Comparative effects of food and water deprivation on movement patterns in the pigeon (Columba Livia)

41

Behavroura/ Processes, 20 (1989) 41-48

Elsevrer COMPARATIVE EFFECTS OF FOOD AND WATER DEPRIVATION ON MOVEMENT PA’ITERNS IN THE PIGEON (COLUMBA LIV’U)

ROBERT W. ALLAN’ and T JAMES MATTHEWS2 i Department Department 10003, USA

of Psychology, Fairlegh Dxkinson Umversdy, Ma&on, New Jersey 07940, USA of Psychology, New York Unrversrty, 6 Washington Place, New York, New York ( Accepted

I August

1989 )

ABSTRACT Allan, R W. and Matthews, T.J , 1989 Comparattve effects of food and water deprtvatton on movement patterns in the ptgeon (Columba Livra). Behave Process. ,20: 4 l-48 Groups of naive pigeons were either food or water deprived and exposed to response-independent ftxed time schedules of reinforcement Reinforcers were always approprtate to the deprtvatton state. After 15 Phase One sesstons under the original deprtvatton state, an addmonal 15 Phase Two sessions were run under the alternate deprrvatton regimen The Phase One results mdrcated that the frequency of movement over time, as measured by floorboard panels, had a characteristic distribution dependmg on deprrvatton state. After state-swatch in Phase Two, both groups exhibited changes in frequency dtstrrbutton over ttme, but the new distributions were dtfferent from those produced m SubJects without any deprivation htstory. There appeared to be sustained effects on movement patterns due to prior exposure to alternate deprtvatron and periodic feeding regimens, results which confirm earlier experimental work. The findings are discussed m hght of research on the effect of prior response-mdependent schedule trammg, the matchmg law, and the effects of contextual condtttonmg. It has frequently been shown that ptgeons ~111engage m stereotyped movement patterns when exposed to response-independent

fixed-time (IT) schedules of reinforcement

(Skmner, 1948;

Staddon& Simmelhag, 1971, Staddon, 1977; Reberg, buns, Mann & Etzenga, 1978, K.tlleen, 1975; Allan & Matthews, 1982). The behavior so produced seems to be a function simply of the deprivanon state m conlunctton with the experimental

environment

(including feeding schedules).

The

chamber is usually limited in size, with a food or water dehvery system on one wall, and with httle else in the enclosure.

In a large experimental chamber, Reberg et al. (1978) observed distmct

behavroral effects when ptgeons were food or water deprived and exposed to the response-independent reinforcer deliveries.

However, after SubJectS were switched to an alternate deprtvatton

regimen, sustained carry-over effects were observed

The carry-over effects mamfested themsel-

ves in a slow drifting of the form and distribution of activity produced by the original motivational state toward those assoctated with the original presentation

of the new deprtvatton state.

Unlike the Reberg et al (1978) study, which also compared food and water-induced the present study employed automated data recording

This meausurement

a more conststent defnution data.

of response classes and allows the recruitment

0376-6357/89/$03

Elsevrer Scrence Publishers BV (Bromedrcal Drvtston)

50

@ 1989

behavtor,

technique provides of a larger body of

42

Figure 1. Schematic diagram of the experimental to measure locomotion frequency and location.

chamber and the segmented floorboard

used

METHOD

Subjects Erght narve, Whrte Carneau pigeons served as subjects. In their home cages, food-deprived and water-deprived

subjects had free access to water and food respectively

Apparatus The experimental

chamber is depicted in figure 1. Its inner drmensrons were 73 cm long, 38

cm wide, and 41 cm high. The front panel was one of two BRS/LVE Prgeon Intelligence

Panels,

the first delivered food reinforcers (mixed grain), the second panel was modified to deliver water (.75 ml). The response key centered on the front panels was left uncovered but was not used The floor of the test chamber was divided into nine panels, each measurmg 16.2 X 12.1 X 0.3 cm (see figure 1). Each corner was supported by a single Honeywell V3L131-D8

rmcroswitch.

Two carriage screws held the panels in place, but m no way restricted their movement against the microswitches

If the prgeon stepped anywhere on the panel, one of the parallel wired mrcros-

watches was acttvated.

On and off trmes were recorded by a Plessy Micro 1 computer for later

analysrs. A video camera positioned 1 m above the clear acrylic cedmg of the experimental chamber allowed visual monitoring of subjects’ actrvrty

43

Procedure The subjects were divided into two groups of four buds, each run m two experimental

Group Food-Water

(FW; subjects 20,21,22,23)

