medium or heavy grazing support higher
population densities of the snakes than
do ungrazed areas that are otherwise
comparable. Small animals that dig ex-
tensive burrow systems also alter the
habitat by rendering it more favorable
for the snakes, which spend much time
underground yet lack capacity for dig-
ging. The mole Scalopus aquaticus,
chiefly in woodland, and the vole Mi-
crotiis ochrogaster, chiefly in grassland,
provide an abundance of tunnels that
arc used by the snakes. The mole is the
more persistent burrower, but its bur-
rows ordinarily are not open until bro-
ken into by larger animals or by running
water. The main tunnel systems of both
these mammals are horizontal and some-
what less than 10 cm beneath the sur-
face. Burrows of the pocket gopher
Geomys bursarius, pine vole Microtus
pinetorum, and bog lemming Sijnaptomtjs
cooperi may also be used. Much more
available shelters are the cylindrical, ver-
tical emergence tubes of cicada nymphs
(Tibicen pruinosa, Magicicada septen-
decim, M. cassini and others). The tun-
nels are usually somewhat more than 1
cm in diameter, large enough for the
snakes to move and turn in, and provide
convenient retreats to depths where
moderate temperature and high humid-
ity prevail. Ordinarily the tunnels of the
larger Tibicen are most available, but in
the emergence years of Magicicada, such
as 1947 and 1964, the soil is riddled with
tunnels and the availability of shelters
is tremendously increased, especially for
small individuals of Diadophis.

Known and suspected predators on
the snakes are mentioned in the section
on predation. Of these predators, the
mole and the two species of shrew also
feed to a large extent on earthworms
and are potential competitors of Dia-
dophis. Other earthworm-feeders that
might compete with Diadophis are
mostly small snakes including Carpho-
phis vermis, Thamnophis sirtalis (espe-
cially first-year young), Storeria dekayi,
Virginia valeriae and Tantilla gracilis.

All of these are much more specialized
than Diadophis in their habitat require-
ments, but overlap Diadophis. None
approaches the high population densities
attained by Diadophis, which is usually
the dominant kind even where it occurs
in association with one or more of the
others.

Tantilla gracilis and Virginia valeriae
are remarkably small snakes; young or
even adults are potential prey for Dia-
dophis. In Arizona Tantilla icilcoxi has
been recorded in the food of Diadophis
(Van Denburgh, 1922:654). A hatch-
ling Ophisaurtis attenuatus is the only
lizard recorded in the prey on the Reser-
vation, but hatchling skinks (Eiimeces
fasciatus and Leiolopisma laterale) are a
potential food source, as are newly meta-
morphosed anurans Acris crepitans,
Pseudacris triseriata, Bufo americanus,
and Gastrophryne olivacea.

FOOD HABITS

In my study, made near the center of
the species' geographic range, the trend
of the food habits differed notably from
the trend indicated by published litera-
ture, or by any single published account.
Significant geographic variation in food
is evident, and seems to be correlated
with parallel changes in size, bodily pro-
portions, habitat and behavioral traits.

Ernst (1962:266), in a brief popular
account based mainly on observations in
Maryland, wrote that earthworms, small
salamanders (especially Plethodon cine-
reus), slugs, giiibs, newborn snakes,
toads, frogs and various insects were
eaten. He stated, "I have observed ring-
necks capturing winged insects that have
fluttered to the ground after being at-
tracted to a night light." Myers (1965:
63) obtained stomach contents of 74 of
the snakes from Alachua County, Flor-
ida; 26 had earthworms, one contained a
salamander (Manculus quadridigitattis),
four contained narrow-mouthed frogs
(Gastrophryne carolinensis), two had ju-
venile frogs (Rana pipiens) and another
contained an unidentified frog, two con-

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

23

tained ground skinks (Leiolopisma lat-
erale), and one contained unidentified
reptile scales. Myers also recorded a
salamander, Eurycea bislineata, from the
stomach of a ringneck snake from Leon
County, Florida. In northern Michigan,
Blanchard, Gilreath and Blanchard (in
press) found from the combined evi-
dence of food disgorged by captured
snakes, and from experimental offering
of possible prey species to those in con-
finement, that the salamander Plethodon
cinereus was preferred and was preyed
upon much more often than any other
species, although earthworms and hatch-
ling snakes were also taken.

At Big Black Mountain in eastern
Kentucky, Barbour (1950:104) obtained
a sample of ringneck snakes (Diadophis
punctatus edwardsi) and dissected 53, 18
of which contained prey. It consisted
chiefly of the salamanders, Plethodon
glutinosus, Desmognathus fuscits (both
adults and larvae), Eurycea bislineata
and Aneides aeneus (eggs). However,
wood roaches were 11.1% in frequency
and 7.5% in volume. Other insect frag-
ments and miscellaneous debris were in-
terpreted as stomach contents of the
salamanders.

Gehlbach (1974:144) stated that in
Texas ringneck snakes (both D. p. arnyi
and D. p. regalis) feed mainly on small
reptiles. In laboratory tests he found
that they would catch and hold small
snakes such as Virginia and Sonora and
would chew or cling to the site of the
original bite, sometimes for several hours,
until the prey was immobilized, seem-
ingly killed or paralyzed by the toxic
saliva, conducted into the wound by the
enlarged posterior maxillary teeth.

Most earlier literature was summa-
rized by Wright and Wright (1957) who
discussed food habits separately for each
of nine kinds of Diadophis currently
considered conspecific, but recognized as
valid subspecies. They are amahilis,
modestus, occidetitalis and similis of the
West Coast, edwardsi and punctatus of
the eastern United States, laetus and

regalis of the arid Southwest and arriyi
of the central states. In discussing their
food habits Wright and Wright cited the
publications of 28 earlier authors. In
most instances the food records for each
kind of ringneck snake were obtained by
a different group of authors. Of the nine
subspecies figuring in the records, seven
were reported to have eaten insects,
seven had eaten snakes, five each had
eaten lizards, salamanders, frogs and
earthworms, two had eaten toads and
one had eaten slugs. The subspecies
arnyi, which includes the population on
which my own study is based, had eaten
prey of all these categories, excepting
slugs.

In summary the literature records
indicate that Diadophis punctatus is
euryphagous; its known food includes
diverse types of small terrestrial animals
of four phyla. Therefore, it is remark-
able that a sample from the Reservation,
much larger than any assembled else-
where, indicates feeding almost entirely
limited to one kind of prey. Stomach
contents were palped from each of 182
snakes captured on the Reservation in
14 different years from 1951 to 1974.
The 224 items included 221 earthworms,
two tipulid larvae and one hatchling
glass lizard (Ophisaurus attenuatus).
Many of the earthworms were identified
as Allolohophora caliginosa, and prob-
ably most were of this species.

A total of 347 fecal droppings were
collected from the ringneck snakes
processed in 1959, 1960 and 1961, and
were examined under a dissecting micro-
scope. In 56 of these "scats" there was
nothing identifiable, but in each instance
the material seemed to consist mainly or
entirely of soil which might have been
residual from the digestive tracts of
earthworms after the soft parts had been
dissolved. In 291 scats some hard parts
of the prey remained, providing a clue
to its identity; 288 contained only setae
of earthworms, three contained frag-
ments of chitin as residue from insect
prey, and two contained both setae and

24

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

chitin fragments. Traces of plant mate-
rial occurred in 177 scats. Some of this
vegetation was still green but most of it
was old and decayed; it might have
come from the guts of earthworms, or
might have been ingested accidentally
as the snake swallowed prey.

Miscellaneous materials found in
scats included masses of nondescript soft
tissue (probably from earthworms in
most instances), crystals of uric acid,
teeth of the snakes themselves, soil par-
ticles, and (in two instances) fine hairs
or hairlike fibers.

Each of 170 snakes checked for stom-
ach contents contained one earthworm
and nine snakes each contained two
earthworms. Probably in many instances
the food that happened to be in the
stomach was undetected when the snake
was processed. The elongate shape of
an earthworm usually caused no discern-
ible bulge in the body of the snake. The
part of the prey in the posterior end of
the snake's stomach disintegrated first,
and palpation did not readily reveal
presence of the object. In April 1973
special effort was made to detect food in
the stomach or fecal material in the hind
gut. Of 83 snakes processed on 6, 11, 12,
13, 17 and 18 April, 34 produced scats
when the rear end of the body was
palped and 17 had stomach contents; 9
of these snakes yielded both stomach
contents and scats. Thus 42 of the 83
snakes had food or food residue in the
digestive tract.

In 70 instances in which both the
snake and its earthworm prey were
measured, the worm averaged 34.1 ±
1.58% (14 to 97) of the snake's length
(S-V). In 45 instances in which the in-
tact earthworm was weighed, it averaged
11.9 ± 0.87% of the snake's weight.
Such relatively small prey might be ex-
pected from the minute head, delicate
jaws, and attenuate body of the snake.
Frequency of feeding is no doubt greater
in these snakes than in those which in-
gest relatively bulky prey.

Henderson (1970) performed a series

of feeding experiments using snakes from
the same population that I studied.
Starting with snakes that had empty di-
gestive tracts, he found that the first
defecation occurred in as little as seven
hours after feeding in snakes maintained
near their maximum temperature toler-
ance, at 35°, and that first defecation
ordinarily occurred in from 15 to 30
hours in snakes kept at 24° to 33.° Di-
gestion can occur even at low temper-
atures; in one snake kept at 7° defecation
occurred after 312 hours. Henderson
found that after an earthworm meal
there were from two to six (average 4.2)
successive defecations, with four occur-
ring most frequently. At 33° last defe-
cation occurred only two or three days
after ingestion, and at 25° it occurred in
four or five days.

Combined evidence of Henderson's
study and my own field data suggest that
the average ringneck snake feeds about
once every eight days, that about half
the interval is required to digest and
assimilate the meal, and that the snake
has an empty digestive tract for the re-
maining half. At this rate of feeding a
snake would take about 27 meals, ingest-
ing approximately three times its body
weight, in a normal growing season esti-
mated at 213 days.

SPATIAL RELATIONSHIPS

In the course of routine field work,
numerous snakes captured for processing
were generally deposited in the same
container without noting the precise cap-
ture points. For these, distances trav-
elled between captures were usually not
known, but, as recapture records accu-
mulated, it became evident that each
snake had usually stayed within the
same small study area such as the quarry,
Quarry Field or House Field. Likelihood
of recapture was greater in those that
stayed in the study areas than in those
that wandered away from them, but
spacing of the study areas was well
adapted to show relatively long move-
ments, when a snake shifted from one

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

25

study area to another, and such shifts
were recorded occasionally.

Some of the capture records that
were obtained involved precise stations
such as those occupied by funnel traps
at the ends of the drift fences, and pieces
of sheet metal put out to attract the
snakes. When two or more of the cap-
ture records for any individual involved
such definite field stations, the distances
between them could be readily measured
on the map. In all, 433 such measure-
ments were obtained and these averaged
80 m (0 to 1700). In 86 instances the
distance was recorded as zero, but the
actual capture points may have been as
much as several meters apart. Table 4
shows the trend toward greater distances
with elapsed time. In general, the longer
the time, the farther the snake would be
from any given starting point. However,
each snake tends to revisit certain points
from time to time, and to turn back into
a known area instead of wandering in-
definitely.

The 433 distances measured include
several types of movements such as day-
to-day foraging for food, gradual or
abrupt shift from one favored area to
another, and regular seasonal migrations
involving travel to and from hibernacula
or egg-laying sites, but in most cases
these cannot be distinguished. In order
to determine the usual distance of travel
between hibernacula and summer ranges,
the records of individuals captured in
March or October (presumably near the
place where hibernation had occurred or
would occur) and caught also between
late April and mid-September (presum-

ably where they were spending the sum-
mer) were examined. Fifty such move-
ments averaged 121 m and half of them
were in the range 67 to 174 m. Many
other snakes, caught in early April or
late September, were probably near their
hibernacula and their relatively long
movements between these sites and their
summer ranges raise the average some-
what.

