.“I- e n
‘Aiaméfl

inn-us... .ua’. II but:

0437--...-

 

This is to certify that the
thesis entitled

EFFECTS OF DENSITY AND ENVIRONMENTAL STIMULI
ON AGGREGATIVE BEHAVIOR AND GROWTH IN

BUFO AMERICANUS TADPOLES
presented by

 

D. Conrad Griffith

has been accepted towards fulfillment
of the requirements for

Master of Science dggreein Zoology

Z/flflfl

Major professor

/
Date / ”$7 (6) Robertmbbins

0.7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

MSU

LIBRARIES
“

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. flfl§§_will
be charged if book is
returned after the date
stamped below.

 

 

 

 

EFFECTS OF DENSITY AND ENVIRONMENTAL STIMULI ON AGGREGATIVE
BEHAVIOR AND GROWTH IN BUFO AMERICANUS TADPOLES

 

By

David Conrad Griffith

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Zoology

1985

ABSTRACT

EFFECTS OF DENSITY AND ENVIRONMENTAL STIMULI ON AGGREGATIVE
BEHAVIOR AND GROWTH IN BUFO AMERICANUS TADPOLES

 

By

D. Conrad Griffith

Light, heat, and food were tested as possible

environmental stimuli which attract Bufo americanus tadpoles

 

to aggregations. Numbers of tadpoles in different sectors of
an artificial pond were recorded with one environmental
stimulus present, and compared with a no-stimulus control.
Food, light combined with heat, heat alone, and light alone
were observed to attract tadpoles, their relative efficacies
being in the order presented.

Tadpoles isolated from early cleavage were tested for
aggregational tendency when introduced to a group at
different deve10pmental states. Much variabilty was observed
in both isolates' and pool-raised tadpoles' reactions to
conspecifics; no trends were observed.

Length and develOpmental state were measured for tadpoles
in isolation, a pool environment, and tanks with one
individual (semiisolate) separated by a screen partition from
a group (cohorts). Both isolates and semiisolates developed

faster than either cohorts or pool-raised tadpoles.

ACKNOWLEDGMENTS

I wish to thank all the people who have helped me with
this study. I am indebted to Dr. Robert Robbins for his
guidance. His knowledge of everything from statistics to
grammar helped me bring order out of chaos. I would also
like to thank the other members of my graduate committee,
Drs. John King and Hiram Fitzgerald. Their input was much
appreciated. A very special thank-you goes to Bruce Jayne
who helped me with all stages of the study, from collection
of specimens to the kind use of his word processor. To all
the other individuals who provided helpful comments and
support, I extend my sincere thanks. Finally, to my parents,
David and Mary Jane Griffith, I would like to say that I
couldn't have done it without your support from the home

front.

ii

TABLE OF CONTENTS

LIST OF TABLES..............................................v
LIST OF FIGURES............................................vi
INTRODUCTION................................................1
LITERATURE REVIEW...........................................6
I. Environmental Stimuli...................................6
II. Isolate Aggregation....................................9
III. Growth Study.........................................13
GENERAL METHODS............................................17
ENVIRONMENTAL-STIMULUS EXPERIMENTS.........................20
A. Overall Introduction...................................20
B. Pilot Study-darkness control...........................23
Introduction..........................................23
Procedure.............................................23
Results...............................................23
Conclusions...........................................27

C. Food as a Stimulus.....................................30
Introduction..........................................3O
Procedure.............................................3O
Results...............................................31
Conclusions...........................................31

D. Hot Spots as Stimuli...................................33
Introduction..........................................33
Procedure.............................................33
Results...............................................34
Conclusions...........................................34

E. Light as a Stimulus....................................38
Introduction..........................................38
Procedure.............................................38
Results...............................................38
Conclusions...........................................4O

F. Heat as a Stimulus.....................................41
Introduction..........................................41
Procedure.............................................41
Results...............................................41
Conclusions...........................................42

G. Light in Pool Center...................................43
Introduction..........................................43
Procedure.............................................43
Results...............................................43
Conclusions...........................................44

H. Stimulus Comparisons...................................46
Introduction..........................................46
Procedure.............................................46
Results...............................................16

8

comeluSionSOOOOOOOOOOO:O:OOO00.......OOOOOOOOOOOOOOOOO
iii

Table of Contents (continued)

ISOLATE/AGGREGATION EXPERIMENTS.
IntrOductionO O O O O O O O 0 O O O O O O 0 O 0 O O O o O O O O O O 0 O O O O O O O O O O O O O O O O .51
Procedure. 0 O O O I O O I O O O O O O O O O O O O O O O O O O 0 O O I O O O O O O O O O O O O O O O O O .52

ResultSOOOOCOOO...OOOOOOOOOOOOOO..0OOOOOOOOOOOOCOOOOOOOOOOSS
COflCluSionBOCOOOOOOOOOOOOOOOOOOOO0.0...0.00.00.00.0000000057

GROWTH STUDIES. 0 O O O O O O O O O O O I O O O O O O O O O O O O O O O O O O O O O C O O C O O O I O .55
IntrOduction. O I O O O O O O O O O O O O O O O O O O I O O I O O O O O O O O O O O O O O O O O O O O .59
Procedure. 0 O O O O O O O O I O O O O O O O O O O O C O O O O O O O O O O O O O O I O C O O O O O O O I .60

...........................51

ResultSOOO0.00000000000000000000000000000IOO0.0.0.0000000061

conCIuSionSOOOOOOOO0.0000000000000000000000000.00.00.00.0065

GENERAL DISCUSSION. 0 O O O O O O O I O O O O O O O C O O O O O O I O O O O I O O O O O O O O O O .69
BIBLIOGRAPHY. O O O O O O O O O O O O O O O O 0 O O O I I I O O O O O O O O O O O O O O O O O O O O O O .75
General References........................................79

iv

LIST OF TABLES

Types of aggregations.......................................2
Aggregational classification................................2
Summary of experiments on sibling recognition..............11
Combinations of stimuli tested.............................22
Results of ring analysis, darkness test....................25
Results of "pie wedge" analysis, darkness test.............26
Mean proportions for 24 ranked sectors, darkness test......28
Food as a stimulus.........................................32
Hot spots as a stimulus....................................36
Light as a stimulus........................................39
Results of test with bulb at center of pool................45
Food and hot spots compared as stimuli.....................47
Food and light compared as stimuli.........................47
Light and hot spots compared as stimuli....................49
Results of isolate aggregation experiment..................56
Results of growth study, June 15...........................64
Results of growth study, June 27...........................64

LIST OF FIGURES

Potential size difference model............................15
Test pool..................................................18
Special tank...............................................18
Number of tadpoles in "pie wedges".........................25
Sector light intensities and temperatures..................35
Sociality apparatus........................................54
Inner and outer one—half areas of bucket...................54

vi

INTRODUCTION

The larvae of anuran amphibians exhibit many specialized
adaptations to an aquatic existance. Behavioral interactions
among tadpoles manifest themselves in several ways, among the
most overt and well documented being the formation of
aggregations. While the majority of anuran species' tadpoles
do not aggregate, numerous examples of aggregation are
documented in the families Pelobatidae, Ranidae, Bufonidae
and RhinOphrinidae.

Several different types of aggregations have been
described, and almost as many systems of aggregate
classifications have been suggested. For clarity, I will
employ the classification system of Bragg (1954, 1968). He

defines, for the spadefoot toad ScaphiOpus hurterii, two

 

major types of aggregations: (1) those based on communal
feeding, and (2) those based on metamorphosis. Feeding
aggregations are further broken down into asocial
aggregations, in which individuals cluster only around a food
source, and social aggregations, that may involve feeding,
but do not require the presence of food (Table 1). Feeding
aggregations can contain anywhere from three to thousands of
individuals (Beiswenger, 1975) as long as there is an abrupt

end to the group at its margins. Aggregations can be
1

Table 1. Types of aggregations.

Feeding. - social - aggregate even when food is absent

- asocial - aggregate only when food is present

Metamorphic - social - aggregate in the stages just before

metamorphosis is complete

School - social - involves coordinated swimming, parallel

orientation

Table 2. Aggregational classification. Degrees of contact and
percent area in which tadpoles are found determine
classification.

D1 -
D2-
D3 -
D4-

D5 -

not aggregating
1-5 cm apart, 20-50% of area covered by tadpoles
mostly in contact, 51-84% of area covered by tadpoles

all tadpoles in contact, one layer thick, cover 90-99%
of area

all tadpoles in contact, multiple layers thick, cover
100% of area

- based on Beiswenger (1975)

3

stationary or moving, demonstrating coordinated swimming
and/or head butting (Wassersug and Hessler, 1971; Beiswenger,
1975). Beiswenger has further classified tadpole
aggregations based on number of individuals present, amount
of contact beween individuals and percent area they cover
(Table 2).

Tadpoles come from far-reaching locations in a pond to a
given area in which they aggregate. The tadpoles are
believed to use specific environmental stimuli to converge on
the same general location. They may simply exhibit this
asocial aggregation, using the environmental stimulus as a
focus of attraction, or they may move close enough together
that they can interact socially (Beiswenger, 1977).
Environmental factors which elicit this taxic behavior to a
given area of the pond include heat, light, stream currents
(rheotaxis), oxygen concentration, substrate pattern and
presence of food (Wassersug, 1973). Details on what is known
about each of these environmental cues can be found in the
Literature Review section of this thesis.

Bufo americanus tadpoles typically form dense,

 

conspicuous aggregations containing hundreds to thousands of
individuals (Wassersug, 1973; Beiswenger, 1975, 1977; Waldman
and Adler, 1979; Waldman, 1980). These tadpoles remain
relatively active on the bottom of the pond throughout the
larval period. Social aggregations are first formed by
tadpoles of Gosner stage 25 (1960), about the time the
opercular flaps cover the gills (Beiswenger and Test, 1966;

Beiswenger, 1977).

4

One of the objectives of the current study was to

evaluate quantitatively the ability of E. americanus tadpoles

 

to respond to external environmental stimuli.‘ I wished to
determine if heat, light, and the presence of food can
attract Bufo tadpoles. Due to time constraints and a limited
number of test animals, my experimental apparatus
necessitated that I pit one environmental stimulus against
another on the same test. This had the added advantage of
allowing me to compare the relative ability of the various
stimuli to attract tadpoles.