Phases

were food deprived during the first Phase and

switched to water deprivatron during the second Phase. Group Water-Food

(WF; SubJects 106,

108, 109, 110) were water deprived during Phase One and switched to food deprtvatron during Phase Two. There was a five day recovery period between deprtvatton regimens durmg whrch all subjects had restrrcted daily access to food and water in the home cage Both expertmental Phases were run for fifteen days In the food deprivatton condrtrons, the experimental sessrons were run on consecutive days and conststed of 40 trials with 30 s inter-remforcer mtervals (IRI) During water deprrvatron, however, the bards were run every other day and sessions consisted of 20 trials with 30 s IRIS Thts was necessary because pilot work showed that with more trrals per session, or wrth dally sessions, the pigeons would not drink throughout the session. RESULTS Figures 2 and 3 depict frequencies of floorboard actrvrty (responses per mmute), plotted in 3Each data pomt is an average over all trrals over the last 5 sesstons

s bms over elapsed trial time

of each phase, wtth filled circles depicting actrvrty on panels l-3 (front), and open circles representing movement on panels 4-9 (back). For group FW, during the food-deprrvatton

Phase, three out of four subjects show a front panel

dominance

The exception, subject 22, moved quickly after the reinforcer to panels 4-6 and paced slowly between front and back panels along the srde wall After the reinforcer, the three remaining SubJects immedtately moved to the back of the chamber for a brief period, followed by a sustained, htgher rate of pacing along the front wall until the next food delivery After the shift from food to water deprivatton, the peaked nature of front and back panel actrvtty seemed to disappear

The very high rates of front panel acttvrty for subjects 20 and 21

remained dominant under water deprivation.

Subject 22 continued to pace along the side wall

and subject 23’s response rate dropped to very low levels after swrtching to water deprrvatron. Group WF SubJects in Phase One (given this particular level of water deprrvatron) generally responded at lower rates than drd the birds m Group FW, Phase One (food depravation). In three out of four cases, the dominant response pattern, in the water deprtvatron Phase, was pacing along the side wall at a very low rate. The exception, subJect 106, spent most of the interval moving at a fatrly rapid rate on the back panels

Only subJect 109 produced a peaked distribution

under

water deprivation conditions. When shafted to food deprivation

in Phase Two, all subjects showed an enhancement

of ac-

tivity, front and back. There was also an increase in the peaked nature of the movement distrrbutions with the highest rate of acttvrty occurring m the first half of the IRI. The effect of prror water deprrvatronon

subsequent food activity rates is an enhancement

of relative middle and back panel

Food -----+

Water

120 100 80

20

60 40 20 ‘;Q,

0

C

60

E b

40

21

SL u$

20

E ’

0

$

60

u

22

40 20 0 60

25

40 20 0

0

6

12

16

24

0

6

12

16

24

3

Elapsed trial time

Figure2. Individual average frequencies of panel closures per minute for group FW, plotted over elapsed trial time. The plots on the IetI represent Phase One (food deprived) performances. The plots on the right depict frequencies obtained alter the switch to water deprivation conditions. Bird numbers are indicated at the right of each set of plots.

45

Food

Water ------w

80

i

106

td

108

6

12

18

24

'0

6

12

18

24

Elapsed trial time

Figure3. Individual average frequenctes of panel closures per minute for group WF, plotted over elapsed trial time. The plots on the left represent Phase One (water deprived) performances. The plots on the right depict frequencies obtained after the switch to food deprivation conditions. Bird numbers are indicated at the right of each set of plots.

movement,

wtth one case (subject

106) showmg higher mtddle

and back panel rates under food

deprrvatton Overall, food and water deprivation one case, when animals animals were shifted relative

amount

food deprived

condttrons

were shifted

produce

from food to water deprrvatron,

of back panel activity produced subjects.

different

formances

produced

preceded

by treatments

cases (#106-food,

also confirmed

to the back of the chamber

water deprivation

as the rates eventually condrttons

represented observed

sequence

Followmg

Followmg

back panel ac-

Phase Two per-

m Phase Two

of behavior

It is because

during the IRI

food reinforcers

a wa!er reinforcer,

but were less likely to commence

condmons

the

seven out of eight buds

briefly, after which pacing acttvrty was mutated

toward the front wall of the chamber the front of the chamber

#20-water)

that the characterrstrc

differed under food and water depravation

rear of the chamber

In general,

was higher than in

in which the relative rates of front and back panel actrvtty were

and in the same duectton

Video momtormg

In contrast,

when

front panel acttvtty rates as high or higher than back panel

Each of the exceptronal

hrghly disparate

subjects

In seven out of eight buds run under water deprtvatron,

run under food deprtvatton

In all but

actrvtty increased,

acttvtty levels decreased

by water deprived

tivity was as high or hrgher than rates under food deprtvatron movement

levels of acttvtty

from water to food deprrvatton,

animals moved

as the animal moved

subjects also retreated

to the

pacing, and did not usually move toward

of the latter tendency

also resultmg

m less disparate

observatrons

provide,

that overall activity was lower m the relative rates of front and back panel

activity DISCUSSION The present

experimental

tween the movement compartson

patterns

of movement

been changed,

associated

frequencies

and (3) confumatton

have been altered

patterns

by the continued

Although

in the present

regtmen

changes m the experimental

acttvtty (Ktlleen,

regimen

has

et al. (1978),

patterns

of the

phases for both groups FW and WF)