Figure 11 shows the recorded move-
ments grouped in 10 m intervals. It
shows that nearly one fourth of the
movements were distances less than 10
m and one third more were between 10
and 70 m and 22 per cent were between
70 and 140 m, with the remainder longer
distances up to 1700 m. The histogram
reveals a trend toward progressively
fewer records for longer distances. There
seem to be rather abrupt reductions be-
yond distances of 10 m, 70 m and 140 m,
respectively, and these steps may have
biological significance. The much larger
number of records between 0 and 10 m
than in any other 10 m interval, espe-
cially, indicates attachment to small
areas. The significance of these many
short movements is believed to be, in
part, the slow rate of travel of the snakes,
so that one recaptured after several days,
or perhaps even after several weeks,
often would still be lingering in the vi-
cinity of its previous capture, without
having covered much of its actual range.
Also, habitual return to a favorite spot
within the home range, even after long
intervals, must have been involved. In
many instances a snake was recaptured

Table 4. — Distances Travelled by Marked and Recaptured Ringneck Snakes in Various

Time Intervals.

1 to5
Days

6 to 10
Days

11 to 20
Days

1
Month

2 to 6
Months

7 to 18
Months

More Than

18 Months

Number of Records .

Average Distance
(meters) --

... 46
... 80

43

106

0 to 444

38
134
0 to 600

38
250
0 to 700

66
300
0 to 1200

120

310

0 to 170C

69
447

Range of Distances
Travelled (meters)

0 to 820

1 0 to 1550

26

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

100 200

DISTANCE OF MOVEMENT:

300

METERS

Fig. 11. — Distances moved by individual ringneck snakes between captures.

at approximately the same place after a
year or more.

Movements up to 70 m are numerous
and their numbers are near the same in
each 10-meter interval. These probably
represent normal movements within a
home range and suggest that a nomial
home range may have a diameter of
approximately 70 m. Between 70 and
140 m, records are less numerous than
they are for shorter distances and tend
to decrease gradually; beyond 140 m
there is an abrupt decrease. It is sug-
gested that the many recorded distances
between 70 and 140 m were mostly
within the home ranges of the individ-
uals concerned, and that they reflect un-
usually large or elongate ranges. The
relatively few distances exceeding 140 m
are probably outside the snakes' normal
ranges and represent shifts or extensions
of original home range, or perhaps small
scale migrations between summer ranges
and hibernacula or egg-laying sites.

In other kinds of snakes it has been
found that the kind and amount of travel
and size of range differs greatly accord-

ing to the sex and age of the individual.
Undoubtedly such differences exist in
Diadophis also. However, the sample
obtained was heavily weighted to adult
males (58 per cent) with relatively few
adult females (11%) and first-year young
(both sexes combined 17%) and even
fewer second-year young (both sexes
combined 14%). Average distances were
44.4 m for first-year young, 55.0 m for
adult males and 102.1 m for adult fe-
males. The proportion recaptured at the
original location differed significantly:
43.8% in first-year young, 17.7% in adult
males and 11.7% in adult females.

The finding that first-year young are
much less vagile than adults is in line
with what is known about other snakes,
and animals in general. It was surprising
to find that adult females averaged
nearly twice as much distance as adult
males, and were relatively seldom re-
captured at an original capture site.
Probably their greater vagility is corre-
lated with their larger size, and with the
fact that at least some of them travel to
and from special nesting areas.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

27

I

REPRODUCTION

Diadophis punctatus is oviparous and
its reproduction is on an annual cycle.
Ovarian follicles are minute at the time
of emergence from hibernation, they en-
large rapidly in late spring, ovulation
occurs in June and egg-laying occurs in
late June or early July. Incubation is
mainly in July and August and hatch-
lings usually appear near the end of
August or in September.

Aggregation and Mating. — Trailing
by olfaction brings the sexes together
and increases opportunity for mating.
Dundee and Miller (1968) showed that
the small aggregations of ringneck snakes
frequently found under sheltering ob-
jects in spring result from following of
the scent trail left by contact of the skin
with the substrate. When one individual
has found a shelter, others may follow it
there. Aggregations are most prominent
in spring when surface activity is at its
peak, but are evident also in fall. Al-
thouo;h the snakes are seldom seen in
summer, they probably aggregate then,
but in deeper and better insulated shel-
ters than are used in spring and fall. The
females at least in some instances use
communal egg-laying sites.

The aggregations found in spring in-
clude both adults and immatures, males
and females. There is some tendency for
the first-year young to aggregate with
others of their size segregated from
adults. However, in general adult males
make up a high proportion of the indi-
viduals in aggregations, which probably
form partly as a result of sexual search.
Skin odors are the basis for trailing
(Noble and Clausen, 1936; Dundee and
Miller, 1968) but it is not clear how
much discrimination is exercised by the
trailing individual in distinguishing sex
and reproductive condition of the one
being followed. Musk from the scent
glands is secreted only when the snake
is frightened or injured, and its odor in-
hibits trailing.

Sexual behavior is inhibited by cap-

ture and handling and unnatural sur-
roundings, and has never been observed
in the many ringneck snakes kept in
confinement at all times of year. Even
under natural conditions sexual behavior
has rarely been observed, and this is re-
markable in view of the great abundance
of the snakes. Active courtship was ob-
served on 24 September 1959 at 4 P.M.
The female involved was in process of
sloughing and was crawling slowly
through the surface litter in open wood-
land, partly exposed. The slough had
peeled back for about half the length of
her body. A male was beside her, facing
in the opposite direction. For several
minutes that the snakes were watched,
the male alternately lay passive or made
jerky, animated courtship movements,
obviously stimulated by scent emanating
from the moist, newly-exposed skin. The
snakes moved out of sight beneath leaf
litter, and when it seemed that they
would not reappear, leaves were gently
brushed aside exposing the pair to view.
Meanwhile another male had joined
them and both males pressed against the
female but the anterior ends of all three
snakes were concealed. Posterior ends
of the two males were momentarily in-
tertwined and were apart from the fe-
male. There was no evident rivalry be-
tween the males as each pursued its
courtship, but confusion and delay re-
sulted from the presence of the extra
male. Soon the snakes again moved out
of sight and could not be relocated.

Copulation was observed at approxi-
mately 5 P.M. on 7 May 1974. The pair
of snakes was found in damp soil be-
neath a heavy decaying plank. They did
not separate when exposed by turning of
the plank nor for about 20 minutes sub-
sequently that they were kept under
observation. The female was nervous
and active, darting out her tongue al-
most continually, shifting position fre-
quently, and sometimes crawling a short
distance, dragging the relatively passive
male after her. From time to time the
snakes coiled together in a fairly com-

28

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

pact mass, the male often pressing his
snout against the female's side. After a
few minutes' gap in the observations, the
snakes were found to have disappeared
and could not be relocated.

Reproductive Cycles. — In four differ-
ent years on dates representative of the
entire season of activity, a total of 484
adult males were tested for motile sperm
by gently palpating the rear of the ab-
domen to eject renal and fecal wastes,
then pipetting a drop of physiological
saline into the cloaca. By gentle back-
ward massaging of the posterior part of
the abdomen, semen was pressed from
the vas deferens into the cloaca, and
into the droplet of saline solution, which
was expressed from the anus onto a glass
slide and examined. Active sperm were
found in 98.0 per cent in April ( N =
204), 95.4 per cent in May {N = 65),
91.0 per cent in September (N := 165),
and 93.4 per cent in October (N = 30).
Of 19 males tested in June, July and
August, only one lacked motile sperm.
Negative results did not necessarily indi-
cate that sperm were absent; they may
have been overlooked. The tests indicate
that most adult males have motile sperm
throughout the year and that insemina-
tion might occur at almost any time,
providing males have sexual drive and
females are receptive.

In adult females sperm can survive
for relatively long periods in vascular-
ized pouches of the oviducts, but prob-
ably their maximum survival in the clo-
aca is only a few days at most. In a few
instances tests revealed massive clumps
of sperm, but more often the numbers
were small, and at all times of year the
majority of females tested were negative.
The 625 females tested had the follow-
ing percentages with active sperm in
different months of the season of activ-
ity: 0 in March (N = 28), 5.8 per cent
in April (Nr=197), 20.2 per cent in
May (N= 178), 16.4 per cent in June
{N = 61), 0 in July (N = 30), 0 in Au-
gust (N = 16), 7.9 per cent in Septem-

ber ( N = 63 ) , and 2.5 per cent in Octo-
ber (N = 80).

On 1 May 1966 seven large females,
all negative for cloacal sperm, were dis-
sected. All but one had quantities of
active sperm in the oviducts. Ovarian
follicles were only beginning to enlarge
(largest 6.0 to 8.5 mm, much short of the
size at ovulation which is usually 25 mm
or more). On 24 April 1955, six females
were dissected and they contained an
average of 6.5 small ovarian follicles
ranging from 1 to 6 mm in length but
with a distinct bimodality in size. The
larger, clustered about a length of 5 mm,
were probably destined for ovulation in
the current season, while the smaller
clustered about a length of 1 mm, prob-
ably would have been retained until the
following year.

In May follicles enlarge rapidly and
ovulation occurs in June. In eight differ-
ent years 86 females captured in June,
their bodies distended with eggs, were
held in confinement until deposition of
their clutches. Average egg-laying date
for this group was 3 July ( ± .87 day ) ,
but dates extended from 19 June to 24
July with a standard deviation of 7.25
days. Much of this variance resulted
from the difference between years; egg-
laying and other phenological events in
the annual cycle of the local population
may be accelerated or retarded by un-
seasonably mild weather or cold weather
at critical times in spring.

On 6 August 1965 a female of 319
mm (S-V) was found to have three
shelled eggs in her oviducts and the eggs
contained living embryos larger (3.3 mm)
than those usually found in freshly laid
eggs. This date, 34 days later than the
average oviposition, indicates a spread
of more than six weeks in egg-laying and
explains the presence of hatchling sized
snakes in late spring when most have
grown well beyond this minimum size.

Peterson (1956:152) noted a remark-
able instance in which a female Dia-
cJophis punctatus stictogenys from the
Miami area of southern Florida gave

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

29

birth to a litter of six living young. In
this case the eggs were retained in the
female beyond the usual time of laying.
In newly laid eggs of Diadophis the
embryos are somewhat more advanced
than are those of many other kinds of
oviparous snakes.

Eggs. — The eggs are elongate and
remarkably variable in shape (see Fig.
7). Often they are nearly cylindrical, but
in some the ends tend to be pointed. In
some both ends are similar but in others
one end is more enlarged. Typically the
eggs are somewhat curved, and sausage
shaped.

In 251 eggs of 72 clutches, width
varied from 21.7 to 46.8 per cent of
length. The length averaged 25.2 ± .277
mm (17 to 43). Width averaged 7.11 ±
.045 mm (5.2 to 10.0). Weight averaged
.89 ± .001 grams (.58 to 1.35).

The eggs are white, but the soft, thin
shells are partly translucent so that the
coloring of the yolk is faintly visible.
The eggs are firm and turgid when laid
and as incubation proceeds they gradu-
ally absorb moisture from the surround-
ing medium, and eventually they almost
double their original weight. As liquid
is absorbed, the eggs become increas-
ingly turgid by stretching of the shell,
and the shape changes slightly, with
more increase in girth than in length.

Besides the eggs of normal appear-

ance, females kept in confinement in late
June and early July sometimes laid eggs
that lacked shells and were inviable.
Often one or more such eggs were laid
in a clutch along with normal eggs. The
inviable eggs averaged slightly smaller,
and are flaccid rather than firm. Clark
(1970) in his study of Carphophis also
found that confined females often pro-
duced unshelled eggs, and he concluded
"Shell-deposition may be a crucial step
in development, such that a gravid fe-
male collected and confined prior to shell
deposition generally reacts by laying her
eggs in their premature condition." In
Diadophis likewise, eggs that lack nor-
mal shells are believed to result from
handling and /or confinement of the fe-
males at a critical stage in their egg
production.

Size of Clutch. — For 300 clutches,
mean was 3.89 ± .083, with a range of
one to ten. Data were obtained from
unlaid eggs, either found in females dis-
sected, or palpated in live individuals,
and from eggs laid by females isolated
in confinement. Some of the more im-
portant factors affecting size of the
clutch are size (and/or age) of the fe-
male, the year, and the stage of develop-
ment at which counts are made, since
apparently some eggs are eliminated be-
tween the time of ovluation and ovi-
position.

Table 5. — Number of Eggs per Clutch and Relative Productivity of Females of

Different Year Classes.