A second objective of this study was to determine

whether isolated B. americanus tadpoles will exhibit

 

aggregational behavior. There is a great deal of population
variability in aggregational behavior (Wassersug, 1973).
Previous studies on aggregating tendency of isolated tadpoles
have involved tadpoles from very socially-aggregating
populations. I wished to determine if less socially-inclined
tadpoles would also exhibit such a strong tendency to
aggregate after isolation, if a latency period occurs, or if
no aggregation occurs. No study to date has tested isolated
tadpoles from different age classes for their tendency to
aggregate, so I wished to determine if age differences in
aggregational tendencies exist.

Tadpole growth rate is negatively correlated with
density; the more tadpoles per unit area, the slower on
average they grow. This phenomenon may be called the

crowding effect. A notable exception to this occurs in

5

R. dorsalis which grow poorly in low density situations
(Foster and McDiarmid, 1982). Rose (1960) demonstrated that
while the mean growth rate is suppressed during crowding of
Rana pipiens tadpoles, a few individuals grow at an
accelerated rate. The rate of development of these few
tadpoles is independant of density. Two distinct size
classes are observed in R. dorsalis tadpoles (Stuart, 1961).
The third purpose of this study was to look into the

problem of differential growth in Bufo americanus tadpoles.

 

I wished to test the merit of Wassersug's (1973) statement

that B. americanus tadpoles are not very susceptible to the

 

crowding effect. I also wished to provide some insight into
the mechanism(s) by which aggregating tadpoles inhibit each
others' growth. Chemo/physical substances, psychological
stress, and exploitation competition have all been suggested

as possible mechanisms for this phenomenon.

LITERATURE REVIEW

I. Environmental Stimuli

There is no one particular cause for tadpole
aggregations. Among the most cited causes is the enhancement
of food gathering capabilities by a group. In
bottom-dwelling detritis feeders, a group may disturb the
bottom more than an individual, thus exposing more food.

This has been been demonstrated with Scaphiopus hurterii

 

(Bragg, 1965, 1968). When'temporary breeding ponds begin to
dry up, food may become patchy and scarce, further leading to
the likelihood of aggregational behavior (Bragg, 1954, 1965,
1968).

Predation may be another reason for aggregational
behavior. Certain anuran species' tadpoles have epidermal
poison glands, notably Bufo tadpoles, which secrete bufonin
toxins. Experiments have shown that bluegill sunfish

(Lepomis macrochirus) prefer tadpoles without poison glands,

 

 

and in fact, learn to avoid the conspicuous dark mat of

aggregating Bufo americanus tadpoles (Voris and Bacon, 1966).

 

Epidermal glands may also be a source of chemical alarm
signals (Pfeiffer, 1966) which would most benefit members of
a group. The fright reaction in tadpoles is strong. A dead

or dying tadpole causes the whole aggregation to scatter

6

7
(Wassersug, 1973). Finally, Carpenter (1953) hypothesized

that .3222 pretiosa tadpoles may be aggregating to avoid
attachment of parasitic leeches (a problem for tadpoles in
mountain lakes) and to dislodge those leeches which have
already attached themselves to the tadpoles.

Bufo boreas tadpoles are highly pigmented. A group of

 

such tadpoles will absorb more heat than an individual,
thereby raising the temperature of the surrounding water
several degrees Celsius (Brattstrom, 1962). This allows the
poikilothermic tadpole to increase its metabolic rate,
causing development to occur faster, a highly advantageous
trait in a temporary pond.

Many species' tadpoles prefer the warmest areas in the
pond, in the 25-30'C range, including Bufo boreas

(Brattstrom, 1962), Bufo canorus (Mullally, 1853). Bufo

 

punctatus (Tevis, 1966), and Bufo americanus (Beiswenger and

 

 

Test, 1966; Beiswenger, 1977). Numerous other species from

the genera Bufo, Hyla, Rana, Pseudacris, and Scaphiopus

 

 

exhibit this behavior (Rugh, 1962; Wassersug, 1973). Some
species' tadpoles do not prefer the warmest areas, especially
those species which live in mountain streams. For example,

Ascaphus truei (deVlaming and Bury, 1970) prefers water below

 

22°C.

Bufo americanus tadpoles are positively phototaxic

 

(Beiswenger, 1977). In the morning, they move to the
shallowest areas of the pond in response to light. Hyla

versicolor tadpoles are also positively phototaxic when

 

disturbed (Noble, 1954). Franz (1913) has described a number

8

of other species whose tadpoles exhibit the same results.
Wassersug (1973) claims the phototaxic response varies
greatly between species.

Rheotaxis has been described as a most visible factor
affecting asocial aggregation (Wassersug, 1973). Those
tadpoles which develOp in streams are regularly observed to
head upstream. Tactile (lateral line) and visual cues are
believed to be involved in this response.

In oxygen-deficient water, tadpoles come to the surface
to gulp air (Bragg, 1954; Stuart, 1961; Wassersug, 1973).
Shallow water may be sought by individuals in
oxygen-deficient situations, as there is more dissolved
oxygen in shallow areas (Wassersug, 1973).

Wiens (1970) experimentally demonstrated a
species—specific selection for certain substrate patterns in

Rana aurora and Rana cascadae tadpoles. A similar phenomenon

 

 

has been suggested for Bufo americanus (O'Hara, 1974). In

 

nature, such a phenomenon would be difficult to detect.
Presence of food appears to be an almost universal
stimulus to form (typically asocial) aggregations. Risser,

in his studies of Bufo americanus (1914), was one of the

 

first to document the fact that tadpoles use olfactory cues
to locate sources of food. Bragg (1954, 1968) has suggested
that food may be the single most important stimulus to
aggregate.

Once tadpoles are sufficiently close together, they may
begin to aggregate socially (in the absence of food)

(Beiswenger, 1977). Visual and/or lateral line stimuli,

9

which are useful only at close range, may be the cues used by

social aggregates (Wassersug, 1973). Xenopus laevis tadpoles

 

can aggregate using only visual cues (Wassersug and Hessler,
1971), but the addition of lateral line stimuli allows for
parallel orientation (Katz et al., 1981; Lum et al., 1982).

Bufo boreas may need mechanical/tactile stimulation (i.e.,

 

actual physical contact) as well as visual cues to aggregate

socially (Wassersug, 1973).

II. Isolate/Aggregation

Studies have shown that within a social aggregation,the
position of individual tadpoles may not be random. Allee
(1931) suggested "similar" tadpoles tend to remain together
in unusual surroundings to minimize the "disturbing effects"

of the strange environment. Bufo woodhousei tadpoles tend

 

to associate with individuals in the same size class within_
the aggregation (Brenden et a1, 1982), although this is not a
strong tendency. A most intriguing discovery is that both

Bufo americanus and Rana cascadae tadpoles, when given a

 

 

 

choice, prefer to associate with siblings over non—sibs
(Waldman and Adler, 1979; Waldman, 1980; Blaustein and
O'Hara, 1981; O'Hara and Blaustein, 1981). R. cascadae
tadpoles reared with sibs or a mixed sib/non-sib group
preferred to associate with sibs (O'Hara and Blaustein,
1981). Those isolated since Gosner (1960) stage 12 (yolk
plug) also preferred siblings when given a choice. Those
reared with sibs preferred unfamiliar sibs over familiar

non-sibs (Blaustein and O'Hara, 1981). B. americanus

 

10
tadpoles within an aggregation tend to be closer to sibs than
non-sibs. In some cases, two aggregations formed, one for

each sib group (Waldman and Adler, 1979). B. americanus

 

tadpoles reared with sibs or in isolation preferred
associating with sibs over non—sibs. Those reared in a mixed
group showed no tendency to prefer siblings (Waldman, 1980,
1981). These observations have been taken one step further.
Tadpoles can discriminate half sibs from full sibs. B.
cascadae tadpoles prefer full sibs over half sibs, and half
sibs over non-sibs. It has been reported that they can even
distinguish between maternal and paternal half sibs, and
prefer maternal ones (Blaustein and O'Hara, 1982). B.

americanus tadpoles raised in isolation can also distinguish

 

paternal half sibs from full sibs (prefer full sibs), but
they cannot discriminate between maternal half sibs and full
sibs (Waldman, 1981). The above results are summarized in
Table 3.

The most commonly cited explanation for these behaviors

is a kin selection argument. B. americanus tadpoles, being

 

darkly pigmented, form a conspicuous black mat in the pond.
Since they are unpalatable, a predator would soon learn to
associate their conspicuousness with their distastefulness.
By remaining in close proximity to kin, the chances of a
tadpole being eaten are reduced, as one or a few individuals
act as altruists, giving their lives (anthropomorphically
Speaking) so that their relatives may be spared (Waldman and
Adler, 1979; Waldman, 1982). The same kin selection argument

is applied to B. cascadae tadpoles, even though they are not

11

mononmmmna o:

mpamladam

mnflm

monoanmnm o:
mpflmimams assumpme
mnfimuwamn
wnamuaaam

mpHm

mm...“ m .HNHHHENHCS.

mnflm

 

ammonm

spamuafisa\mpamuofimn assumpse
mpamuflasm\mpnmumfime Hmeumpma

mpwmlco:\wpfim

mpHmIGOC\mpHm

mpflmiwam: HmGHumm\mpamlmHmn assumpms

mpfimlnon\mpflmlmfims
mpfimiwam:\mnflmnaazm

mpflmlcon\mnam

mpamaco: Hmaaaemw\mnfiw HmHHHemHG:

mpfimlco:\mpwm

mnoapmo

.QOfipfizmoomm mafianam

cowpmfiomfi
:oapmaomfi

soapmaomfi
mnamuso:\mnflm umxfls
asohw cmxfie

macaw cmxfle

moohm Bonds
scavmaomw

mpfimuco:

mpflmlcon\mnfim vmxHE

 

 

spa: emnmmm

mangHHosm .m

 

mmnmommo .m

mofiommm

no mpnmEHnmmxm Ho mHMEEdm .m manna

12
distasteful to predators. Tadpoles may be able to warn each
other of a predator's presence, and by being close to kin, a
tadpole is most likely to receive an inclusive—fitness
benefit by warning its neighbors (Blaustein and O'Hara,
1981).

It is currently unknown if sibling recognition is
genetically based (Blaustein and O'Hara, 1981; O'Hara and
Blaustein, 1981), learned (Waldman and Adler, 1979; Waldman,
1982) or a combination of the two (Blaustein and O'Hara,
1982). The mechanism of recognition appears to be
chemo-olfactory (Blaustein and O'Hara, 1982), possibly
pertaining to some factor in the jelly mass which surrounds a
clutch of sibling eggs (Waldman and Adler, 1979; Waldman,
1982).

Similar studies performed on Rhinophrynus dorsalis

 

tadpoles produced very different results. Isolates
introduced to a mixed group of about 200, or a small
aggregation of 10 siblings actively avoided the others for
the first ten minutes. They became more tolerant of the
group over the next few hours, but were never observed to
join the aggregation (Foster and McDiarmid, 1982). It is
unknown whether these isolates would continue to refuse to
aggregate, or if they were going through a latency period,
much like that exhibited by schooling fish reared in
isolation (Shaw, 1960).

13

III. Growth Study

Rose found that water conditioned by the larger tadpoles
was enough to inhibit growth of other tadpoles. It was
hypothesized that algal-like cells (perhaps sloughed
epithelial cells) found in the intestinal tract and feces of
tadpoles was responsible for this growth inhibition
(Richards, 1958, 1962). It is currently unknown exactly what
these cells are, but they appear to resemble unicellular
green algae. Licht (1967) found the algal-like cells in the

feces of Bufo valliceps, B. speciosus and B. woodhousei, and

 

 

 

observed the same crowding effect Rose and Richards did.

Gromko et a1. (1973) demonstrated the crowding effect in
B. pipiens in the absence of the algal-like cells. They
suggested the phenomenon is a behavioral response to social
stress, causing adrenocortical hormone secretion. This
stress response may be due to increased contact with other
tadpoles (Bilski, 1921; John and Fenster, 1975). Shvarts and
Pyastolova (1970b) argue that if the inhibitor is a chemical
substance, it would probably be a cellular metabolite which
reduces the activity of the genes responsible for a
particular stage of deveIOpment of tissues or organs. Growth
inhibition is such a widespread phenomenon in plants and
animals, it would be difficult to explain the evolution of a
Special purpose growth inhibitor so many times.

Wilbur (1977) proposed yet another explanation for the
crowding effect. He hypothesized that larger tadpoles were

the ones better able to exploit resources. If food were a

14

limiting resource, larger tadpoles could monOpolize it,
preventing the smaller ones from growing as quickly
(Brockelman, 1969; Wilbur, 1977). Wilbur and Collins (1973)
proposed a simple model which diagrams how an exploitation
scheme might work (Figure 1).

High predation rates may prevent the crowding effect in
nature. It may only be an artifact of laboratory conditions
that we see this reduced growth rate and delayed
metamorphosis (Brockelman, 1969). Regardless of the cause of
the crowding effect, it is highly specific in its effects on
B323 arvalis (Shvarts and Pyastolova, 1970a), with the
greatest effect on the most closely related tadpoles.
Shvarts and Pyastolova believe the crowding effect is a
mechanism which ensures genetic diversity. They claim that
growth inhibition is greatest among close relatives. By only
permitting a few individuals from each family to survive (the
ones which can suppress the growth rate of the others), it is
more likely that more families will be represented in the
next generation. This mechanism, of course, necessitates a
group selection argument for its existence, as
family-specific growth inhibition would not normally be
expected to occur due to the advantages of kin selection.

Shvarts and Pyastolova (1970b) found that the presence
of a group of small B. arvalis tadpoles actually accelerates
the growth rate of the larger ones at the same time it
inhibits the growth of the smaller ones. This gets the
larger ones out of the ponds quickly, allowing the smaller

ones a chance to utilize the remaining resources in the

15

 

 

 

no

 

 

   
  

    
  

initiate
metamorphosis

additional
growth

  

 

 

weight of tadpole

w:
b = minimum size necessary to
metamorphose
b + c a maximum size at which
metamorphosis must occur
g = growth rate below which it is

advantageous to initiate
metamorphosis

- from Wilbur and Collins (1973)

Figure 1. Potential size difference
model. Tadpole continues to grow if
its growth rate is "good" enough, but
will metamorphose if it is not.

16
temporary pond on the outside chance that they may
metamorphose before it dries up.
Wassersug claims some species' tadpoles, such as B339

americanus, are less susceptible to growth inhibition due to

 

selection against it. Constant selection pressure for
individuals which are not affected by growth inhibition may
have created a species-wide immunity to the crowding effect.
Other Species may deal with the problem by metamorphosing at

a smaller size.

GENERAL METHODS

I collected two pairs of Bufo americanus toads in

 

amplexus on the night of May 18, 1984. One pair was obtained
on Bennett Road (on the road itself) one mile south of the
Michigan State University campus. The other pair was found
in a small temporary pond, about 500 meters south of Bennett
Road. Each pair was transferred to a 3.5-liter plastic
bucket where oviposition occurred. Fertilized eggs were
collected on the morning of May 19th, within 12 hours of
ovulation.

Seven-hundred tadpoles were raised under laboratory
conditions from Gosner stage 9 (late cleavage) to stages
21-26 in a circular plastic wading pool, 91.5-cm diameter,
filled 1.5-cm deep with dechlorinated tap water. Air and
water temperature remained a nearly constant 22°C. Light was
provided by overhead fluorescent fixtures set on a 12L, 12D
cycle. The bottom of the pool was divided into 24 sectors,
each with an area of 273.7 cm2, by strips of black electrical
tape affixed to the inside of the pool (Figure 2). Tadpoles
were fed boiled lettuce 2g libitum. Water in the pond was
changed at three-day intervals, using dechlorinated tap

water.

17

18

 

 

Figure 2. Test pool. Pool is divided
into 24 sectors of equal area.

 

 

 

 

 

 

 

 

 

 

 

 

 

F I
I .. I ......_I L_....._J

Figure 3. Special tank. On left is front view of tank as
seen through clear plastic front. 0n right is clear
plastic divider with screen.

 

 

19

Fifty tadpoles were raised in isolation in one-pint
opaque plastic cups, 9-cm diameter, filled 1.5-cm deep with
dechlorinated tap water. In addition, 24 special tanks (17.5
cm X 9.5 cm) were constructed of plywood and white pine
(Figure 3) and painted gloss white inside. A screened
partition made of 1-mm2 mesh nylon window screen glued to a
clear Plexiglas frame was positioned to isolate one third of
the tank. The tanks were lined with clear plastic film. I
was careful to expose the tadpoles only to inert plastic
materials and Silicone glue to prevent chemical or metal ion
contamination of the water. A single tadpole was raised on
the Side with the larger area (two-thirds of the tank), and a
group of 11-19 tadpoles was raised on the side with the
smaller area (one-third of the tank). Care and feeding of
tadpoles in the pint cups and Special tanks was the same as

for those in the pool.

ENVIRONMENTAL—STIMULUS EXPERIMENTS

A. Overall Introduction
I wished to test three environmental stimuli in their

ability to attract B. americanus tadpoles: heat, light, and

 

presence of food. Due to time constraints and a limited
number of test animals, I chose to look at two stimuli at
once, one stimulus on each Side of the pool. This had the
added advantage of allowing comparisons of the relative
abilities of the three stimuli to attract tadpoles.

B. americanus tadpoles in the pool were raised to stages

 

24-27 before testing began. Tadpoles are mobile at stage 21,
so they had been free to move about for several days. A
removable cardboard partition was fashioned to prevent light
from the fluorescent fixtures from reaching one side of the
pool. This partition went down to the water's edge, yet
allowed the tadpoles to move freely between sides. Two
incandescent lamps, one with a 15-watt bulb and the other
with a 50-watt bulb were affixed on the side opposite that
with the fluorescent light. The intensities of the
incandescent and fluorescent light sources were adjusted so
as to be nearly equal. The incandescent light's intensity
was 93.4% that of the fluorescent light, as measured at the

water surface by a Photovolt model 502M photovoltmeter.

20

21

Due to time and equipment constraints, not all
combinations of the three environmental stimuli were tested.
I looked at those which were possible to set up within the
time framework I had to work with, and complied with the
physical constraints of my laboratory. A summary of these
combinations is found in Table 4.

In each test, the starting positions of the tadpoles
were initially randomized by thoroughly mixing the water.
This also reduced the likelihood of the formation of an
oxygen-gradient. The environmental cues were established in
either Side of the pool and the tadpoles were allowed to move
about freely for 15-20 minutes, after which I counted the
number tadpoles on each side. Four such trials were made for
each test (exception: those with heat as the stimulus on one

side - see Table 4).

22

Table 4. Combinations of stimuli tested.

 

 

 

Stimulus 1 vs. Stimulus 2 No. Trials
heat darkness 1
light, heat light 4
food light 4
light, heat food 4
light, heat light, food 4
darkness darkness 4
light in center darkness 3

23

B. Pilot Study - darkness control

Introduction

 

My first objective was to get some baseline data
concerning what tadpoles do in the absence of any external
stimuli. The possibility exists that tadpoles continue to
aggregate without external stimuli, or a random orientation
may occur, although a behavior intermediate to the two could

also be possible.

Procedure

 

As previously outlined, initial tadpole orientation was
randomized. I then turned off the lights which resulted in
complete darkness. There was no source of food, nor any
known thermal gradients in the pool. The tadpoles were
allowed to acclimate for 15-20 minutes, after which a red,
low-wattage overhead lamp was switched on. I then quickly
recorded the number of tadpoles in each sector. Because B.

americanus tadpoles are rather lethargic, very few changed

 

position during the counting, which was done with a
thumb-activated hand counter. Four such trials were made,

each on a different day.

Results
I analyzed these four darkness trials in two ways.
First, I looked at the number of tadpoles in the outside,

middle, and inner rings, of eight sectors each. Note that

24

results are reported only for analysis on the four trials
combined. Results of analyses on the four separate trials
were the same as the combined results. The outside ring
contained 1074 tadpoles, the middle ring 855, and the inner
ring 906 tadpoles. I compared two rings at a time using
chi—square analysis (Table 5). My expectation was an equal
number in each ring, since each sector and hence, each ring
was of equal area. Next, I analyzed the data by "pie wedge",
adding the outer, middle and inner sectors for each wedge
together (Figure 4). Here, each of the four trials was
analyzed separately (Table 6).