Length

(S-V)

of Females

Most Probable
Year-class
of Females

Number of

Ovigerous

Females

in Sample

Mean
Number

of Eggs

Range

Percentage
of Females

in Each
Year-class

Percentage

Eggs Produced

by Each

Year-class

210"- 256

3rd

50

2.70 ± .160

1-5

26.6

20.0

257 - 272

4th

82

3.23 ± .109

1-6

21.4

19.0

273 - 280

5th

46

3.67 ± .155

2-5

12.1

12.4

281 - 287

6th

25

3.68 ± .208

2-6

8.8

9.0

288 - 294

7th

28

3.93 ± .210

2-6

7.6

8.2

295 - 300

8th

30

4.17 ± .225

2-10

5.1

5.9

301 - 305

9th

12

4.25 ± .392

2-6

4.1

4.8

306 - 310

10th

16

5.06 ± .225

4-7

3.8

5.3

31 1 or larger 1 1th or older

50

5.26 ± .310

3-8

10.5

15.4

I

* Most females that are fecund or have reached their third year are 235 mm or more.

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

-10

CO

CD
UJ

cr
lij

CD

34

39

-5

15 28

50 34

65

10

24

'

4

^

3

I'

J-

O.

J.

Fig.

250 300 350

SNOUT VENT LENGTH IN MILLIMETERS

12. — Number of eggs per clutch in female ringneck snakes, showing relationship of

clutch size to length of female. Mean, standard error, standard deviation and range are

shown for each of the larger series.

Eggs laid averaged 3.36 ± .153 per
clutch (N = 50), whereas those palped
or dissected averaged 3.96 ± .096 ( N =
267). These figures imply that 15 per
cent of the eggs are eliminated between
ovulation and oviposition. However,
most of the records of eggs laid were
obtained in 1965 when clutch size may
have been smaller than in other years.
In 1974, 53 unlaid clutches that were
palped averaged 3.81, and 17 clutches
that were laid averaged 3.47. All the
females that oviposited in captivity had
been palped two or three weeks earlier
and some loss may have resulted.

In Table 5 females are grouped on
the basis of size according to most prob-
able age, and it is shown that the mean
size of clutch increases from 2.37 in
probable third-year individuals (produc-
ing first clutches) to 4.95 in those that
are probably seven years old or more.
In Figure 12 females arc further divided
according to length with a 10 mm in-
terval; a steady increase in size of clutch
according to size of female is indicated,
except that females in the 290-299 mm

class averaged higher (4.74) than those
in the 300-309 or 310-319 classes (4.33
and 4.65). Seemingly size, as such,
rather than age determines productivity.

Table 6 shows mean numbers of eggs
per clutch in different years, arranged in
order from the largest clutches (in 1963)
to the smallest clutches (in 1965), along
with weather factors that might be ex-
pected to affect fecundity. Amount and
distribution of precipitation affect avail-
ability of earthworms, which provide
most of the food. Ova are minute when
the snakes emerge from hibernation in
March. Therefore, availability of food
in the spring period of April, May and
the first half of June might be expected
to control the development of eggs, and
the availability of earthworms as food
would depend on both the amount and
distribution of moisture. In Table 6 the
number of days having precipitation and
the amount of precipitation during the
supposedly critical spring period and
during the 11/2 months preceding the
time of ovulation are shown. There is no
evident correlation. In a study of Car-

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

31

Table 6. — Fecundity and Weather.

Mean Number

of Eggs

Per Clutch

Year

Days With Rain,
April, May,
Early June

Precipitation
April, May,
Early June

Precipitation

July Through

Early June

5.15 iN=l4)

1963

23

180

897

4.70 (N=17)

1962

18

168

956

4.43 (A^=14)

1961

25

281

946

4.23 (N=30)

1966

25

280

827

4.00 (N=14)

1973

18

231

1340

3.97 {N=60)

1969

16

326

1070

3.83 (iV=12)

1960

21

236

922

3.75 (N=eO)

1974

27

273

1195

3.43 (N=5l)

1970

23

396

900

3.23 (N=23)

1965

25

405

1117

pJwphis vermis in the same general area,
Clark ( 1970 ) found seemingly signif-
icant correlation between precipitation
and size of clutch. Like Diadophis, Car-
phophis is an earthworm eater.

Reproductive Effort. — The amount of
energy and material used in producing
eggs or young and the frequency of such
reproductive effort differs greatly in dif-
ferent kinds of snakes. Also it differs
intraspecifically, according to the age
and size of the female (Fig. 13). More
information is now available for Dia-
dophis punctatus than for any other kind
of snake.

In 248 eggs of 71 clutches, weight of
egg averaged 0.890 ± .01 gram, and
eggs that seemed normal varied from
0.58 to 1.35 gram. Clutches averaged
3.89 eggs, hence average clutch-weight
is calculated to be 3.47 grams.

Most typical adult female weights
can be obtained at the end of the grow-
ing season when females are not gravid
and those that produced clutches in
early summer have had time to recuper-
ate. In a composite September sample
from 1965, 1966 and 1969, average
weight of adult females was 4.26 dz .11
grams ( N = 72 ) . From these samples it
is calculated that weight of the newly-
laid clutch averages 81.4 per cent of the
non-breeding weight of females.

Lower ratios of weight of clutch to
weight of snake were obtained by weigh-

ing individual gravid females that were
almost ready to lay, then reweighing
them and their newly-laid clutches sep-
arately. In a series of 52 females, clutch-
weight was 39.4 =i= .88 per cent of pre-
laying weight. Clutches ranged from
20.8 to 51.1 per cent of pre-laying weight
of snake. To determine whether repro-
ductive effort was affected by size of
female, the sample was divided among
large (S-V 295-340, N = 19), medium

^^^

^^^

o

-6

«

•

•5

o

o

0

o

o

o

o

CD

5

-4

o

fi

8

•

•

,

<

8

o

•

o

o

1-

-3

o

o

f

B

■

X

O

8

•

o

o

o

0

LlJ

ft

o

o

•

8

5

-2

0

o

0

•

•

-1

250 300 350

SNOUT VENT LENGTH IN MILLIMETERS

Fig. 13. — Weight of clutch in relation to length

of female; solid circles show means and small,

open circles show individual records.

32

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

(S-V 275-294, N=12) and small (S-V
225-274, N = 2l) females. Clutch-
weight (as percentage of snake-weight)
differed as follows in these three groups:
large, 40.9 ± 1.10 per cent; medium, 39.1
± 1.69 per cent; small, 38.1 ± 1.67 per
cent. These figures suggest that repro-
ductive effort is greater in larger females
than in smaller, but the difference is
relatively small.

Even with equal reproductive effort
the larger females produce more eggs
than the smaller. Table 5 shows the
quota of eggs calculated for each annual
age group. It shows that because of
their relatively large numbers third year
females produce the most eggs (one-
fifth of the total ) , despite their relatively
low individual productivity. Because of
declining numbers, successively older
age groups produce slightly fewer eggs,
although individually the snakes become
more productive.

Nests. — Eggs were rarely found un-
der natural conditions, and most of those
observed were obtained from gravid fe-
males held in confinement. However, on
several occasions eggs were found in
natural nests in damp soil beneath large,
sunken flat rocks. At the Reservation
headquarters on 3 August 1970, a com-
munal nest was found by workmen tear-
ing down a brick structure enclosing an
electric water pump. The brick wall was
cracked and crumbling as a result of
weathering; 15 eggs, probably represent-
ing four clutches, were found in the
cracks. These eggs were partly incu-
bated and the females had left the area.

Probably communal nesting is the
rule, judging from the fact that large
concentrations of hatchlings were found
year after year at certain favored spots.
Such concentrations were found in Sep-
tember, October and even April; mean-
while, hatchlings were relatively scarce
in the areas where most adults were
found. The 93 snakes obtained 1-15
April 1973 were from three locations,
each showing a characteristic biased ra-
tio of hatchlings to adults (and second-

year young) as follows: nesting area at
east end of quarry (N ^ 40), hatchlings
78%; rocky upper slope near hibernacula
at "Tomb Rock" 165 m southwest of first
location (N = 28), no hatchlings;
"House Field" (N = 25), hatchhngs 16%.

The nesting area at the east end of
the quarry was the best known one, and
it must have been used by dozens of
females from the adjacent field, judging
from the numbers of hatchlings found.
The actual nests were not found, as
search for them would have damaged
the habitat. Numerous potential nest
sites were provided by niches and deep
crevices in the limestone outcrop shat-
tered by quarrying operations, with
blasting, more than 25 years before, and
by the rock-strewn slope below it. A
second nesting area was near the middle
of the old quarry, approximately 60 m
southwest of the first and perhaps not
entirely distinct from it. Two other im-
portant nesting areas were known at Rat
Ledge.

These nesting areas were all situated
along hilltop outcrops of the Oread
Limestone, in relatively open situations
of southward exposure. Another nesting
area was in bottomland approximately
245 m west and a little south of the
Reservation headquarters, where the
driveway crossed a gully. An old and
weathered stone check-dam, a cracked
concrete slab, and a rock fill provided
shelters with favorably moist nesting
sites, but there was some risk of flooding
when the gully filled with run-off water
after unusually heavy rains.

Although hatchlings tend to be con-
centrated in preferred nesting areas,
many others that seem to be recently
emerged are scattered in open habitats
where the adults spend most of their
time. Probably some of the females find
nest sites without shifting from the areas
where they live. Abandoned burrows
such as those of prairie voles and cicadas
are abundant in the snakes' habitat, are
often used by the snakes, and on occa-
sion may serve for egg-laying sites.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

33

GROWTH

Growth was studied by ( 1 ) direct
measurement of gains in length and
weight of individuals marked and re-
captured, and (2) average differences in
length and weight between series in dis-
crete annual age groups. Both methods
are fallible and need to be used with
caution, and in combination.

The growth in marked individuals is
not typical of that in the entire popu-
lation because these small and delicate
snakes are adversely affected by routine
processing and fail to grow normally for
an indefinite period afterward. Of 55
recaptured after one to ten days, seven
had gained weight and 48 had lost with
a mean loss for the whole group of 5.44
per cent. In those recaptured after 11 to
20 days five had gained weight and 17
had lost with a mean loss for all of 5.36
per cent. In 21 to 30 days four of an-
other group had gained weight and 13
had lost with mean loss of 11.1 per cent.
Of 18 caught in spring and recaptured
in fall after intervals of 139 to 197 days,
12 had gained and six had lost with a
mean average gain of 8.4 per cent. Each
sample consisted of different individuals
and all were fully adult so that growth
in length had stopped or greatly slowed.
From these records it can be concluded
that the deleterious stunting effects con-
tinue to be severe as much as a month
after handling, but that by four to six
months these effects have largely worn
off. The direct cause of stunting is not
definitely known but perhaps handling
and processing causes some shock, ag-
gravated by bruising or tearing of deli-
cate tissues when the snake is grasped
firmly and stretched to full length for
measurement, and results in a period of
fasting.

In any one year most eggs are laid
within a few days in midsummer and the
hatchlings that emerge constitute a co-
hort which at first is readily distinguish-
able from older and larger young. How-
ever, some of the hatchlings, although
seemingly normal in most respects, are

much undersized whereas others are
much larger than the average. These
differences increase with elapsed time
and by late spring some of the smallest
young are still of a size typical for hatch-
lings whereas at the other extreme the
largest young of the year overlap in size
the smallest second year young. There-
fore, although the population consists of
discrete annual age groups, there is a
continuum in size from the smallest to
the largest individuals at most times of
year. Throughout the first year of life
most young are recognizable as members
of their cohort. Only a small percentage
overlap in size those that are in their
second year. Second year snakes are
usually recognizable as such, but their
size-range has increased and they are
overlapped in size by both first- and
third-year individuals.

The typical course of early growth is
revealed best by comparing mean size
of series of young from different times
of year with the size of newly hatched
individuals to determine the amount of
gain in a given time. The course of later
growth is revealed best by the records of
individuals marked and measured early
in life and recaptured after intervals
sufficiently long that stunting from the
original handling had been overcome
and was a relatively minor factor.