The results of the first analysis indicate that there is
a preference for the outside ring of the pool. There appears
to be no significant difference between the middle and inside
rings. Results of the second analysis are more difficult to
interpret. In every trial, there is a significant deviation
from random, but the wedges with high and low counts are not
the same in each trial (Kruskal-Wallis test, H = 12.95, p 5
.10). These data suggest that even in darkness, there
appears to be a tendency to aggregate, usually toward the
outside, but not necessarily in the same location around the
pool each time.

In order to provide accurate values for the expectations
in future analyses, I needed a way to take this natural
aggregating tendency into account. To do this, I ranked the
sectors in each darkness trial according to the number of
tadpoles per sector. I calculated the proportion of the

total number of tadpoles in each sector. I then averaged the

Table 5. Results of ring analysis, darkness
test. Outside, middle and inner rings of 8
sectors were compared for number of tadpoles
present by chi-square analysis.

 

 

 

Comparison* B3 Significance Preference*
o - m 24.9 E 5 .005 o
m - i 1.5 (D050) "
o - i 14.3 ‘p 5 .005 o

* o = outside ring

m = middle ring

i = inside ring

404

296

373

319

 

404

428

 

315

301

Figure 4. Number of tadpoles

in "pie wedges".

Total

number of tadpoles in all
four darkness trials are

shown.

26

Table 6. Results of "pie wedge"
analysis, darkness test.

All eight "pie wedges" were
compared by chi-square analysis
for four trials.

Trial 33 Significance

1 28.6 p 5 .005
2 34.9 2,5 .005
3 41.6 2’5 .005
4 16.8 p 5 .025
total 55.6 p 5 .005

27

proportions for each rank for all four trials. For instance,
each trial had a sector of rank 1 (fewest number of
tadpoles). The mean proportion was calculated for the four
sectors with rank 1, one sector from each of the four trials.
The result was 24 mean proportions (Table 7). These were

used as expectations in experimental tests. In the following

 

analyses, I ranked the sectors in each trial as described
above, and compared the observed proportion of the total
number of tadpoles in a given sector with the corresponding
expected proportion for that same ranked sector. I used
these observed and expected proportions to do chi—square
analyses. In actual practice, I converted expected
proportions to expected numbers and compared these with
observed numbers.

This method of analysis tests against the "worst case"
for the expectation. In the worst case, the tadpoles
naturally aggregate to the degree described in the pilot
study. Any observed aggregation must be significant above
and beyond this natural aggregation. In this respect, the
test is quite good at preventing type II error in analysis,
because any aggregation significant in this worst-case
situation would also be so in anything less than the worst

case.

Conclusions

 

It appears that B. americanus tadpoles will aggregate

 

loosely even in the absence of any known environmental

stimuli, specifically light, heat, an oxygen gradient, or

28

Table 7. Mean proportions
sectors, darkness test.

.017
.022
.027
.029
.031
.032

R7 =
28 =
29 =
1310=
I914"=
«1312=

.034
.035
.036
.037
.044
.039

213
B14
B15
216
217
1918

for 24 ranked

.043
.044
.045
.046
.049
.050

E19
220
221
B22
223
224

.051
.053
.055
.058
.062
.068

29

food. AS the statistics Show, tadpoles tended toward the
outside, but not toward any one particular area in the pool.
My subjective observations confirm this; the aggregations
were in the D1 to D2 range (Beiswenger, 1975), but no single
area or substrate pattern was preferred. The water was
slightly shallower toward the very edge of the pool, a
consequence of its leped sides. Perhaps tadpoles seek
Shallow water under the "assumption" that this is where food
or some other important resource is located.

In any event, the data generated from this study were
quite useful in looking at the rest of the experiments, as
chi-square expectations could now be modified to reflect more

accurately a null (no stimulus) expectation.

C. Food as a Stimulus

Introduction

 

Food was tested as a possible cue for aggregation. Food
in vernal ponds is typically found in patches around the
edges. Tadpoles may use some chemo-olfactory cue from the

food to home in on it, hence forming asocial aggregations.

Procedure

 

Boiled lettuce leaves served as the source of food.
When used as an experimental attractor, the food was not
randomly scattered about the one side of the pool, but was
confined to four sectors, randomly chosen for each trial.
This more closely resembles the patchiness of the
distribution of high quality food that is found in B.

americanus' natural environment.

 

The tadpoles were not fed for at least a few hours
before this test was run to ensure they were hungry. Tadpole
starting location was randomized and the lettuce was added.
I waited for 15-20 minutes, then counted the number of
tadpoles in the four food-containing sectors and the eight
other sectors on the same Side of the pool for each of the

four trials.

30

31

Results

In comparing the four food—containing sectors with the
others on the same side of the pool (and in all subsequent
tests), I used chi-square analysis with the expectations
calculated earlier (see pilot study). A total of eight
trials were looked at from two different tests, food versus
fluorescent light, and food versus light and heat. In all
cases, the food sectors were significantly preferred over the

non-food sectors on the same side (Table 8).

Conclusions

 

Tadpoles grow as quickly as possible in order to
metamorphose. As pointed out in the literature review,
feeding rate and growth rate are directly correlated. It is
therefore highly advantagous for a tadpole to be able to find
sources of food. Since tadpole eyes are poorly develOped, it
may be that a chemo-olfactory stimulus is being detected by
the hungry tadpole. Feeding aggregations are the most
commonly described type of asocial aggregation (Bragg, 1954),

and my results indicate that B. americanus tadpoles are quite

 

capable of finding isolated sources of food, forming dense

asocial aggregations about them.

32

Table 8. Food as a stimulus. First
light, then light and heat were

pitted against food. Number of tadpoles
in food sectors was compared with

number in non-food sectors on the same
side of the pool using chi-square
analysis.

- Food vs. light

 

Trial xi Significance
1 75 p 5 .005

21 p_< .005
84 2,5 .005

#WN

86 2’5 .005

- Food vs. light, heat
Tgigi BE Significance
1 35 2’5 .005

290 p 5 .005

180 p_5 .005

#Wm

20 p < .005

33

D. Hot Spots as Stimuli

Introduction

 

"Hot spots" are areas of increased light intensity and
warmer temperature. This combination of two stimuli was
looked at as a potential environmental cue for tadpole

aggregation.

Procedure

 

When light and heat were to be used in combination, the
incandescent lamps were used. The bulbs were approximately
25 cm from the water surface. They provided virtually the
same light output as the fluorescent light, and were capable
of raising the temperature of the water below them by 2°C.

Light intensity and temperature were not uniform in the
incandescent light experiments, nor was the light from the
fluorescent bulb of uniform intensity at all points on the
pool. I measured light intensity and temperature in all 24
sectors with the the incandescent bulbs set up on one side
and the fluorescent light on the other (partitioned by the
cardbord divider). Light readings were made with a Gossen
Luna-pro light meter at the water surface. Light meter
readings ranged from 117 to 613 lux; temperature ranged from
19.4 to 21.1‘C. Both stimuli were rescaled from 1 to 100 to
give them equal ranges. It is not known whether the tadpole

adds the two stimuli together or multiplies them in its

34

perception of the total stimulus, so both methods are
represented in Figure 5. In both cases, there were four
sectors on the incandescent (light, heat) side whose light
intensities and temperatures exceeded the others. These were
the ones directly under the bulb. I have called these four
sectors the hot spots; they are pictured in Figure 5. On the
fluorescent (light, no heat) Side, there were six sectors
with light intensities greater than the others, and
temperature was a constant 19.4'C on that side of the pool.
Counts of the number of tadpoles in each of the four hot
spot sectors and the other eight sectors on that side of the

pool were made.

Results

Hot Spots were compared with non-hot spots on the same
side of the pool. In this case, eight trials from the two
tests of light and heat versus light, and light and of heat
versus food were used. In seven of the eight trials, the hot
spot sectors were significantly preferred over the

non—hot—spot sectors (Table 9).

Conclusions

 

In nature, heat generally accompanies light and
vice—versa. The pool's hot spots were therefore probably not
an artificial anomaly. The edges of a pond are the most
likely places for hot spots to occur, where the water is

shallow enough that temperature can be quickly raised, and

35

10000“ fl“
IOOOOIOOOO

 

 

 

 

 

32 .9- 3 71
I396 150 75
l396 90 H5
40 ‘99 41 u
40 so 41 u
9' ‘ 4O 25 9. ~ H 4| 25 2'
40 40 4| 41
4o 25 4| 26
LIGHT X TEMPERATURE LIGHT «4 TEMPERATURE

Figure 5. Sector light intensities and temperatures. Both
stimuli are scaled from 1 to 100 and multiplied (left) or
added (right) together, resulting in four hot-spot sectors,
within heavy borders.

36

Table 9. Hot spots as a stimulus.
First light, then food were pitted
against light and heat (hot spots).
Number of tadpoles in hot-Spot

sectors was compared with number in
non-hot-spot sectors 0n the same side
of the pool using chi-square analysis.

- Light, heat vs. light

Trial 13 Significance

 

1 18.9 2 5 .005
2 3.4 p 5 .10*
3 37.8 p 5 .005
4 5.9 E 5 .025

* not significant
- Light, heat vs. food

Trial B3 Significance

1 8.3 2 5 .005
2 23.0 p 5 .005
3 35.0 p 5 .005
4 120.0 2 5 .005

37
light easily penetrates to the bottom (where the benthic

tadpoles are found).

Clearly, hot spots, as they are defined in this
experiment, are attractors for tadpoles. The hot spots were
strongly preferred over non-hot spots in all the trials but
one, and in that case, they were preferred, but not
significantly so. Any attractor of tadpoles is a candidate
for a stimulus to aggregate. If, indeed, hot spots occur in
nature, they would be ideal environmental stimuli for asocial

aggregation.

38

E. Light as a Stimulus

Introduction

 

Light has often been cited as a stimulus for attacting
tadpoles, as the literature review indicates. Whether or not
it could act as a focus for aggregation in my experimental

apparatus was tested next.

Procedure

 

The source of light by itself (with no heat) was the
overhead fluorescent light under which the tadpoles had been
raised. I counted the number of tadpoles in each of the
twelve sectors on the fluorescent light side of the pool for
each of the four trials. Recall from Section D (hot spots as
stimuli) that six of the fluorescently-lit sectors had

greater light intensities than the other six.

Results

I compared the six fluorescently-lit sectors with the
highest light intensities with the six others on the same
side, using the data from the experimental tests of light
versus light and heat, and light versus food. In six of the
eight trials, there were no significant deviations from the
expectation (Table 10). In the two cases which did deviate,
there were significantly more tadpoles in the less—intensely

lit sectors than was expected.