At the time of hatching the sexes are
already somewhat different in size and
proportions; the females are larger with
shorter tails. As development proceeds,
the differences become more pro-
nounced, with females attaining larger
average and maximum sizes. Therefore,
data for the sexes need to be separated.
In some years hatchlings are not found
in substantial numbers until September
or even October after the surface soil has
been moistened by heavy rain; probably
such individuals are already several
weeks old and they have had time to
grow, utilizing egg yolk still unassimi-
lated at the time of hatching, and then
beginning to feed. Thus, the normal

34

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

a
<
u
>^

a

z
o

u
u

Q
Z
<

H

en
CS

CO

<

o
z

o

(^
o

c«
H

O

a
o
u

o

en
S
H
O
Z
U

z

PQ
<

O

CO

CD

'X

iX

'X

o
I—

05

iX

en

CD
05

iX

00
05

iX

l^ 0)

CD -w

(35 "55

—I o

ti &

in G

05 O

'X

"^ o

t^ 00

CO ^

c<i cq

+1 +1

Ol 00

oi in

CO <M

l-H ^

Tf CO

00 o

+1 +1
o (>i

CO

+1

p
oq

CO

+1

CO

XT in CO l-H

r-( '^ t- (N

00 O 00

a; CO CO

+1 II

in in

d CO

+1
o

o
d

C<1 Ol CM Ol

CO in

O '^

tT .-h

■-H oi

+1 +1

00 Ol

05 t^ CO X -1*
CO (M ^H -H (M

O ^H CO ^

r- TT TT o
— I oi Ol oi

+1

00 t-

CO ^

o in Ol

CO 05 00

+ 1 +1 +1 +1

•—I CO 00 in

^ t- CO -H
CO CO CO "^

CO
Ol

in

I> 00

^ CO

II +1

q ^

CO c^

— ( Ol

^ Ol

+1 +1

p Ol

-^ oi

Ol Ol

in

Ol

» o

00 ^-^

i-H oi

+1 +1

CD in

^ in

in CO o

00 00 Ol

01 p p

CO ^ in

+1 II +1

p O; t-;

CTJ oi ai

Ol CO Ol

Tf O --H

in 05 CO

0 00 in

01 00 oj

•— i " oi

II II II

p p p

^ oi d

Ol Ol CO

m o

Oi rr

CO CO
1—1 CD

II II

in CO

CO t-
Ol Ol

^ oi

II II

p p

in CO

O Ol

o

Ol

Ol

CO

II

CO

CO

00

in

CO

oi

II
in

ai

CO

CO

1— 1

Ol
Ol

o

"^

in

CO

II

II

II

in

CD

o

CO

I— 1

d

d

o

1 — 1

oq

1— (

CO
CO

Ol

o

CO

00

in

CO

oi

oi

II

II

II

II

o

o

m

in

00

00

I— t

ai

00

1— 1

in

00
1— 1

CO

>— 1

GO
CD

in

Ol

in

CO
CO

Ol
Ol

O

in

00

CO
in

. — 1
o

o

Ol

I— 1

1 — (

1— 1

oi

oi

II

II

II

II

II

. — 1

m

Ol

o

in

o ^ CO t^ in

00 00 OD CO 00

05 Ol CO

CO Ol 1-H

-* in CO

CO t-; Ol

oi oi TP

II II II

Ol Ol t-

t- t- r-

Ol ^H 05
Ol Ol 1— I

o o o

^_ p p

CO oi CO

II II II

00 p ^

d '^' 00

i> i^ t-

CO

CO
I— I

cq

+1
oq

00
Ol

CO

oq
oq

+1

05

CO

Ol

in
+1

CO

CO

o

Ol

O

05

l-H

CO

T-H

f-H

l-H

Ol

Ol

in

l^

00

00

C35

l-H

T-H

l-H

l-H

1— (

1— (

1— (

r-H

i-H

oq

o

in

l-H

in

05

t^

CO

Ol

CO

o

in

CO

05

05

l-H

CO

1—1

t-

Ol

05

CO

CO

05

in

l-H

CO

f-H

CO

'^

in

in

oq

p

Ol

m

Ol

Ol

O

o

02

in

in

■<*

d

^

d

d

d

in

in

d

CD

CO

oi

d

Ol

■-H

Ol

Ol

Ol

't*

^

t-

00

00

00

00

00

f— (

1—1

1— 1

1 — 1

l-H

^H

1 — 1

^H

^H

1—1

1—1

'~'

1 — (

5

4->

a

C/D

O

o

r-i

a
<

v.
a
<

»rH

c

! 1

Oj
1— I

2>

a

a

Oj
CD

CJ

o

o
O

CJ

a
<

a
<

0)

C

1—1

w

^

>>

^.

>.

^>

"G

>.

_>.

^

>,

^

25

oi

<3j

0;

1;

ij

0;

c5

OJ

0;

5

Cj

OJ

0;

5

S
^

cS
CD

*1

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

35

hatchling size is best determined from
young hatched in captivity.

Young were hatched from 21 clutches
in 1965/1967, 1969, 1970 and 1974; 50
males averaged 109.3 di .91 mm and 33
females averaged 114.1 ± 1.22 mm. Most
series of young collected in the field
averaged several per cent larger than
these laboratory hatched young, and pre-
sumably they had made some growth
between the time of hatching and the
time of capture.

Tables 7 and 8 show mean lengths
of young males and young females at
biweekly intervals for six different years,
and for a composite of 11 earlier years
when samples were generally smaller.
Each vertical column in the tables traces
the growth of one cohort of young snakes
from the time of their appearance soon
after hatching to an age of about 21
months in late spring. The months of
July, August, and November through
early March are not included because
no samples of snakes were obtained then.
Many other biweekly intervals in Sep-
tember, October and March through
June of specific years lack data because
no snakes were found, or samples were
inadequately small. Only samples of ten
or more are included.

In some instances average length in
a sample is slightly less than for a sample
of the same cohort in an earlier period.
Such aberrations are due to chance and
to the fact that successive samples in-
volved different groups of individuals.
In most instances slight gain is indicated
for each biweekly interval, but the aver-
age increment was small in cool weather
just before hibernation and just after
emergence. The following figures repre-
sent average increments for the several
annual samples: late September to early
October, 3.63%; early October to late
October, 1.32%; late March to early April,
0.48%; early April to late April,' 2.08%;
late April to early May, 2.16%; early May
to late May, 2.72%; late May to early
June, 3.29%. Growth is fastest in sum-
mer, but few snakes were obtained dur-

ing late June, July, August and early
September. The second-year snakes have
increased by an average of 46.7% (males)
or 49.8% (females) over the length of
first-year young at the time of their
emergence from hibernation.

In Tables 7 and 8 some annual co-
horts of young are definitely accelerated
or retarded compared with others at the
same stages in their history. For exam-
ple, the young hatched in 1973 were
larger in April 1974 than the April young
of other years. Development may be
delayed or hastened by early or late re-
tirement into hibernation, but such early
differences tend to become smaller or
disappear before the end of the first year.

Another aspect of growth brought
out by Tables 7 and 8 is difference be-
tween the sexes. In every sample of
first-year snakes in which there were 20
or more of each sex, the females aver-
aged larger. In 13 such series, the fe-
males were 101.2 per cent to 104.7 per
cent of male length, with a combined
average of 102.9 per cent. In 11 corre-
sponding series of second-year young,
the females were 100.5 to 107.8 per cent
of male length, for an average of 104.2
per cent.

Figure 14 is based on the snout-vent
lengths of first- and second-year young
distributed over the year except for No-
vember through early March, and July
and August, when too few snakes were
found for an adequate sample. It shows
that the females are consistently larger
than males and that young make rapid
growth in their first few weeks. How-
ever, growth slows with the advent of
cooler weather in autumn. A high pro-
portion of all first-year young are be-
tween 120 and 130 mm in length, and
the means for both sexes are usually
within this range from August through
the autumn and spring until May. Those
sampled in March and early April are
only slightly smaller than those sampled
in early May. From late April to early
September young make rapid growth,
increasing from about 130 mm to about

36

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

180 mm; then, during their second au-
tumn, winter and spring gains are rela-
tively slight.

An arbitrary dividing line between
second- and third-year snakes is set be-
tween 214 and 215 mm in males and
between 223 and 224 mm in females, on
the basis of the gains made by marked
individuals. Some snakes seem to suffer
little or no stunting from the effects of
processing, and most have overcome the
effects within a month. Those recaptured
after the lapse of an entire growing
season or a longer interval were assumed
to have made normal growth, the initial
stunting being a relatively minor factor.

Figure 14, showing the growth curve
through the tenth year in females and
the eleventh year in males, is based upon
the records of all marked individuals
recaptured after a growing season or a

longer interval. Since little growth oc-
curs from September through April, and
since nearly all the records are from
spring and fall, each record was assigned
to a particular age-class, with disregard
for fractional years, although the actual
age of those classed as two-year-olds, for
example, ranged from 24 months to
about 33 months. Also, the figure in-
cludes data from the series of unmarked
young that formed discrete age classes
at the time of hatching, and in the fall
and spring following hatching. The two-
year-old and three-year-old classes are
based upon recaptured individuals that
were marked as hatchlings. The remain-
ing age-classes are based partly on
those marked as hatchlings, but also on
those marked as second- and third-year
young and recaptured as adults. Because
second-year young overlap in size both

AGE IN YEARS

Fig. 14. — Lengths of male and female ringneck snakes marked eady in life and recaptured. In each
series mean, standard error, standard deviation, range, and sample size are shown. Growth is rapid
in the first three years but is slow in the older snakes. Females average larger and plot above males.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

37

<
a

Q

Z

o
u

Q
Z
<

H

s

z

<

o

z

o

o

H

K

o

K
O
U

[^

o

X
H
O

z

w

hJ

z
■ <

00

u
1-1

<

CO
05

iX

cq
I-

05

'X

02

iX

o

Ir-
as

iX

CD
CD
05

'X

00

CO
05

iX

tr-

OJ

ee

-w

02

'S

o

1

a

t-

r^

in

o>

5

<x

lO
CM

00

ci ^

+1 +1

00 o

t-^ d

CM CO

in

00

+1
o

(M

CO
+1

02

CO
CO

CO ^ >— I

CO I— I ^

t- CM CD 00

TP 03 rH (M_

^ oi CO c<i

+1 +1 II II

p ^ p p

CO CO CD 00

CO CO 'a^ "^

CO

CD

p

CO

II

CT>

CO ^

CO Tt^

(M ■— I 05

p CO C<1

1— I I— I o-i

II II II

^ lO CO

CO d 00

(M CO (M

CO

CO

in
oi

II
o

O <M

Oi 05

co' c<i

II II

N in

iri oi

cq CQ

o
cq

00 00

■^ c4

II II

CI 05

CO ^

cq cq

t-

co

0 CO

01 00

II II

t- o

CO cq

CM cq

—I 02

cq 1—1

^ in

p p

c^i co'

II II

O 00

in i^

CO CO cq
cq i-H ^

05 OT CD

p CO p

c^i CO CO

II II II

"*. °° t".

in c<i 1-'

CO CO CO

CD t— t- o in

t- CO '^ i-H 1— I

m t- CO t- o T)<

cq 00 CO in p o]

T-H ' ^' CO ' co'

II II II II II II

cq ^ CO cq in in

'^' 00 in d 00 i>

cq cq CO -^ Tt< -*

CO o

CD cq

CD O

I — I ^H

1— i c^i

II II

t- p p

1-J CD 1>^

cq cq -"iH

in o CO CD CO 00 cq cq cq

coo in-^cocDcot-t-

p in p 00 p p in c] cq
00 -^ CO CO C7> '-f' in -r CD
cq cq cq cq ci CO CO ^ ^

in

00

1—1

CD

in

II

II

t-

o

in

05

d

00

cq

II

00

CO

00

o

II

05

i> c^; in 1— I CD o CO
Tt i> cq CO 1— I ^ ^

o

^ CO t-

in oi t-

o t- t>

c-1 o in

cq c<i CO cq CO cq CO

II II II II II II II

o o in cq 00 cq '-t^

CD CD t-^ Tf CO 05 t-^

00 GO CO 05 CO C5 00

00 CD cq

1— I C^l rt

00 cq in

05 t- ^

cq in

II II II

in ^. o

cq 05 t-^

t- t- t-

t- CD ^

^H O 1— I

t> ^ cq

cq cq tp

II II II

P TT i-H

cq' d r-!

00 CO 05

cq

05

oi

II

00

CO
05

o
CO

cq

II

o

00
05

in in
cq in

O 00

— ; d

CO 05

00 Cl 1> ^

^ CD ^ CO

in p CD -*
oi co' d CO

C35 05 05 05

•*^

•*~*

p

a

o

O

o

1^

>,

2>

<U

■ ;_

(D

CU

^

4-*

-t-i

.-2

•- & jJ -jr;

^ Qi ^ oj -^ gj ^

S JS

a

CD

"^^^^

-r; a; i:: CD

S lis S ^

2 3 C

s :5

38

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

first-year and third-year young, there is
some uncertainty as to the age of any
individual marked when partly grown
and recaptured as an adult. Each such
individual was arbitrarily assigned to the
age class most typical for snakes of its
size at the time of its first capture. The
figure shows that growth is most rapid
just after hatching and in the growing
season following the first hibernation,
that in the second full growing season
the rate has slowed to about two-thirds
and in the third season to one-third, and
that thereafter, as adults, the snakes
make only slight gains from year to year.
At the time of hatching females are

already somewhat longer than males, at
an age of 20 months they are five per
cent longer, as young adults at 32 months
they are nine per cent longer, and in old
adults they are 15 per cent longer. Fig-
ure 15, showing average and individual
length increments in successive years
after hatching, indicates that females
grow faster. After the third year the
snakes gain weight relatively slowly.
Actual length gains in millimeters aver-
aged 54, 43, 23, 10, 2 and 4 for the first
six years in males, 63, 51, 19, 9, 7 and 3
for females. In the adult snakes the
usual annual growth is not much greater
than the expected margin of error in

<

3

<

uj

z

LlJ
O

X

I-
z

LlJ

_J

LU
CD
<

I-
Z
Ld
O
(T
UJ
Q.