39

Table 10. Light as a stimulus. First
light and heat, then food were pitted
against light. Number of tadpoles in
the Six most intensely-lit sectors
was compared with the number in the
six least intensely-lit sectors on
the same side of the pool using
chi-square analysis.

— Light vs. light, heat

Trial 33 Significance

 

1 .09 p 5 .80*
2 1.09 2,5 .30*
3 4.86 2,5 .05

4 1.23 2 5 .30*

* not significant
- Light vs. food

Trial BE Significance

1 .46 2’5 .50*
2 .76 p 5 .50*
3 7.40 p 5 .01

4 .11 p 5 .80*

* not significant

40

Conclusions

 

In most cases, tadpoles were not preferentially
aggregating in the sectors with the higher light intensities;
in the two cases which did Show a preference, it was toward
the less-intensely lit sectors. These results indicate that
fluorescent light itself may not be a good stimulus for
aggregation.

The quality of the light may be very important for

pr0per attraction of B. americanus ltadpoles. Perhaps

 

tadpoles are more sensitive to certain wavelengths of light
than others. Since incandescent light has the added factor
of heat, it may be that both stimuli are necessary to obtain
a response. This hypothesis would seem logical, Since in the
vernal pond, one does not usually find light without heat.
Future investigations could look into the ability of tadpoles
to detect and respond to different wavelengths of light, all

in the absence of heat.

41

F. Heat as a Stimulus

Introduction

 

It may seem like an unlikely situation where there is a
source of heat in a pond but no light, but one purpose of
this study was to attempt to dissect out the important
component(s) of a complex stimulus-filled environment which
the tadpole uses to aggregate. For this reason, heat alone
was looked at as a potential environmental cue for

aggregation.

Procedure

 

Heat was provided by a submersible aquarium water heater
which was capable of raising the water temperature from 22°C
nominal to 28°C. As with the food tests, I quickly counted
the number of tadpoles in each sector on the heat side (and
also on the other side of the pool) under a red, low-wattage

lamp suspended from the ceiling.

Results

Unfortunately, only one trial of heat versus darkness
was performed before the aquarium heater malfunctioned. The
four warmest sectors (which were the same as the four
hot-spot sectors) had a temperature of 28°C. They were
compared with the other eight on that side (with temperatures
ranging from 22—26'C), and no significant deviation from

expected was observed (X2 = 2.84, p 5 0.10). These four

42
warmest sectors did, however, have significantly more
tadpoles than expected when compared with the twelve sectors
on the other side of the pool, each being 22°C (X2 = 5.34,
p 5 0.025).

Conclusions

 

Heat, it appears, is a good attractor stimulus over a
broad area, as witnessed by the significant deviation from
expected number of tadpoles in the four warmest sectors
versus the twelve on the other side. No significant
difference from expectation was found when the four warmest
sectors on the heat side were compared to the other sight on
that side because temperature gradually went from 28° to 22°
on that side; there was no sudden change in temperature
between sectors.

Since the light-as-a-stimulus experiments showed that
light (at least fluorescent light) is not a very good
stimulus for attraction, and Since the current study
demonstrates that heat is a good "broad" area attractor, I
suggest that it was the heat in the hot-spot sectors which
attracted tadpoles to the general area, and the more intense
light that brought them specifically to the four hot spot

sectors.

G. Light in Pool Center

Introduction

 

As the pilot study indicated, there is a tendency for
tadpoles to aggregate, even in darkness, toward the edge of
the pool. I wished to determine if a light source in the
center of the pool could draw the tadpoles away from edges,
or if the edge preference took precedence. In addition, any
bias toward one side or the other due to location of the

light sources or other unforeseen reasons, could be checked

this way.

Procedure

 

A 50-watt incandescent bulb was suspended 38 cm from the
water surface with a 6-cm diameter cardboard tube (29 cm
long) around it, acting to restrict the light into a 9-cm
spot in the center of the pool. The light was far enough
away that it did not affect water temperature below it.
After randomizing initial location by mixing the water, I
allowed the tadpoles 15-20 minutes to become accustomed to
the stimulus, then counted the number of tadpoles in each
sector. This procedure was repeated three times, each trial

on a different day.

Results
I compared the outside, middle and inside rings of

sectors, in pairwise combinations, using chi-square analysis.

44

As Table 11 shows, there were more tadpoles in the middle and
inner rings than expected in darkness; there were
significantly more in the middle than in the inner ring.
There were significantly fewer tadpoles in the outer ring

than expected.

Conclusions

 

Results of this experiment indicate that light appears
to be a stronger attractant than the edges of the pool. This
result is interesting when one considers that fluorescent
light does not appear to be a good stimulus. Perhaps this
difference has something to do with the quality of the light
after all, as suggested by the hot Spots and heat-as-stimuli
experiments. In any event, this experiment demonstrates that
at least one stimulus, incandescent light, is a greater
attractor for asocial aggregation than the innate tendency to
aggregate toward the pool edge. My subjective observations
tend to confirm this. The densest aggregations were right
around the 9—cm spot of light at the pool's center; density

dropped off markedly outside this small area.

45

Table 11. Results of test with bulb at
center of pool. A spot of light at pool
center was used to test tadpole preference
for outside, middle and inner rings of
sectors.

 

 

 

gomparison Preferred BE Significance
out vs. mid mid 8.7 2 5 .005
mid vs. in mid 32.7 p 5 .005
out vs. in in 18.5 p 5 .005

out - outside ring
mid - middle ring
in - inside ring

46

H. Stimulus Comparisons

Introduction

 

While the main purpose

 

of this study was to determine if

light, heat and food are used by tadpoles as environmental
stimuli to aggregate, the experimental design and method of
analysis were conducive to making comparisons between
stimuli, assessing their relative importance to tadpoles.
The following analyses were done with this in mind.
Procedure

Comparisons were possible due to the pairwise
combination of stimuli presented in the pool, with the

cardboard divider
following pairwise
versus food,

I counted the

isolating the stimulus on each side.

combinations

The

were looked at: hot spots

food versus light, and hot spots versus light.

number of tadpoles in all sectors in each of

the four trials for each of the above three combinations.

Results

I compared the four hot spot sectors with the four food

sectors in the light

chi-square analysis
pilot study.

preferred over hot

were preferred (Table 12).

and heat

In three of the trials,

versus food test, using

with the expectations calculated in the

food was significantly

spots, but in the fourth, the hot spots

47

Table 12. Food and hot spots compared
as stimuli. Sources of food were
compared with hot spots (heat and
light) in ability to attract tadpoles.
The Six most intensely-lit sectors
were compared with the hot-spot
sectors using chi-square analysis.

2

 

2355; '§_ Significance
1 6.9 2 5 .005*
2 16.3 p 5 .005*
3 12.6 B 5 .005*
4 40.0 p 5 .005**

* food preferred
** hot spots preferred

Table 13. Food and light compared as
stimuli. Sources of food were compared
with light in ability to attract
tadpoles. Food-containing sectors

were compared with the six most
intensely-lit sectors using chi-square
analysis.

Trial xi Significance

 

1 40.0 2 5.005*

2 15.0 p 5 .005*

3 35.0 ‘p 5 .005*

4 65.0 p 5 .005*
* food preferred

48

The four food—containing sectors were compared with all
twelve fluorescently-lit sectors since there was no
observable difference between the six with highest intensity
and the six with lowest intensity. In all four trials, food
was Significantly preferred over light (Table 13).

The four hot-spot sectors were compared with all twelve
fluorescently-lit sectors, for the same reason described
above. In this case, three of the four trials Showed no
significant deviation from the expectation (Table 14). In
one case, there were significantly more tadpoles on the

hot-spot side than expected.

Conclusions

 

Any conclusions drawn from these limited tests are
purely speculative, but are worth exploring. My results seem
to point to food as the most effective environmental stimulus
to. aggregate. Since it was significantly preferred over hot
spots in three of the four trials, and significantly
preferred over light in all four trials, it appears to be a
strong stimulus indeed. This makes intuitive sense, since a
tadpole must eat as much as possible in order to grow as
quickly as possible. This ensures it will get out of the
vernal pond before it dries up.

AS a second-best environmental stimulus, I would have to
choose hot spots. They were significantly preferred over

food in one trial (food was preferred in the other three),

49

Table 14. Light and hot spots compared
as stimuli. Light was compared with
hot Spots (heat and light) in ability
to attract tadpoles. The six most
intensely-lit sectors were compared
with the hot-spot sectors using
chi-square analysis.

Trial 53 Significance

1 1.19 2 5 .30*

2 023 E S 070*
3 13.40 E s 0005**
4 004 2 S 090*

* not significant
** hot spots preferred

50
and over light in one trial (no difference in the other
three). In any event, my results indicate that it is not
nearly as good as food as an attractor. Clearly, more tests
need to be done to determine if hot spots are superior to

light as a stimulus to aggregate.

ISOLATE/AGGREGATION EXPERIMENTS

Introduction

 

In their aggregation behavior, the tadpoles used in my
experiments were very unlike those described by other

authors. Most B. americanus tadpoles have been described as

 

forming tight-clustering social aggregations, D4-D5 on
Beiswenger's (1975) scale (Table 2). My tadpoles seemed to
form only loose, D2-D3 aggregations, and only in response to
external environmental stimuli. They did not appear to be
socially aggregating.

I wished to determine: (a) if individual B. americanus

 

tadpoles, isolated since an early stage, would join an
aggregation immediately, (b) if there was a latency period
(see literature review), or (c) if they refused to join the
aggregation. Such experiments have already been done with B.

americanus, but to my knowlege, all previous work has been

 

done with individuals taken from p0pulations exhibiting
tight-clustering social aggregation. I wished to determine
if the possibility exists for intraspecific variation in
joining an aggregation, both for the isolated individual and

for individuals previously exposed to other tadpoles.

51

52

Procedure

 

To test the tendency for tadpoles raised in isolation to
join an aggregation, I needed an apparatus which would
"force" some of my asocially-inclined tadpoles into a tight
cluster. I glued a 4-cm diameter screen tube (5-cm high)
constructed of 1-mm2 mesh nylon window screen to a thin
square of clear Plexiglas. This apparatus will hereafter be
known as the "corral". This was placed in the bottom of a
3.5-liter white plastic bucket with a 51-cm diameter (Figure
6).

I filled the bucket to a depth of 2.5 cm with
dechlorinated tap water, and positioned the bucket directly
below the fluorescent lamp fixture so that light was of
uniform intensity. Between 20 and 30 pool-raised tadpoles,
age class 25-30, were introduced in the corral and allowed to
acclimate for 15 minutes. They were crowded enough to
resemble a D5 aggregation. A single isolated tadpole, age
class 29-38, was released in the bucket at a random location
outside the corral and allowed five minutes to acclimate.
Presumably, any visual, chemical or lateral line cues from
the "aggregation" would pass through the screen tube. I
observed tadpole feces easily passing through the screen, due
to the turbulence caused by the tadpoles' tails beating.

The isolated tadpole's position was recorded every 30
seconds for five minutes (10 recordings). This was repeated
15 minutes later, and again 15 minutes after that for a total

of 30 position recordings in a 45—minute period. The

53
tadpole's age class was also noted. The water was changed
and a new group of pool-raised tadpoles was placed into the
corral before another isolate was tested. Each isolate was
returned to its plastic cup when finished. I tested 15
tadpoles in this manner.

As a control, five of the isolates were retested in the
same manner, except no pool-raised tadpoles were placed in
the corral. This was to see if there was a tendency for the
tadpoles to prefer one area or the other in the absence of
any stimuli. As a further control, I tested 4 pool-raised
tadpoles in the manner described above, first testing them
with other tadpoles in the corral, then without. This was
done to see if any difference existed between the behavioral
response of the isolates and that of tadpoles raised in a
group.

Of the 50 isolated individuals, all but 15 died en mass
in unusual circumstances one time when I added water to their
cups. I am not sure of the Specific cause of this, but it
severely limited the possibilities for study. Only five of
the fifteen remaining individuals were tested both with and
without tadpoles in the "forced school" corral. Due to time
constraints, the other ten were only tested with others in
the corral and four pool-raised tadpoles were tested both

with and without tadpoles in the corral.

Results
Results of this set of experiments are summarized in

Table 15. The location of a tadpole in the plastic bucket

54

"T

 

 

 

[E W

L_....._J

Figure 6. Sociality apparatus. At left is top view of
3.5-liter bucket with corral insert. At right is side
view of screen-tube corral.

 

 

 

OUTER ONE-HALF AREA

 

INNER ONE-HALF1AREA

 

Figure 7. Inner and outer one-half areas of bucket.

55
had been recorded during the experiment. A count was made of
the number of times a tadpole was in the inner one-half of
the area of the bucket and the number of times it was in the
outer half (Figure 7). The number of times the individual
was in the inner half is recorded in Table 15.

I found no significant difference between the number of
"in" scores (in the inner one half of the bucket) for the two
groups: a) with and b) without others in the corral, when
subjected to a Kruskal Wallis test (2.2 0.25). I obtained
the same results using the matched-pair t-test (p $_O.10).
This indicates that there does not seem to be any more or
less time spent socially, toward the center, when tadpoles
are present in the corral. Similar comparisons were made
with the pool-raised tadpoles, and there was no difference in
the number of "in" scores between these and the isolated
individuals (Kruskal Wallis test, 2 3 0.25).

There did, however, appear to be some individual
variation. Since only five isolates were tested both with
and without tadpoles in the corral, it is impossible to draw
any conclusions, but there were some marked changes in
behavior in two of the individuals. As noted in Table 15,
all five of these tadpoles had 10 or fewer "in" scores with
no tadpoles in the corral. While three individuals exhibited
the same behavior (10 or fewer "in" scores) after tadpole
introduction to the corral (matched-pair t-test, p Z 0.25),
two individuals, numbers 1 and 2 in Table 15, changed their
behavioral pattern so that each had 29 "in" scores. This is

a significant change (matched-pair tetest, p g 0.025).

56

Table 15. Results of isolate aggregation experiment.
Number of tadpoles on inside half-area of bucket
was counted, before and after addition of tadpoles
to corral.

Age No. In No. In Significantly
Tadpole Class Before After Different*

1 29 9 29 yes
2 36 7 29 yes
3 38 7 0 no
4 3a 0 e no
5 36 10 1 no
6** 28 1O 8 no
7** 3O 5 19 no
8** 3O 9 6 no
9** 30 7 12 no

* matched-pair t-test,‘p 5 .025
** pool-raised tadpoles

57

Pool-raised individuals were also tested for any change
in behavior after the addition of tadpoles to the corral, but
no change was noted (matched-pair tgtest,‘p.3 0.25).

I performed a correlation analysis on the five isolates
to determine if there was a correlation between the age class
of a tadpole and its relative behavior; its ability, desire,
etc., to switch from asocial to social behavior. I found no
correlation (3_ = 0.009). I then included the pool-raised
sample in my analysis, but again, there was no correlation

between age class and sociality (3’: 3.7 x 10'6).

Conclusions

 

I cannot draw any firm conclusions from this study due
to a lack of adequate number of test subjects. Some
interesting trends were noted, however, which should be
looked into further in future research. There was a great
deal of variability in isolate response to the addition of
tadpoles to the corral, even with only five isolates tested.
Two individuals had a marked change in behavior, suggesting
that some tadpoles are more inclined to aggregate socially
than others.

Remember that this population of tadpoles was very
non-social to begin with. I have no explanation for this, as
all other studies I have read describe very tightly-knit

social aggregation in B. americanus. Perhaps there is

 

geographic variability in social inclinations. My test
subjects came from two clutches which presumably would have

been deposited in the same general location, quite probably

58

in the same pond. In noting other tadpoles in that same
area, both myself and a colleague have commented on the lack
of distinct aggregations. Perhaps the genetic requirements
for social aggregation have been lost in the local B.

americanus p0pulation.

 

GROWTH STUDIES

Introduction

 

Previous studies have looked at differential growth and
the crowding effect in many species' tadpoles, including B.

americanus (see literature review). As noted previously,

 

three hypotheses have been suggested for the crowding effect:
(1) a physical or chemical substance, (2) an adreno-cortical
response to stress, and (3) exploitation competition.

This experiment attempted to fill in some of the gaps
created by previous research on the subject of growth
inhibition. It was confined to a series of observations on
tadpoles in the special tanks described in the general
methods section, the tadpoles kept in isolation, and a sample
of pool-raised individuals. I wished to compare the growth
rate of aggregating individuals with that of isolates. In
addition, the special tanks provided a unique opportunity to
observe tadpoles in a partially isolated state.

I wished to investigate Wassersug's (1973) claim that B.

americanus tadpoles may be less susceptible to the crowding

 

effect than other species' tadpoles. Although this study did
not involve any experimental manipulations, I also hoped to
provide some insight into the mechanism(s) involved in the

crowding effect in B. americanus.

 

59

60

 

Procedure

Tadpoles were assessed for differential growth in two
ways. The total length (snout to end of tail) was measured
after transferring a tadpole to a watch glass using a
large-bore 20-cc syringe. This minimized the effects of
handling, an important factor if psychological stress is
involved in the crowding effect. When completely straight
(i.e. no flex in the tail), tadpoles were quickly measured to
within one millimeter with a small plastic ruler and assessed
for Gosner age class. While age class is a somewhat
subjective value, I estimate that I can accurately determine
the age class of a tadpole to within one stage out of the 46
total (a 2.2% error). Tadpoles were immediately returned to
their respective containers.

Individuals which had been isolated since Gosner stage 9
(late cleavage) were introduced to the special tanks at three
different times. Three individuals were introduced at stage
12 (late gastrula), five were introduced at stage 17 (tail
bud) and four were introduced at stage 24 (opercular flaps
nearly complete). Between 11 and 19 tadpoles of mixed
parentage had been raised together in the small-area side of
the tank since stage 9. I call the introduced isolates in
the large-area side of the tank "semiisolates" because they
are no longer in complete isolation. I term the group in the
small-area side the "cohorts". Originally, 24 such special
tanks were set up, eight for each stage of introduction, but
in 12 of them, the grouped tadpoles in the small-area Side

were able to get under the screen partition which separated

61
them from the lone individual. I was subsequently unable to
determine which one was the original isolate. Data from
these were discarded.

Total length and age class were assessed for all
semiisolates and cohorts on three occasions, 16 days after
the experiment began (6-5-84), 10 days later (6-15-84),and
12 days after that (6-27-84). On the latter two occasions
(6-15 and 6-27), I also assessed the total length and age
class of 19 isolate tadpoles and a random sample of 20

pool-raised tadpoles, using the methods described earlier.

Results

The mean length and mean age class of the isolates,
semiisolates, cohorts, and the pool-raised sample were
compared with each other, two groups at a time, using
Student's Estest. I treated the Six comparisons of mean age
class independently of comparisons of mean length. I also
analyzed the data from June 15th independently of those from
June 27th.

With B_different treatment groups, by convention, g - 1
independent pairwise comparisons are acceptable, each with a
significance level of 3.: .05. In this case, 2_- 1 = 4 - 1 =
3 comparisons. At the .05 level, the experiment-wide
protection against type I error is 1 - (.95)-13»-1 or 1 -
(.95)3 = 0.143 in this experiment.

Because I did more than three comparisons, two problems

arise: (1) inflated experiment-wide 3((type I error), and (2)

62

lack of independence. The problem of increased type I error
can be remedied by adjusting the significance level of
individual comparisons such that the experiment-wide
probability of type I error is less than or equal to that
with 2 - 1 independent comparisons (= 0.143 in this
experiment). Since I did six comparisons, they could not all
be considered independent (remember, I actually looked at
four groups of six comparisons each: June 15 age class, June
15 length, June 27 age class, and June 27 length).

The experiment-wide probability of type I error is the
union of the probabilities of type I error for the individual
comparisons (corollary 1 to the addition rule in Rohtogi,
1976). Furthermore, the probability of the union of any set
of events is less than or equal to the sum of the individual
probabilities of the separate events. Therefore, regardless
of the dependence relationships among tests, the
experiment-wide probability of type I error is always less
than or equal to the sum of the individual probabilities of
type I error. If an individual comparison's level of
significance is selected such that the sum across all six
comparisons equals 0.143, one can be sure the experiment-wide
protection against type I error is as good or better than
that which would occur with three independent comparisons at
the 0.05 level.