^70

o
o

•

50

8

o

40

°»
O

o

9
o •

■30

o
o

•

o:

o

20

•

•

o

•

of

■

10

•

o0

•

■

1

o

•

•

:•

ST 2ND

3RD

4TH 5TH 6TH

7TH

Fig
the

FULL SEASONS OF GROWTH AFTER HATCHING

15. — Percentage gain in length (S-V) in individual ringneck snakes from one fall and/or spring to
next in successive, or first to seventh, seasons of growth. Open circles, females; .solid circles,

males; large circles, mean for each series.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

39

measuring; hence, it is difficult to deter-
mine. In 42 adult males six years old or
more average annual gain was 3.6 ± .41
mm, and in 16 females the correspond-
ing figure was 2.9 ± .89 mm.

Weight. — For snakes of any given
length the heaviest individuals generally
weighed from 2.5 to 3.0 times as much
as the lightest (Fig. 16) and this wide
range occurred at all times of years. The
heaviest individuals obviously were
thriving. Those that were recaptured
often had grown much faster than the
average rate. The lightest individuals
were those that were seriously handi-
capped by disease, injury or undernour-
ishment. Often such individuals were
destined for early elimination.

No well defined seasonal trend in
weight was discernible. Snakes emerg-
ing from hibernation appeared as well
nourished as others, and there was no
conspicuous tendency to store quantities

Table 9. — Deviation from Mean Weight in

Monthly Samples, Mainly from Periods of

Subnormal Precipitation ( for Each of the

Samples A' = 100).

Time of Sample Sample Weight:
Percentage of
Normal Weight

April 1955 106.4 94

October 1959 .... 96.9 103

Early April 1965 88.8 112

March 1966 ...... 89.5 111

June 1966 86.5 115

Aug.-Sept. 1966 88.4 127,

October 1966 .... 75.3 132

March 1967 ...... 84.7 118

September 1970 104.5 95

March-April 1971 93.0 107

Nonnal Weight:

Percentage of

Sample Weight

0± 1.67(63-128)
1 ± .87 (80-130)

6 ± 1.32 (71-147)

7 ± 1.47 (78-171)
.5± 1.08(82-171)
.7± 1.94(94-183)
.8 ± 2.34 (87-194)
.0± 1.73(86-163)
.6± 1.88(67-155)
.4± 1.95(65-180)

of fat at any time of year. However,
climatic fluctuations affect amount and
availabilty of the snakes' food, and may
result in depression or elevation of aver-
age weight. Mean weight for the popu-
lation as a whole may increase more
than 40 per cent in response to improved
conditions. Sex and size classes are af-
fected similarly. Table 9 compares 10

separate monthly (or bimonthly) sam-
ples of 100 snakes each. For each indi-
vidual the weight is compared with the
mean weight for all individuals of that
length (Fig. 16). Interest focused on
1966 because snakes recaptured in that
year were rather consistently retarded in
their growth and precipitation was un-
usually low. Four of the 10 samples are
from 1966; they show that in March of
that year, the snakes were 89.5 per cent
of normal weight, and that they con-
tinued to deteriorate till October when
they were only 75.3 per cent of normal
weight.

POPULATION DENSITY

The secretive habits of snakes in
general render them difficult subjects for
census. Few figures concerning popula-
tion density have been published, and
in general these inspire little confidence.
The most satisfactory censuses of natural
populations are those in which every
individual on a sample area of represent-
ative habitat can be seen and counted
in the course of a brief sampling period.
Since ringneck snakes are rarely found
away from shelter, best information
about their numbers can be obtained by
moving the sorts of objects they use for
shelter. Where such objects are abun-
dant in favorable habitat, it is evident
that the snakes attain high population
densities, as many may be found within
a small area. However, it is doubtful
whether all those present, or even a high
percentage of them, can be found at any
one time.

Most revealing data concerning ac-
tual abimdance were obtained on 13
April 1968; in approximately half an hour
of field work 279 ringneck snakes were
captured in Quarry Field, a nearly flat,
triangular hilltop area of 3.8 hectares
bounded on two sides by steep rocky
slopes, and on the third by a county
road. All snakes were found in aggrega-
tions beneath approximately two dozen
corrugated metal strips that were scat-
tered about the field.

10

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

cn

<

or
o

X
CD
UJ
5

^9

8
7
6
5
4

•3

'

150 200

SNOUT VENT LENGTH

250

IN MILLIMETERS

300

Fig. 16. — Relationship of length to weight in ringneck snakes, grouped in fi\'e-mil]imeter intervals.
For each length interval sample the standard error, and standard deviation and range of weight

are shown.

The 279 snakes caught on this occa-
sion represent a minimum popuhition of
73.5 per hectare. On numerous other
occasions from 1964 through 1974, the
same series of shelters (and others added
subsequently) were turned in routine
search but on no other occasion were so
many snakes found at one time. In eval-
uating the significance of the series of
snakes to determination of population
density, it is necessary to know whether
individuals were residents or transients,
how large an area might be covered by
individuals and to what extent the local
population was using the specific shel-
ters checked. This same 3.8-hectare field

was the locale of intensive population
study for four years before the collection
and seven years afterward (including
the 1968 season); therefore, much infor-
mation is available about some of the
individual snakes caught. Of the 279
total, 156 were caught only once, on 13
April 1968, never before or after. Fifty-
three others were caught one or more
times before that date but not afterward;
51 were recaptured subsequently but not
before, and 19 were caught both before
and afterward. The snakes in this group
tliat were caught on other occasions
were distributed by years as follows: 4
in 1964, 7 in 1965, 32 in 1966, 2'4 in 1967,

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

41

21 in 1968 (7 before 13 April and 14
after that date), 27 in 1969, 20 in 1970,
4 in 1971, 5 in 1972, 5 in 1973 and 4 in
1974. The 279 snakes inchided 245
adults (203 males and 42 females), 24
second-year young ( 17 males and 7 fe-
males ) , and 10 first-year young ( 6 males
and 4 females). The sample has a high
proportion of adult males (believed to
be especially active in sexual search at
this time and place); immatures consti-
tute only 12.2 per cent.

The 13 April search that yielded 279
snakes resulted in the escape of many
others because the aggregations under
the metal strips scattered in all directions
when they were uncovered. Females
and young, missing from this sample
that consisted mostly of adult males, may
have been present within the field but
less inclined to trail other individuals and
form aggregations (see section on Com-
position of the Population). Or they
may have been absent from the field,
still concentrated at hilltop limestone
outcrops where they had hibernated.

Consideration of all available data
suggests that the 279 snakes captured on
13 April 1968 were mostly residents of
Quarry Field and that they were only
a small part of the population present
there. Subsequent study by Richard L.
Lattis, trailing ringneck snakes equipped
with radioactive tags, has shown that the
metal strips are not focal points of the
snakes' activities, and are not used regu-
larly by any individuals nor do they
constitute an essential part of the local
habitat. Instead the strips are used from
time to time by snakes in the immediate
vicinity when they happen to provide
shelter with optimal thermal and hygric
conditions. Most of the snakes most of
the time are using shelters such as the
surface mat of dead vegetation, where
they cannot be found readily. Although
many of the 279 snakes in this sample
were caught at other times, over a ten-
year period, they collectively never made
up more than a small percentage of any
one sample.

In the spring of 1968 after 13 April,
there was little opportunity for further
field work, but 23 snakes were captured
on 21 April and eight on 10 May and
these 31 included only one recapture
from the 279 on 13 April. This one-to-31
ratio suggests that the original sample of
279 represented an actual population of
8650. This figure, based on just one re-
capture, is obviously of little significance
but it does serve to demonstrate that the
279 snakes of the 13 April sample were
drawn from a much larger pool and their
marking did not constitute substantial
progress in attaining a fully-marked pop-
ulation on the study area of Quarry
Field. In late September and October
of 1968, collections of snakes were made
in Quarry Field on 16 days and a com-
bined sample of 184 was obtained, not in-
cluding the hatchlings that had emerged
in late summer. This sample included 14
individuals from the 13 April collection.
The one-to-13 ratio in this collection
suggests a total population in spring of
13 X 279 or 3630 (950 per hectare) in
this field. However, the long interval
between 13 April and the October-
November resampling lessens the value
of the calculation. Population changes,
with dispersal and replacement of some
individuals, would have occurred in the
interval.

A population index based upon
capture-mark-recapture is useful only if
certain assumptions can be made about
the stability and spatial relationships of
the population: it should be limited to
a definite area and the individuals
marked should distribute themselves ran-
domly; their survivorship or difficulty of
capture should not be affected; there
should be no reproduction in the popu-
lation or else it should be measurable.
From what is known about the Dia-
dophis population, the following state-
ments bearing upon these problems may
be made. 1. Most individuals most of
the time stay within areas much smaller
than those sampled. There is sometimes
a shift between summer range and hi-

42

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

bernaculum or place of egg-laying. 2. As
in other vertebrates some individuals
may wander in a random manner as
transients, but most stay in the same
place, and sometimes are recaptured
within a few meters of a former capture
site after several years. 3. Individuals
experience some deleterious effects from
marking and handling, but certainly
most survive. No definite instances of
mortality were observed. 4. Reproduc-
tion occurs only once in the annual cycle
and young constitute a discrete size
group distinguishable from older snakes,
so that they do not complicate capture-
recapture ratios.

In view of these traits and in view
of the fact that no other method is avail-
able to measure populations even in
general terms, it seems worthwhile to
apply a "Petersen Index" to the Dia-
d aphis population studied. Some work-
ers have rejected capture-recapture in-
dices as a means of censusing snakes on
the grounds that the published figures
from studies of this type are too erratic
to be acceptable. Yet there is nothing in
the known population ecology of snakes
that makes capture-mark-recaptin-e tech-
nique inapplicable to them. In general,
snakes are difficult to census because
they are secretive and it is difficult to
obtain adequate samples. However, each
case should be judged on its own merits,
depending on the size of the sample and
the extent to which the theoretical as-
sumptions of a Petersen Index can be
demonstrated to be satisfied.

In my study Quarry Field (3.80 hec-
tares) and House Field (5.76 hectares)
were sampled on hundreds of occasions
but with the possible exception of 13
April 1968 the snakes captured on any
one day were too few to be used as a
separate sample. By combining many
day to day samples, adequate presam-
pling and resampling periods within a
spring or fall season were obtained in
several instances.

Table 10 shows seven censuses in five
different years in House Field and

t/3

a
<

CO

w
«

t-

<
u
a

I

u
s

H
<

U

I

<
o

Q
W
H
<

U
i-l

<

U

<

<

H
CO

O
£f

H

z
o

11

u

<
z

w
o

H
Z
O

z

O

c«
U

I— (

H

z
u

Q

z

o

M

H
<

^

Oh
O

►J

<

.3 t- --< CD

'^ o

o

'4-*

QJ ^ C

— -^ TS

a b
3 -<"

lU

o IJ

i:li

CC -rt

.3 c
Qj a o

03 lU
PM

bc
c

a o
S 'C

w Ph

'a'?
Pm

C

a> o^ ^H 10
xt< ^ oq o

00 i> -* ^

01 o o o

O CO o o

O t- Tt^ 01

t^ CO in ^

00 O 1> CO

t^ a> 00 ^

0000

.M .W 4-> 4-*

O CD CO CO

CO CO CD t^

in 00 '^ in

CO I— I CO

in in

a a

< <

o cj

.2 S S

>> CD t-

J.H CO CD

« 2 2

•c -d
a a
< <

I I

o o

-f 00 00
t- O 00

t- CO ^

in

CO

o
'^
00

CO

t- c^l O

CD CO CO

CO in 01

000

CO

CO

t- 00

CO CO 00

CO CO CO

CO CO CO

c

3

>x

>,

0

>>

>^

<A

03

(—

03

rt

S

^

s

^

S

^

>

^

^

0

0"

0

0

0

CO

CO

D

(U

;s

CO

CO

CO

CD

c

c

i

CD

CD

. — (

f— 4

I—'

3

0)

I-H

1— 1

T.