This study employed six pairwise comparisons, all with
identical levels of significance. Since the six comparisons'
combined levels of significance must be less than or equal

to 0.143, each must have a level of significance less than

63
or equal to 0.024 (0.143/6 = 0.024). I will adopt the
convention that a comparison is significant if a_£ 0.024, or
not significant if’ a. > 0.024. In addition, because of
unequal variance, certain comparisons are done with Cochran's
Bf—test (Snedecor and Cochran, 1967) instead of Student's
B—test.

From the June 15th data (Table 16), both the isolates
and the semiisolates appear both older and larger than either
the pool—raised sample or the cohorts. There were no
significant differences in the mean age or size between the
isolates and semiisolates; the cohorts and pool-raised
tadpoles also did not differ significantly in size or age
class. Note that while mean age class and length are treated
independently, they both elicit the same results.

By June 27th, some changes in these relationships were
observed (Table 17). Isolates and semiisolates remained
indistinguishable in mean size or age class, as did cohorts
and the pool-raised . sample, but I also observed no
significant difference in length between semiisolates and the
pool sample, or the semiisolates and the cohorts. At the
same time, the isolates were still significantly larger than
the cohorts. The disparity in developmental state remained,
as the semiisolates and isolates were both at a significantly
more advanced age class than the cohorts or pool sample.
Comparing Tables 16 and 17 for isolate length, one notes that
length decreases over time. This is due to the metamorphosis
of some of the isolates between June 15 and June 27,

resulting in a smaller sample size for length studies.

64

Table 16. Results of growth study, June 15. Comparisons
were made for the four groups independently for degree

of deve10pment and length, using pairwise Student's ietests
or Cochran's Ef-tests. Means for each class are in
parentheses. Significance was restricted to the .024 level.
A line connecting two or more groups implies no significant
difference exists between them.

 

pool
Q. isolates semiisolates cohorts sample
Age class. _(33,45) (32.67) £28.20) (27.35)
pool
. semiisolates isolates sample cohorts
Length- (19.67) (19.50) (14.25) (13.60)

 

 

Table 17. Results of growth study, June 27. Comparisons
were made for the four groups independently for degree

of develOpment and length, using pairwise Student's Estests
or Cochran's i'-tests. Means for each class are in
parentheses. Significance was restricted to the .024 level.
A line connecting two or more groups implies no significant
difference exists between them.

 

 

pool

, isolates semiisolates cohorts sample

Age Class° (39.43) (37.08) (31.38) (30.90)
pool

Length° isolates semiisolates cohorts sample

(21.65) (19.22) (17.53) (17.14)

 

65

Recall that the semiisolates were introduced to the
special tanks at three distinct Gosner stages, 12, 17, and
24. Correlation analysis of the data from June 15 revealed
that no significant correlation exists between the stage at
which a semiisolate is introduced to the test tank and the
difference between the mean length of the semiisolate and its
cohorts (5 = —0.15). There is also no correlation between
the stage of semiisolate introduction and the difference
between the mean age class of the cohorts and their
semiisolate (5 = 0.12). This suggests that growth inhibition
may occur at any time during early development; it does not
require that the tadpoles be in association from the start.

As an interesting aside, I calculate that there is a
significant positive correlation (£_= 0.74) between the size
difference and age class difference of the semiisolates and
cohorts. This correlation coefficient increases to 0.94 when
only one data point is removed. This confirms that at least
during early develOpment, length corresponds fairly well with

degree of develOpment.

Conclusions

 

The results of the comparisons of the four groups
suggest a few major points. First, isolates and semiisolates
seem to grow at the same rate. They remained about the same
size and age class throughout the experiment. The same can
be said for the pool-raised sample and the cohorts. Notice
that, while food was not a limiting factor, both the isolates

and semiisolates had more space than the cohorts or pool

66

sample. Isolates had 63.6 cm2 each and semiisolates 166.3
cm2, while pool-raised tadpoles had 9.4 cm2 each and cohorts
4.0 to 6.9 cm2 each. This rather large disparity seems to
suggest that the adreno-cortical stress response may be
acting in this situation.

A second point to note is that age-class differences
remained unchanged between June 15th and June 27th, but
length differences broke down. There is most likely a
maximal size the tadpoles reach, at about stage 30—32. While
isolates and semiisolates continued todevelop in advance of
the cohorts and pool sample, they were no longer at a size
advantage. Length fails to be an accurate indicator of
degree of develOpment at that time.

The third point is that the crowding effect will occur
regardless of the time during develOpment in which the test
tank is established, at least early on (stages 12—24). I got
distinct age-class disparity when I introduced a semiisolate
to the test tank at all three stages tested, 12, 17, and 24.

From these observations, it is clear that the crowding

effect does occur, quite readily, in B. americanus. It is

 

difficult to refute Wassersug's (1973) claim that B.

americanus tadpoles are less susceptible to growth inhibition

 

than other species because of a lack of data to compare my
results to. It seems unlikely, however, that other species
could have a more pronounced effect than I observed,
especially when one considers the lack of reports in the

literature on such an obvious effect.

67

It is clear that exploitation competition is not
necessary to cause a growth differential. Since tadpoles had
unlimited access to food, this hypothesis seems unlikely.
This is not to say, however, that it is impossible to induce
the crowding effect via exploitation competition.

Semi-isolated individuals were exposed to the same
environment as the cohorts, yet were able to develop
significantly faster. Indeed, they developed as quickly as
isolated individuals. ' Any chemical or physical substance
responsible for the crowding effect could easily have passed
through the screen partition. The probability that, by
random chance, the semiisolate (as opposed to one of the
cohorts) is always the fast-growing individual which secretes
this substance is approximately 3.6 x 10-15. In eleven of
the twelve experimental test tanks, the semiisolate was
indeed the fastest-growing tadpole. It is extremely unlikely
that this is due to random chance (p = 5.7 x 10-14). If an
alleged chemo/physical substance worked over a 3331 short
range (say, less than 5 mm), such results would be possible,
otherwise, my results suggest that the presence of a
chemo/physical substance in unlikely.

As pure speculation, I suggest that the initial
disparity in growth rate could have been caused by a stress
reaction. There must be a mechanism by which certain
tadpoles gain an initial growth rate advantage. The cohorts,
being much more crowded than the semiisolates, would be under

such stress. In the pool, I observed various tadpoles in a

68

wide range of deve10pmental states at one time. Obviously,
not all individuals were developing at the same rate, yet all
were exposed to the same conditions. This implies that the
stress reaction is not the only factor which could effect
growth rate, assuming that all individuals are equally
susceptible to the stress reaction (which may not be true).
I‘ suggest that secondarily, exploitation competition may be
employed by some tadpoles to gain a growth advantage,
possibly supplemented by a short-ranging chemo/physical
substance.

Another possible explanation for this phenomenon
pertains to energy utilization. Tadpoles in crowded
conditions may not be under stress per se, but because they
are in close proximity to each other, they are constantly
bumping into each other and continually moving. When
disturbed, tadpoles tend to swim away from the disturbance.
This keeps crowded tadpoles in nearly constant motion, while
the uncrowded tadpoles remain quiescent. It is a well
documented fact that anuran growth rate is positively
correlated with net energy gain. The semiisolates and
isolates, because of their relative quiecence, may have
Simply had a larger net energy gain, and hence, grew at an
accelerated rate relative to that of the cohorts or pool

sample.

GENERAL DISCUSSION

This study looked at three aspects of aggregational

behavior in B. americanus tadpoles. In the first experiment,

 

I investigated three environmental factors which may lead to

asocial aggregation. I observed that B. americanus tadpoles

 

have a tendency to aggregate in the absence of these three
stimuli (light, heat and food), particularly toward the
shallow edges of the pool. I also observed that some stimuli
are better than others in attracting tadpoles. Food is a
very good asocial aggregational stimulus, as is incandescent
light (light and heat). Fluorescent light, however, does not
appear to be a good attractor. Tadpoles do not seem to be
able to discriminate well between various intensities of
florescent light. Heat alone is a good "broad area"
attractor.

Shallow water seems to be an important common
denominator for these stimuli of attraction. The warmest
water in a pond duing the daytime is toward the edge, in the
shallows, where the sun's rays can heat it quickly. Food is
also very abundant in shallow water, where light intensity is
strongest (personal observations). All this suggests that a
tadpole programmed to seek cues corresponding to shallow,

food—rich water would be at a great advantage. Natural

69

70

selection would act strongly on such tadpoles. My studies
indicate that tadpoles are able to find food in and of itself
in a small pool, but in a large pond, where food might be
more widely-scattered, other cues such as a temperature
gradient or water depth could become very important in
bringing a tadpole close enough to a food source that the
tadpole could locate the food by chemo-olfactory cues, or by
whatever other means it uses.

Optimal-foraging theory predicts that individuals will
prefer to live in groups if food is spatially arranged in
high quality patches (Pulliam and Caraco, 1984). It is
easier to locate these food patches as a group (by whatever
means). A group—living individual has a higher net energy
gain than if it remains on its own under such circumstances.
Natural selection would therefore be SXpected to favor
individuals which exhibited social tendencies in this
situation. I believe this is why we see many anuran species'
tadpoles forming social aggregations.

Not all species' tadpoles are social aggregators, and
the possiblity exists for variation in aggregating ability in
some species, as suggested by my second experiment. While
only observing a limited number of replications, I did
observe a wide range of social aggregations in tadpoles
tested, from both isolated and pool-raised samples. While I
realize that many more replicates need to be done before any
substantial claims can be made, I hypothesize that a single

Species' tadpoles may exhibit variation in social behavior.

71
Recall that my tadpoles came from ponds where tadpoles did
not form the tight social clusters described by other
authors. While it is possible that food does not occur in
high quality patches in these ponds, thereby nullifying an
assumption necessary to explain social optimal foraging
groups, I see no reason why this should be the case.
Something, however, is different here. Perhaps some genetic

anomaly has gotten into the local B. americanus population.

 

Since American toads may return to the same area, possibly
even the same pond to breed each year (depending on local
availability of bodies of water), and since their offspring
are not likely to travel very far away, it is quite possible
for a genetic condition such as this to develop on a local
basis. Lack of sociality may be a genetic drift phenomenon,
due to chance.