'd

>.

>.

0

4->

'd

't-

a

a

rt

03

CJ

a

a

<:

<

s ;s

0

<

<

4 s

a

ii in in

a ^ 't

(73 '-' ^

3 a a

bC < <<

< -c ^
7 o o

CO

o

CT) 05

O

S S

in CD t~

CO CO CD
05 ro C35

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

43

Quarry Field. Calculated figures for
population density ranged from 719 to
1849 per hectare, with a mean of 1266
per hectare for all seven calculations
combined. Individuals are long-lived,
and year-to-year variation in reproduc-
tive success would not be sufficient to
bring about the fluctuations which must
be caused by biased or inadequately
small samples. The highest and lowest
figures are probably far from the mark.

Rapid seasonal changes in kind and
amount of activity and in the ratios of
sex- and age-groups in samples hinders
census and prevents eff^ective use of re-
finements of the capture-recapture index,
such as have been proposed in the
"Hayne Method" and the "Jolly Method."
A crude capture-recapture index does
not provide a precise measure of the
population, as has been demonstrated by
several investigators. Nevertheless, in
this instance, and with secretive animals
in general, it does provide better infor-
mation about numbers than do direct
counts, and at least indicates the order
of magnitude of the population. It sug-
gests that in favorable habitat population
densities of ringneck snakes are often
between 1000 and 1500 per hectare.

It cannot be stated with any degree
of assurance whether the census figures
obtained tend to err on the low side or
the high side. Even the lowest census
figure, of 719 snakes per hectare, would
seem excessively high for most kinds of
vertebrates, but all evidence suggests
that Diadophis pinwtatiis actually does
attain high densities. The methods of
obtaining samples, by returning regu-
larly to certain favored spots to turn
sheltering objects or to check live-traps,
would tend to favor recapture of certain
localized individuals and would increase
the ratio of recaptures to new captures,
resulting in an erroneously low census
figure. On the other hand, wandering
tendencies of some individuals, which
might leave the study area after being
marked, with replacement by newcom-
ers, would result in an erroneously low

recapture ratio and calculation of a fig-
ure for the population that is too high.

COMPOSITION OF THE
POPULATION

Age structure. — Age structure of the
adult population was calculated after
analysis of the records of individuals
marked while young and recaptured as
adults (Figs. 14-19, Tables 11-14). These
records, combined, show that length is
highly correlated with age, so that each
individual may be assigned to a most
probable age group on the basis of its
measurements. Within each age group
lengths vary over a wide range but tend
to form normal curves, with a clustering
of records near the mean and progres-
sively fewer in successive stages of de-
viation from the mean. Each age group
exceeds younger age groups in mean
length, but successive annual increments
become progressively smaller. Samples
are smallest for the relatively old snakes,
so that the trends of growth and survival
become increasingly obscure.

The graph in Figure 14 shows some
inconsistencies in the older age groups;
for instance ten-year-old males in the
sample averaged slightly larger ( 274.5 ±
1.84 mm S-V) than eleven-year-olds (273.0
±2.10). Such irregularities may result
from inclusion of occasional stunted in-
dividuals or those of unusually acceler-
ated growth. Recognizing that growth
continues in adults, but becomes pro-
gressively slower with increasing age,
and that there is a consistent difference
in sizes of the sexes at any given age,
the growth curve has been adjusted by
slight alterations as shown in Table 14.
The table indicates that the annual re-
duction from any age class to the next
older class is consistently between 20
and 30 per cent, and averages 23.4 per
cent for males and 26.8 per cent for fe-
males. For example, the 1458 first-year
males make up 25.2 per cent of the en-
tire sample of 5814 males and the 92
males in the size range (275-277 mm)
typical of eleventh year individuals make

44

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Table 11. — Composition of a Population of 9162 Diadophis punctatus Captured and

Marked on the Reservation, 1958-1969.

Number Size Range Number of Size Range Total Male to

of Males of Males Females of Females Number Female Ratio

Adults 3436

Second-year young __ 920
First-year young __._ 1458
Total 5814

215-351

1851

221-392

5416

1.85 to 1

147-214

513

160-220

1444

1.79 to 1

90-146

984

85-159

2552

1.48 to 1

90-351

3348

85-392

9162

1.73 to 1

up 23.7 per cent of the snakes that are
of this size or larger (presumably more
than 10 years old). Non-age-specific
annual mortality of approximately one-
fourth for both males and females is
implied by the figures. Incidentally the
figures show a gradual increase in the
sexual difference in size, with females
averaging 109 per cent of male length in
newly matured third-year snakes, but
increasing to 112.0 per cent in the eighth-
year group and 115.0 per cent in the
14th year group.

Although any individual can be as-
signed to a "most probable" age on the
basis of size alone, there is no certainty
that such allocations are correct. The
chances that such assignments will err
by a year or more are increased greatly
in the old snakes, in which any one age

class includes individuals that cover a
wide range of size.

Behavior differs between the sexes
and between young and adults so that
most field samples are probably subject
to some bias, with certain classes better
represented than others. Table 11 di-
vides the population into six classes:
adult males ( third year and older), adult
females (third year and older), first-year
males and females, and second-year
males and females. The table includes
the 9162 snakes caught over a 12-year-
period, when those captured were indi-
vidually marked and released. Recap-
tures are not included. These snakes
were not all contemporaneous; rather,
the 12-year interval spanned many gen-
erations, with constantly changing ratios
of the sexes and of age groups. How-
ever, all are here combined to represent

-

llJ

_J

Q_

N

80 214 353 264 483 609 474 369 578694 1174537 318 883 1067 332 733 381 785

<

70

-

C/)

60

-

u.

n

O

50

-

III

o

40

-

<

t:

30

n — I •■

r

UJ

o

■20 ^ r

(—1

n n

T r- ■:

on

-\

■"1 ::;•; _ >:-

1 ^

LlJ

- 10 S I

" ^

■■ ^

■: :^ T- v :': :;

•:■

O.

I i-il l:iil 1 p 1

1 1 i,M ij-l 1 1,^ 1

M i i M

:• :.•••

I960

1965

1970

1975

Fig. 17. — Percentages of adults (solid columns), second-year young (shaded columns), and first-year
young (open columns) in annual spring population samples from The University of Kansas Natural
History Reservation and nearby areas over an 18-year period.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

45

one population in this composite, mini-
mizing the seasonal and year-to-year
differences that characterize the smaller
samples. In general the samples repre-
sent either fall or spring in the pause
between the summer growth periods;
therefore, there are discrete annual age
groups and in the younger snakes there
are fairly distinct size groups.

The 3436 adult males make up 39 per
cent of the total in Table 11 and are
much better represented than any of the
other live classes. For example, the adult
male to adult female ratio is 1.85 to 1.
However, Table 12 shows that this ratio
changes throughout the season of activ-

Table 12. — Changing Sex Ratio of Adult

Diadophis punctatus Throughout Season of

Acttvity.

Time

Total

Number

Number

period

Sample

! of males

of females

16-31 March

285

182(63.9%)

103(36.1%)

1-15 April

1291

927(71.9%)

364(28.1%)

16-30 April

1187

818(68.9%)

369(31.1%)

1-15 May

939

662(70.5%)

277(29.5%)

16-31 May

509

323(63.4%)

186(36.6%)

1-15 June

351

191(45.5%)

160(54.5%)

16-30 June

297

90(30.5%)

207(69.5%)

1-15 July

83

24(29.0%)

,59(71.0%)

16-31 July

50

26(52.0%)

24(48.0%)

1-15 August

57

23(40.4%)

34(59.6%)

16-31 August

79

,50(63.3%)

29(36.7%)

1-15 September 288

220(76.4%)

68(23.6%)

16-30 September 667

500(74.8%)

167(25.2%)

1-15 October

560

385(68.7%)

175(31.3%)

16-31 October

268

141(52.6%)

127(47.4%)

1-15 November 18

11(61.2%)

7(38.8%)

ity, with females constituting only 23.6
per cent ( 1 to 4.24 males ) in the early
September sample of adults but increas-
ing to 71.0 per cent of the early July
sample. Throughout the spring months,
when the snakes are found in greatest
numbers, females comprise less than 40
per cent of adult samples, and often less
than 30 per cent. In the hot weather of
June and July, the snakes are much
harder to find. Samples are small then

but the ratio of females rises. These fe-
males are gravid and their relative con-
spicuousness seems to be correlated with
a tendency to frequent superficial situ-
ations where body temperatures are
readily elevated by basking. The re-
mainder of the population retreats to
deeper shelters. In late summer after
egg laying and in early fall females again
become relatively scarce in samples.

Sex-ratio. — The seasonally fluctuat-
ing sex ratio in samples raises the cjues-
tion whether males are actually more
numerous than females, or are merely
easier to find. The following ratios of
adult female to adult male captures in-
dicate that females, regardless of their
actual numbers, are more secretive and
less readily taken:

Individuals with only one cap-
ture: 1 female to 1.85 males

Individuals with 2 or 3 captures:
1 female to 3.34 males

Individuals with 4 to 12 captures:
1 female to 6.85 males

There may be some bias in most sam-
ples arising from the fact that sexually
mature males are more active than ma-
ture females, or immatures of either sex,
causing these males to be found more
often.

Physiological and behavorial differ-
ences between the sexes must result in
some differential mortality. Female mor-
tality may be a little higher, judging
from the fact that the oldest individuals
recorded were all males (Table 13).
However, the figures in Table 14 suggest
tliat mortality rates are not much differ-
ent in the sexes. Presumably in pre-
adolescent young the sexes are alike in
behavior or at least differ much less than
adults, hence more acceptable ratios can
be obtained from them. As shown in
Table 11, males are more numerous than
females even in the first-year young with
an approximate ratio of 1.48 to 1. First-
year and second-year young combined
made up 40.8 per cent and 43.0 per cent.

46

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Table 13. — Records of Large (and old) Males of Diadophis punctatus Captured and

Recaptured, Indicating Potential Longevity.

Dates of Capture
And Recaptvire

Size (S-V)

at Capture

And Recaj)ture

Time Span
of Records

(Years)

Most Probable Age
First Captvire Last Capture

Possible Age

Range at
Latest Record

25 April 1966 -

11 Sept.

1973

260-284

8

7

15

10-17 +

5 May 1966 -

20 April

1970

274-282

4

10

14

9-14+

18 April 1966 -

10 May

1968

276-293

2

11

13

8-15+

12 May 1966 -

16 May

1970

283-291

4

13 or more 17 or more

13-19+

9 May 1966 -

13 April

1970

273-280

4

10

14

9-14+

2 Oct. 1964 -

2 Oct.

1970«

264-290

6

8

14

9-14+

25 April 1966 -

9 April

1970"

267-278

4

8

12

6-12+

17 April 1965 -

30 May

1970^

267-268

5

8

13

-14+

4 Sept. 1965 -

13 Oct.

1968

272-275

3

10

13

7-13+

8 April 1965 -

15 April

1970

265-263

5

8

13

8-13+

4 April 1966 -

17 April

1974

267-281

8

8

16

7-16+

13 April 1968 -

17 April

1974

273-283

6

10

16

9-16+

13 April 1968 -

14 May

1974

274-282

6

10

16

9-16+

a Recaptured also in April 1967, 1968, 1969 and 1970.
'' Recaptured also in April 1970.
"^ Recaptured also in May 1968.

respectively, of the male and female
samples in Table 11.

Figures 17 and 18 show the changing
ratios of adults to first-year young and
to second-year young over an 18-year
period. The ratios are believed to reflect
varying reproductive success in different
years. Also, they are affected by the
times and places that samples were col-

lected. In some years large concentra-
tions of young were found in nursery
areas, whereas in other years the collec-
tions were made mostly in the areas
where adults predominated.

Fecund females are estimated to
comprise approximately 20.2 per cent of
the population. If the 1851 females rep-
resented in Figure 19 produced an aver-

1960 1965 1970 1975

Fig. 18. — Percentages of adults (solid columns), second-year young (shaded colmuns), and first-year
young (open columns) in annual fall population samples from the University of Kansas Natural His-
tory Reservation and nearby areas, 1959 to 1974.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

47

age of 3.89 eggs per clutch, their total
output would amount to 7210 eggs and
this annual increment would add 78.6
per cent to the spring population. It is
calculated that by the following spring
the hatchlings from these 7210 eggs
would be reduced by approximately
two-thirds to 2442 (1458 males, 984 fe-
males ) . The 66 per cent mortality would
be attributable to several causes, includ-
ing resorption or abortion of eggs, in-
fertility or inviability of some of those
laid, genetically defective hatchlings,
predation on hatchlings and winterkill
of the hibernating young.

The age pyramid in Figure 19 is
based upon the actual numbers of snakes
captured, but shows also the numbers of
missing adult and second-year females
that might have been present but un-
detected (shaded ends of columns) if
the sex ratio of 1.48 males to 1 female
that was found in hatchlings is the real
ratio for the older snakes also.

PREDATION

There are few published records of
predation on Diadophis. In the present

study 94 records of eight species of pred-
ators on the Reservation and nearby
areas were assembled; large series of
pellets of raptors and scats of snakes
were examined, hawk nests were visited
on many occasions to check for remains
of prey, and hundreds of live snakes of
other kinds were palped to identify food
objects. The records of predation on
Diadophis included 35 by Agkistrodon
contortrix, 25 by Biiteo jamaicensis, 15
by Coluber constrictor, 13 by Buteo
platypterus, 2 by Lampropeltis triangu-
him and 1 each by Bubo virginianus,
Crotahis horridus and Rana catesbeiana
(Table 15). Sample size varied from
2763 stomach examinations of Rana
catesbeiana to only 20 records of feeding
in Lampropeltis triangulum. The 192
pellets of Bid)o virginianus represent
mainly the colder half of the year, with
reptiles including Diadophis probably
much less well represented than they
would have been in summer. Table 15
shows sample sizes and frequency of
Diadophis in food for the eight predator
species, and attempts to show the rela-
tive severity of predation on the basis of

Table 14. — Sample of 9162 Ringneck Snakes Captured Over a 12-Year Period ( 1958-
1969) Arranged in Size Groups and Supposed Annual Age Groups to Show Progressive

Reduction in Numbers with Increasing Age.

Males

Females

Most

Probable

Age in

Years

Most

typical

Size Range

(mm S-V)

Actual
Number

In
Size Class

Total Alive
At Start
of Year

Percent Lost
During Year

Most

Typical

Size Range

(mm S-V)

Actual
Number

In
Size Class

Total Alive
At Start
of Year

Percent

Lost

During Year

1st

110-146

1458

5814

25.2

110-159

984

3348

29.4

2nd

147-214

920

4356

21.1

160-223

513

2364

21.6

3rd

215-233

851

3436

24.7

224-256

464

1854

25.1

4th

234-245

673

2585

26.0

257-272

380

1387

27.4

5th

246-253

484

1912

25.3

273-280

231

1007

21.6

6th

254-258

300

1428

21.0

281-287

169

776

21.8

7th

259-263

261

1128

23.2

288-294

163

607

26.8

8th

264-267

211

867

23.4

295-300

101

444

22.8

9th

268-271

154

656

23.4

301-305

89

343

26.0

10th

272-274

114

502

22.7

306-310

74

254

29.1

11th

275-277

92

388

23.7

311-315

49

180

27.2

12th

278-280

76

296

26.7

316-320

33

131

25.2

13th

281-283

56

220

25.4

321-325

27

98

27.5

14th

284-286

43

164

26.2

326-330

21

71

29.6

15th or

older

287-351

121

331-392

50

48

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

1000

NUMBERS

500
OF SNAKES

Fig. 19. — Age pyramid showing relative numbers of successive annual age classes of male and female
Diadophis punctatus, based on figures in Table 14. Shaded portions of columns on female side rep-
resent a part of the population not actually recorded but suspected to be present if the ratio of 1.42
males to 1 female found in first-year young is correct for adults also.

frequency of Diadophis in the food, pop-
ulation density of the predator, and
frequency of its feeding. Number of
meals consumed annually by each kind
of predator on the 590-acre Reservation
have been estimated from censuses of
the predators and observations on the
frequencies of their feeding. Most kinds
are present and active only in the
warmer part of the year. For Bxiho vir-
ginianus and Buteo jamaicensis, both
present throughout the entire year, the
estimated total number of meals is di-
vided by two because it is unlikely that
predation on Diadophis would occur
during the colder half of the year.

The figures suggest that the copper-
head and blue racer combined account
for 94 per cent of the known predation
whereas none of the other six predators
accounts for as much as 1.5 per cent.
The combined kill of the eight known
predators would account for approxi-

mately 1700 Diadophis annually, a little
less than 7.4 per hectare on the Reser-
vation. However, this toll seems trivial
in view of the finding that Diadophis
attains population densities of several
hundred per hectare over extensive areas
of favorable habitat. Presumably the
population is regulated by some other
mortality factors.

Other predators may be even more
important than those studied, although
no records of their predation on Dia-
dophis were obtained. The eastern mole
Scalopus aquaticus is one such suspect.
It is known to feed largely on earth-
worms and other invertebrates, but feed-
ing experiments with captive moles on
the Reservation demonstrated that they
would catch and eat small reptiles at
every opportunity. The spotted skunk
(Spilo<i,a]e putoriiis), striped skunk
(Mephitis mephitis) and opossum (Di-
delphis marsupialis) all are known to

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

49

cS

u
Oh

O

I—"

t2

^.— s.

^ J2n3

2-? 8

a

•2 S r-

c:

« O *3

a

iJ2 -^ — ■

«

0, =« >

^ CO ^

*3

~ >< ^

-c p ^

^■^

c:

^ J- r^

••^ o ■■

i. 1-1 Oj

2

S CO .ti

u

^ oq

•2

s

««.

S

s ><

^-^ ^ W3

1

1^-

s

j^

kJ

K

S ^

5a

ft >- en

o

■§§1

O

S o

o ^

S

o

-S >>

"C

■— v „

o

■w i= ?3

B a S

.Co

■:= oq i!

.5

IS'"

CiC

<

^

s >.

o

5; ?^ en

^

<♦<*

^ EB

cS

■^ CO -

r-g ^H

-2 X

Co o

o

C v^ en

^

■Ji ft P

t^

o --

■*

•~ (M aj

CQ

.« '-'

Co tn

•^ ,4^

o

S Ji

.2 13

•S ^

c5

?JD (M

-S 05

» --1

<D

'Cu

r^

«

en

M-(

O

m

_N

c/5

^ o

^ o o
^ o o

CO

-r< o

O —I

C<1

^ o
in ^ o

t— I o Ol

CO 05 (M

— I O

in -H TfH

CM O

in

o o
o ^

o
o
in

in

CM

O

o

CO

in

C\l

o

o

o o

in

in

r— ( T-H

(N

<M

o
o
o
o

CO

I— I

CM

00

o

o

in CD

oc

Tf<

CO o

i-H

^

-*

CM

t-

05

o

o

d

o

CO
CO

o
o

o

CD

00

CO

00

o
m

CD

CO

m

CD

in

^ 2

Q eu "rt
^ - ^

^ F, r r >^

g .2

^ 5

'J-' _

c «

OJ P

O .S

Oj en

c
c

0-)

<^ ^ -2 S
.2 ;= i-^ C/D

o
^ I

O n3

c S
a^ o

feed upon small reptiles, and, as all are
fairly common on the Reservation, they
may be among the major predators on
Diadophis there.

The wood mouse Peromtjscus leuco-
pus is known to take a variety of animal
matter in its food. Evidence that it may
prey on hibernating Diadophis was ob-
tained in November and December of
1964, after several of these and other
snakes were released at known hiber-
nacula at hilltop limestone outcrops, with
thermisters implanted and with fine and
extremely pliable wires attached to them
and extending to the surface. After sev-
eral weeks it was found that all the wires
had been cut by the mice and there was
evidence that the reptiles themselves had
been eaten. Perhaps in these instances
the snakes were handicapped by the at-
tached wires, so that they were prevented
from reaching the deep and secure shel-
ters where they would have hibernated.
However any small reptile hibernating
where its body surface is accessible to
mice would seem to be vulnerable to
predation. The shrew Blarina brevicauda
also is a potential predator, especially on
hatchlings. Many of the snakes bore
scars on their bodies, and especially on
their tails, that could have been inter-
preted as shrew bites.

These worm-sized snakes, especially
the smaller individuals are potential prey
for birds other than raptors, such as the
crow Corvus brachyrhijnchos. Blue Jay
Cyanocitta cristata. Crackle Quiscalus
quiscula. Yellow-billed Cuckoo Coccyzus
americanus and Brown Thrasher Toxo-
stoma rufiim.

Disease may be an important mortal-
ity factor. One common ailment caused
the snakes to develop blisterlike swell-
ings in the skin, and sometimes these
produced open lesions. The snakes so
afflicted were in varying degrees emaci-
ated and sickly in appearance. Ringneck
snakes kept in confinement dehydrated
rapidly unless they had access to water,
but those kept in damp surroundings
rapidly developed the ailment described.

50

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

and succumbed. Blanchard, Gilreath and
Blanchard (in press) described this same
ailment in the population that they stud-
ied in northern Michigan, and Hender-
son (1970) was handicapped by it in his
study of food and water consumption in
confined ringneck snakes.

Winterkill may involve considerable
mortality in addition to predation of the
hibernating snakes. Since Diadophis
lacks capacity for digging and can only
follow existing passageways or force its
way through loose soil, entombment or
entrapment might frequently occur
through deposition or compaction of the
soil around the hibernating snake. Freez-
ing is also a potential source of winter
mortality, but the species penetrates into
regions having much more severe winter
climates than that of northeastern Kan-
sas, where this must be a relatively
minor factor.

DISCUSSION

Compared with small vertebrates in
general, the ringneck snake is K-selected
{semu Dobzhansky, 1951; MacArthur
and Wilson, 1967; Pianka, 1970) in the
following respects: sexual maturity de-
layed (first breeding in third year),
brood size reduced (3.9 eggs per clutch,
produced once annually), young rela-
tively large and well developed at hatch-
ing (36% of maternal length and 12% of
maternal weight), life expectancy in-
creased (annual mortality in adults less
than 25 per cent, with some individuals
attaining ages of 15 years or more). All
these traits make for stability of popu-
lations and except for the climax vegeta-
tion itself the ringneck is one of the most
stable elements in its community. Snakes
in general are notable for longevity, and
associated stability of populations. How-
ever, considering the small size of the
ringneck, it has progressed farther than
most in the direction of K-selection, a
trend correlated with its secretive and
partly subterranean habits.

The ringneck snake is a predator on
various sorts of small vertebrates and

invertebrates but the trend of food habits
differs in different parts of the range.
The locale of this field study is in the
ecotone between deciduous forest and
grassland, at the eastern edge of the
Great Plains, near the center of the geo-
graphic range of the ringneck snake,
which is a wide-ranging, polytypic spe-
cies. Selective pressures in this ecotonal
region have produced a population dif-
ferent from any in other parts of the
range, but many of the differences are
subtle. Snakes of the population studied
are relatively small and stubby, highly
variable in extent and pattern of dark
ventral markings, and more secretive and
subterranean in habits as compared with
eastern populations (but perhaps less so
than far western and southwestern pop-
ulations). In feeding they have become
specialists dependent almost entirely on
earthworms, whereas elsewhere popula-
tions are much more versatile but with
a tendency to specialize on plethodont
salamanders (northeastern United States
and Appalachian region), on insects and
other invertebrates (southeastern United
States) or on small reptiles (Southwest).
Geographic variation in size and
morphological characters occurs along
with variation in food habits, but the
morphological variation does not closely
parallel the behavioral variation. Gehl-
bach (1974) noted that in Diadophis
punctatus armji of Texas, near the range
of D. p. regalis, size remains relatively
small, but armji in that area resembles
regalis in its ophiphagous habits. When
individuals of the two subspecies were
confined together, mating occurred de-
spite the disparity in size. Cannibalism
did not occur in these mixed pairs al-
though other small snakes were readily
eaten when confined with the large D. p.
regalis. Gehlbach found intergradation
between the subspecies arnyi and regalis
in central and western Texas, but in the
Guadalupe Mountains he found isolated
populations of both types with no inter-
breeding. Gehlbach speculated that
small ringneck snakes similar to arnyi

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

51

and the more eastern subspecies, of
worm-eating habits and mesic habitats,
are ancestral and that the large, ophi-
phagous regalis-type populations evolved
from these snakes in Pliocene Madro-
Tertiary woodlands. An alternative pos-
sibility is that regalis is the more primi-
tive and that arntji and far eastern and
far western subspecies are derived from
a regalis-like ancestor. Although evolu-
tion conceivably could have proceeded
in either direction, regalis seems more
generalized, that is, more like related
genera of xenodontine snakes, but with
less typical exhibition of the characters
that distinguish Diadophis. Among such
generalized characters are: euryphagous
habits and predation on small verte-
brates, lack of a neck ring, and large
body size associated with large numbers
of ventral and subcaudal scales and with
17 body scale rows, whereas other sub-
species have only 15 to 13 rows.