On average, the growth rate of tadpoles in an
aggregation is reduced over that of an isolated or
semi-isolated individual (as defined by my experiments). An
important goal for a tadpole is to get to metamorphic size
and get out of the water as quickly as possible. Thus, there
seems to be a conflict of interests here. On the one hand,
joining a group causes a reduction in growth rate, but on the
other hand, it may increase the individual's feeding
efficiency, assuming Optimal—foraging theory is correctly
applied here. It can be assumed that aggregation, if a
genetic phenomenon, would not occur if there wasn't an

overall advantage to it. If one considers the other

72

advantages to being in an aggregation; (1) increased heat due
to formation of a large black mat in the water, (2) reduced
predation due to the conspicuous nature of the black mass of
tadpoles in association with distastefulness, and (3) the
possibility for increased fitness due to kin selection in the
group, it is quite conceivable that the advantages outweigh
the disadvantages.

In any event, the crowding effect does occur in B.

americanus, apparently regardless of whether the aggregation

 

is social or asocial. Many experimenters have shown that
some tadpoles grow quite normally while the growth rate of
the others is inhibited (Richards, 1958; Licht, 1967;
Steinwascher, 1978). This is an inhibition phenomenon, not
an accelerated growth phenomenon as some have hypothesized
(Rose, 1960; Shvarts and Pystolova, 1970b).

My experiments seem to support the idea that the
crowding effect is either an adreno-cortical response to
stress caused by physical contact with other tadpoles, or the
result of a greater net energy gain due to quiscence in
less-crowded conditions. Those individuals which are not
being severely affected may be producing some physical or
chemical substance which further reduces the growth rate of
their fellow tadpoles while not affecting their own. Having
a size advantage would allow such tadpoles increased access
to food via exploitation competition, implying they could
further increase their growth advantage. In a positive
feedback loop such as that outlined in Figure 1, this system

could go on until the larger tadpoles metamorphosed, leaving

73
the pond and all its resources to a few individuals.

This study attemped to fill in some of the voids left by
previous research on tadpole aggregative behavior. While
attempting to answer some questions, it raised others which
must be addressed by future researchers. More work needs to
be done on exploring population differences in aggregation.
Observations and tests need to be done comparing the local

East Lansing B. americanus tadpoles with those from

 

surrounding regions, using standardized tests to determine if
there are differences in degree of sociality between
p0pulations.

To my knowledge, no one before has suggested that
aggregating tadpoles are optimally foraging. This is a
fascinating idea which undoubtedly should be researched
further. Quantitative measurements of the food base of
vernal ponds in different areas should be taken to confirm
that food exists in high quality patches.

It may be that a single p0pulation can exhibit social or
asocial aggregation depending on prevailing conditions. Such
a hypothesis could be explored by varying conditions under
which groups of tadpoles from a single clutch are raised.

Finally, more work needs to be done on the problem of
differential growth and the crowding effect. Specifically,
the question of whether a chemo-physical substance or a
stress reaction being primarily responsible for this

phenomenon needs to be addressed in more detail. The notion

74

that it could be primarily a stress reaction supplemented by

a chemo-physical substance Should be explored further.

BIBLIOGRAPHY

BIBLIOGRAPHY

Allee, W. 1931. Animal aggregations. University of Chicago
Press, Chicago, IL. pg 74.

Beiswenger, R. 1975. Structure and function in aggregations
of tadpoles of the American toad, Bufo americanus.
Herpetologica 31:222-233. 1

 

 

Beiswenger, R. 1977. D181 patterns of aggregative behavior in
tadpoles ‘of Bufo americanus, in relation to light and
temperature. Ecology 58:98-108.

Beiswenger, R., and F. Test. 1966. Effects of environmental
temperature on movements of tadpoles of the American
toad, Bufo terrestris americanus. Papers Michigan Acad.

Sci. 51:127-141}

Bilski, F. 1921. Uber den einfluss deS lebensraumes auf des
wachstum der kaulquappen. Pflugers Archiv. 188:254-274.

Blaustein, A. and R. O'Hara. 1981. Genetic control for
sibling reconition? Nature 290:246-248.

Blaustein, A. and R. O'Hara. 1982. Kin recognition cues in
Rana cascadae tadpoles. Beh. Neural Biol. 36:77-87.

Bragg, A. 1954. Aggregational behavior and feeding reactions
in tadpoles of the Savannah spadefoot. Herpetologica
10:97-102.

Bragg, A. 1965. Gnomes of the night. University of
Pennsylvania Press, Philadelphia, PA. pgs 70—81.

Bragg, A. 1968. The formation of feeding schools in tadpoles
of spadefoots. Wasmann J. Biology 26:11—16.

Brattstrom, B. 1962. Thermal control of aggregation behavior
in tadpoles. Herpetologica 18:38-46.

Brenden, F., A. Lum and R. Wassersug. 1982. Body size and
orientation in aggregations of toad tadpoles Bufo
woodhousei. Copeia 1982:672-680.

 

75

76

Brockelman, W. 1969. An analysis of density effects and
predation in Bufo americanus tadpoles. Ecology
50:632-644.

 

Carpenter, C. 1953. Aggregation behavior of tadpoles of Rana
p_. pretiosa. Herpetologica 9:77-78.

deVlaming, V. and R. Bury. 1970. Thermal selection in
tadpoles of the tailed-frog, Ascaphus truei. J. Herp.

 

Franz, V. 1913. Die phototaktischen erscheinungen im
tierreiche und ihre rolle im freileben der tiere.
Zoologische Jahrb. Abt. fur Physiologic 33:259-286.

Foster, M. and R. McDiarmid. 1982. Study of aggregative
behavior of Rhin0phrynus dorsalis tadpoles: design and
analysis. HerpetEIOgica 38:395-404.

 

Gosner, K. 1960. A simplified table for staging anuran
embryos and larvae with notes on identification.
Herpetologica 16:183-190.

Gromko, M., F. Mason and S. Gill. 1973. Analysis of the
crowding effect in Rana pipiens tadpoles. J. Exper.
Zoology 186:63-72.

 

John, K. and D. Fenster. 1975. The effects of partitions on
the growth rates of crowded Rana pipiens tadpoles. Amer.
Midland Naturalist 93:123-130.

 

Katz, L., M. Potel, R. Wassersug. 1981. Structure and
mechanisms of schooling in tadpoles of the clawed frog,
XenOpus laevis. Animal Behavior 29:20-33.

 

Licht, L. 1967. Growth inhibition in crowded tadpoles:
intraspecific and interspecific effects. Ecology
48:736-745.

Lum, A., R. Wassersug, M. Potel and S. Lerner. 1982.
Schooling behavior in tadpoles: a potential indicator of
ototoxicity. Phar. Biochm. and Beh. 17:363-366.

Mullally, D. 1953. Observations on the ecology of the toad
Bufo canorus. Copeia 3:182-183.

 

Noble, G. 1931. The value of life history data in the study
of the evolution of the Amphibia. Ann. New York Acad.
Sci. 30:31-128. -

Noble, G. 1954. The biology of the amphibia. Dover
Publications, Inc. New York, NY. pgs 344-345.

77

O'Hara, R. and A. Blaustein. 1981. An investigation of
sibling recognition in Rana cascadae tadpoles. Animal
Behavior 29:1121-1126.

 

Pulliam, H. and T. Caraco. 1984. Living in groups: is there
an Optimal group size?, in Behavioral ecology: an
evolutionary approach. J. Krebs and N. Davies, eds.
Sinauer Associates, Inc. Sunderland, MA. pgs 122-147.

Richards, C. 1958. The inhibition of growth in crowded Rana
pipiens tadpoles. Physiol. 2001. 31:138-151.

Richards, C. 1962. The control of tadpole growth by
algal-like cells. Physiol. Zool. 35:285-296.

Risser, J. 1914. Olfactory reactions in amphibians. J. Exptl.

Rohtogi, U. 1976. An introduction to probability theory and
mathematical statistics. John Wiley and Sons, New York,
NY. pg 26.

Rose, S. 1960. A feedback mechanism of growth control in
tadpoles. Ecology 41:188-199.

Rugh, R. 1962. Experimental embryology; techniques and
proceedures. Burgess Publishing Co. pgs 49-109.

Shaw, E. 1960. The deve10pment of schooling in fishes.
Physiol. 2001. 33:79-86.

Shvarts, S. and 0. Pyastolova. 1970a. Regulators of growth
and development of amphibian larvae I. Specificity of
effects. Soviet Journal of Ecology (Ekologiya) 1:58-62.

Shvarts, S. and O. Pyastolova. 1970b. Regulators of growth
and deve10pment of amphibian larvae II. Diversity of
effects. Soviet Journal of Ecology (Ekologiya) 1:122-134.

Snedecor,G. and W. Cochran. 1967. Statistical methods. Iowa
State University Press, Ames, Iowa. pgs 114-118.

Stuart, L. 1961. Some observations on the natural history of
tadpoles of Rhinophgynus dorsalis. Herpetologica
17:73-79-

 

Tevis, L., Jr. 1966. Unsuccessful breeding by desert toads
(Bufo puctatus) at the limit of their ecological
foIerance. Ecology 47:766-775.

Voris, H. and J. Bacon. 1966. Differential predation on
tadpoles. 00peia 1966:594-598.

Waldman, B. 1980. Sibling recognition in toad tadpoles: the
role of experience. Amer. Zoologist 20:854.

78

Waldman, B. 1981. Sibling recognition in toad tadpoles: the
role of experience. Zeitschrift fuer Tierpsychologie
56:341-358.

Waldman, B. 1982. Sibling association among schooling
tadpoles: field evidence and implications. Animal
Behavior 30:700-713.

Waldman, B. and K. Adler. 1979. Toad tadpoles associate
preferentially with siblings. Nature 282:611-613.

Wassersug, R. 1973. Aspects of social behavior in anuran
larvae, in Evolutionary biology of the anurans. J. Vial,
ed. University of Missouri Press, Saint Louis, MO. pgs
275-297.

Wassersug, R. and C. Hessler. 1971. Tadpole behavior:
aggregation in larval Xenopus. laevis. Animal Behavior
19:386-389-

Wilbur, H. 1977. Density-dependant aspects of growth and
metamorphosis in Bufo americanus. Ecology 58:196-200.

 

Wilbur, H. and J. Collins. 1973. Ecological aspects of
amphibian metamorphosis. Science 182:1305-1314.

General Reference

 

Schefler, W. 1979. Statistics for the biological sciences.
Addison-Wesley Publishing Company. Reading, MA.