The abundance and availability of
earthworms as a food supply has per-
mitted attainment of remarkably high
population density and biomass by the
snakes in the area of my study. Through-
out much of its range the ringneck is
one of the less common species of snakes,
but in northeastern Kansas it probably
outnumbers all other species combined.
On the basis of known population struc-
ture and the average weight of individ-
uals of different sizes it is calculated that
3.6 grams is a typical weight. A popula-
tion of 1266 snakes per hectare (the
average of seven capture-recapture cen-
suses) would thus have a biomass of 4.58
kilograms. Of local vertebrates only the
prairie vole (Microttis ochrogaster) has
been found to attain higher biomass over
extensive areas, with 7.41 kg per hectare
in an actual sample, but the vole is a
primary consumer, and its population is
subject to rapid and drastic change.
Other vertebrates of the Reservation
known to occur, at times, in biomass ex-
ceeding 1 kg per hectare include the
cottontail, Sijlvilagus fioridanus, 3.88 kg
(in the only instance in which a popu-

lation was adequately measured; how-
ever, it probably often exceeds this figure
and is the most important primary con-
sumer among the local vertebrates);
slender glass lizard, Ophisaurus attenu-
atus, 1.96 kg (in brushy former pastures
dominated by tall brome grass); Great
Plains skink, Enmeces obsoletus, 1.86 kg
(in rocky, open hilltop habitat in 1949,
but subsequently almost eliminated by
unfavorable changes in the course of
vegetational succession); worm snake,
Carphophis vermis, 1.67 kg (on rocky,
wooded hillside and hilltop); five-lined
skink, Eumeces fasciatiis, 1.34 kg (in
open woodland soon after livestock were
removed, but succcssional trends subse-
quently caused marked reduction); cop-
perhead Agkistrodon contortrix, 1.23 kg;
cotton rat, Sigmodon hispidus, 1.11 kg
(probably often exceeded by this primary
consumer, but the cotton rat is also sub-
ject to drastic reduction and is usually
much below this figure even at times and
places where habitat is favorable).

These incomplete and preliminary
figures indicate that Diadophis attains
biomass levels comparable with those of
the most abundant vertebrate primary
consumers and much exceeding the other
local reptiles that feed upon inverte-
brates such as insects and annelids.

Traits peculiar to the local popula-
tion of Diadophis must have arisen over
a long period of selection and adjustment
to climate, habitat and biotic associates.
On the other hand profound changes in
the species' ecology must have been
brought about within the past 125 years
as the region was altered from its orig-
inal state by settlement, agriculture and
industry. Within the ecotone between
deciduous forest and prairie, there was
formerly an abrupt transition from wood-
land to grassland, but tree-cutting, elim-
ination of prairie fires, and pasturing of
livestock destroyed the clear-cut bound-
aries and created a much more extensive
transitional zone with a mixture of wood-
land and prairie types of vegetation that

52

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

provided better habitat for ringneck
snakes than either type in its pure form.
Meanwhile the common and adaptable
earthworm Allolohophora caJU^inosa, in-
troduced from the Old World, became
established in a thriving population that
provided an abundant new food supply
for the ringneck snake. Although there

are on record no actual observations of
population changes in Diadophis, it may
be speculated that the vastly improved
habitat and food supply resulting from
man's alteration of the original forest-
prairie ccotone, caused a population ex-
plosion of the snakes. Perhaps they are
more abundant now than ever before.

LITERATURE CITED

Anderson, P. 1965. The reptiles of Missouri.
LTm\'. Missouri Press, Columbia, xxiii +
330 pp.

Barbour, R. W. 1950. The reptiles of Big
Black Mountain, Harlan County, Kentucky.
Copeia, 1950:100-107.

Blanchard, F. N. 1926. Eggs and young of
the eastern ringneck snake Diadophis punc-
tatiis edwaidsi. Pap. Michigan Acad. Sci.
Arts, Lett., 7:279-292.

Blanchard, F. N. 1930. Further studies of
the eggs and young of the eastern ringneck
snake Diadophis punciatus edwaidsi. Bull.
Antivenin Inst. Amer., 4(1):4-10.

Blanchard, F. N., 1937. Eggs and natural
nests of the eastern ringneck snake, Dia-
dophis punctatus edwardsi. Pap. Michigan
Acad. Sci. Arts, Lett. 22:521-532.

Blanchard, F. N. 1942. The ringneck snakes,
genus Diadopliis. Bull. Chicago Acad., Sci.,
7(1):1-144.

Blanchard, F. N., Finster, E. B. 1933. A
method of marking living snakes for future
recognition with a discussion of some prob-
lems and results. Ecology, 14:334-347.

Blanchard, F. N., Gilreath, R., Blanchard,
F. C. in press. The eastern ring-necked
snake (Diadophis punctatus edwardsi) in
northern Michigan. Jour. Herpetology.

Buikema, a. L. Jr., Armitage, K. B. 1969.
The effect of temperature on the metabolism
of the prairie ringneck snake, Diadophis
punctatus arnyi Kennicott. Herpetologica,
25:194-206.

Clark, D. R., Jr. 1966. A funnel trap for
small snakes. Trans. Kansas Acad. Sci.,
69(l):91-95.

Clark, D. R. Jr. 1967. Experiments into se-
lection of soil type, soil moisture level, and
temperature by five species of small snakes.
Trans. Kansas Acad. Sci., 70:490-496.

Clark, D. R. Jr. 1970. Ecological study of
the worm snake, Carphophis vermis ( Ken-
nicott). Univ. Kansas Publ. Mus. Nat.
Hist., 19(2):85-194.

CoNANT, R. 1948. Regeneration of clipped
subcaudal scales in a pilot blacksnake. Chi-
cago Acad. Sci., Natural History Miscel-
lanea, no. 13:1-2.

DoBZHANSKY, T. 1951. Gcuetics and the ori-
gin of species. Columbia Univ. Press, New
York.

Dundee, H. A., Miller, M. C. 1968. Aggre-
gative behavior and habitat conditioning by
the prairie ringneck snake, Diadophis punc-
tatus arnyi. Tulane Studies Zool. Bot., 15:
41-58.

Ernst, S. G. 1962. Notes on the life history
of the eastern ringneck snake. Turtox News,
40 (10): 266-267.

Fitch, H. S. 1951. A simplified type of funnel
trap for reptiles. Herpetologica, 7:77-80.

Fitch, H. S. 1956. Temperature responses in
free-living amphibians and reptiles of north-
eastern Kansas. Univ. Kansas Publ. Mus.
Nat. Hist., 8(7):417-476.

Fitch, H. S. 1958. Home ranges, territories
and seasonal movements of vertebrates of
the Natural History Reservation. Univ.
Kansas Publ. Mus. Nat. Hist., 11(3) :63-326.

Fitch, H. S. 1960. Autecology of the copper-
head. Univ. Kansas Publ. Mus. Nat. Hist.,
13:85-288.

Fitch, H. S. 1963. Natural history of the
racer, Cohiher constrictor. Univ. Kansas
Publ. Mus. Nat. Hist., 15:351-468.

Fitch, H. S. 1965. An ecological study of the
garter snake, Thamnophis sirtaJis. Univ.
Kansas Publ. Mus. Nat. Hist, 15:49.3-564.

Fitch, H. S., Fleet, R. R. 1970. Natural
history of the milk snake (LampropeUis
trian^tdum) in northeastern Kansas. Herpe-
tologica, 26(4):387-396.

Gehlbach, F. R. 1974. Evolutionary relations
of southwestern ringneck snakes (Diadophis
punctatus). Herpetologica, .30:140-148.

Greene, H. W. 1973. Defensive tail display
by snakes and amphisbaenians. Jour. Herpe-
tology, 7(3) : 14.3-161.

Henderson, R. W. 1970. Feeding behavior,
digestion and water requirements of Dia-
dophis punctatus arnyi Kennicott. Herpe-
tologica, 26(4):520-526.

Klauber, L. M. 1956. Rattlesnakes. 2 vols.
LIniv. California Press, Berkeley.

MacArthur, R. H., Wilson, E. O. 1967. The
theory of island biogeography. Princeton
Univ. Press, Princeton, New Jersey.

FITCH: A DEMOGRAPHIC STUDY OF THE RINGNECK SNAKE

53

Myers, C. W. 1965. Biology of the ringneck
snake, Diadopliis punctatus, in Florida.
Bull. Florida State Mus., 10:43-90.

Noble, G. K., Clausen, H. J. 1936. The ag-
gregation beha\'ior of Storeria dekayi and
other snakes with especial reference to the
sense organs involved. Ecol. Monogr., 6:
269-316.

Petersox, H. W. 1956. A record of \'iviparity
in a normally oxiparous snake. Herpetolog-
ica, 12(2): 152.

PiAXKA, E. R. 1970. On r- and K-selection.
American Naturalist, 104:592-597.

Smith, A. K. 1973. Feeding strategies and
population dynamics of the bullfrog (Rana
catesbeiana). ( Unpublished dissertation,
University of Kansas Library).

Vax Dexburgh, J. 1922. The reptiles of west-
ern North America. Occas. Papers Cali-
fornia Acad. Sci., 10:1-1028.

Wright, A. H., Wright, 1957. Handbook of
snakes of the United States and Canada.
Comstock Publ. Associates, Cornell Univ.
Press, vol. 1, xviii + 564 pp.

\

AVAILABLE MISCELLANEOUS PUBLICATIONS

IN HERPETOLOGY

UNIVERSITY OF KANSAS MUSEUM OF NATURAL HISTORY

52. Reproductive cycles in lizards and snakes. By Henry S. Fitch. Pp. 1-247, 16 fig-
ures in text. June 19, 1970. Paper bound, $5.00 postpaid.

53. Evolutionary relationships, osteology, and zoogeography of leptodactyloid frogs.
By John D. Lynch. Pp. 1-238, 131 figures in text. June 30, 1971. Paper bound,
$7.00 postpaid.

55. Middle American Hzards of the genus Ameiva (Teiidae) with emphasis on geo-
graphic variation. By Arthur C. Echtemacht. Pp. 1-86, 28 figures in text. De-
cember 14, 1971. Paper bound, $3.00 postpaid.

57. A systematic review of the Teiid lizards, genus Bachia, with remarks on Heter-
odactylus and Anotosaura. By James R. Dixon. Pp. 1-47, 15 figures in text.
February 2, 1973. Paper boimd, $1.50 postpaid.

59, Systematics and evolution of the Andean lizard genus Pholidobolus (Sauria:
Teiidae). By Richard R. Montanucci. Pp. 1-52, 8 figures in text. May 14, 1973.
Paper bound, $1.75 postpaid.

61. Reproductive strategies in a tropical anuran community. By Martha L. Crump.
Pp. 1-68, 13 figures in text. November 15, 1974. Paper bound, $4.50 postpaid.

62. A demographic study of the ringneck snake (Diadophis punctatus) in Kansas. Pp.
1-53, 19 figures in text. April 3, 1975. Paper bound, $3.00 postpaid.

MONOGRAPHS IN HERPETOLOGY

1. The hylid frogs of middle America. William E. Duellman. Pp. 1-753, 324 text
figures, 72 color plates. 1970. Two volumes, cloth bound, $25.00 postpaid.

PUBLIC EDUCATION SERIES

1. Amphibians and reptiles in Kansas. Joseph T. Collins. Pp. 1-283, 8 figures, 91
maps, 103 halftones in text. 1974. Paper bound, $5.00 postpaid.

2. Illustrated guide to amphibians and reptiles in Kansas / An identification manual.
Daryl Kams, Ray E. Ashton, Jr., and Tom Swearingen. 28 pages, many illustra-
tions. Paper bound, $1.00 postpaid.

I

f

oA,

erne

Bookbinding Co.. Inc.
300 Summer St(e«it
Btston. Mass. 0??!') .

3 2044 093 361 632

Date Due