to a prior deprtvation

a change in deprivation

be-

(2) a clear wrthm-group

once the deprtvatton

It seems that the typical adaptive

exposure

dtstmctton

effects noted by Reberg

many sources of the distinct movement

in the level of schedule-induced acttvtty, as observed

carry-over

at some level, throughout

There are, perhaps, water deprivation

elicited

those ehcrted m the first experimental

remam predommant,

of a reliable

with food and water deprrvatron, and patterning

of quahtatrve

at least m the form of locomotton prgeons (presumably

(1) a descrtptton

patterns

These

assoctated

level could be expected

patterns

environment with food and

to produced

a change

1975) a shift in the shape of the drstrtbutron

study, would not be expected

of

to result from a change m deprtva-

tion level alone An alternative duced by pertodtc these reinforcers over patches

approach

recognizes

the consptcuous

food and water presentations Food acqutsttron

of high food denstty

m the natural while water

parallel between

the behavior patterns

and the natural foraging patterns envn-onment

requires

IS assocrated

very httle

locomotor

associated

mwith

with raped pacing behavtor

The

47

schedule-Induced

behavior

tated by species-typical

produced

patterns

effects shown here and elsewhere cement

history in determining

The observation terns of movement to a new deprivation (alternation

(Reberg

of remfor-

behavior IS not

history of an organism

of these effects that is strlkmg

Phase One condutons

effect of the behavioral

1s to confhct with a new, and perhaps

If the pat-

are representative

perseveration

of some

after the switch

more appropriate

This fmdmg also seems to suggest, quite powerfully, from front to back panels and relative frequencies

die-

The carry-over

the importance

effects due to the past experimental

then one potential

was not entirely

mechanisms

of schedule-induced

It IS the relative permanence

regimen,

however,

et al, 1978), demonstrate

elicited during the respective

sort of foragmg pattern, mg responses

experiment, (deprivational)

the form and distribution

of carry-over

unusual nor unexpected

m the present

tred to motivational

set of forag-

that the locomotion

thereof)

patterns

are related to the depriva-

tion state of the organism It is clear from the present being recorded,

results that, (1) there was modiftcatton

(2) there was development

of new alternating

front and back panels which did not resemble (3) some of the response donunant

patterns

any behavior

of responses

then, provides response

for both groups changed

quantitative

dimension

-- frequency

In light of the present havioral probabilities deprivation

high frequency),

state shift

The present

et al (1978) while utthzing

experiment,

perhaps

descrtpttons

established

responses

dtswork,

a single and simple

and the response

Both response

of potential

patterns

and deprivation

alteration

should be expanded

of present

be-

to include the prior

ekczted (not only emitted)

effects need to be considered

analysis by Pear (1988)

produced

during deprivation

the generalized conditions

matchmg

However,

law was applied

The results of the present

port Pear’s claim that the value of the bias term is affected

changes

(e g ,

and (4) The temporal

under

when examm-

responding

In a recent

delivery

animals,

were ehmmated

of panel closure

history of the orgamsm

mg current

between

seen m Phase One deprived

post-deprivation

of Reberg

by previously

the earlier conditions

movement

support

patterns

of locomotton

seen during Phase One performance

front or back panel pacing with relatively

trrbution

of the movement

patterns

by alterations

instead of simply altering the ongoing schedule,

in deprivatron

we should consider

state

If the matching

the potential

carry-over

law 1s to be applied effects of changing

to stereotypic

study seem to sup-

m the schedule

of food

the present

data resulted from

to response

stereotyples,

deprivatton

condittons

then

and its ef-

fect on the bias parameter Many of the carry-over

effects may be due to discrunmative

toires by the experimental stimuh might be assessed state training phases.

context

(Balsam & Tomie,

Perhaps

behavior

patterns

accompamed

The comparison induced

behavior

the change m deprivation

of motivatronal is strained

mechanisms

by the dlstmctiveness

This putative

of those recorded

and mamtamed

reper-

control by contextual

contexts with the different

more indicative

One for both groups FW and WF would be elicited roundings

1985)

by pairing umque experimental

control over the established

deprivation during Phase

if unique experimental

sur-

regimen and their role m the development of the form of the behavior

of schedule-

associated

with the

48

two motrvatronal

mechanisms

board acttvity, enables of the behavior

recording

Although

observer-coding

schedule-induced methods

compartson

Moreover,

of mdivrdual animals IS strengthened

sible by automated companies

It 1s for this reason that a single, oblectlve

a defensible

of behavior

by the breadth

there 1s some sacrtftce

of behavior,

behavtor

the

based on replicable

behavior

the rehabtlty

establishment

measurement

measure,

floor-

of the characterrzatton

of coverage

that 1s made pos-

m the richness

of detail that ac-

of general

wrll require

prtncrples

mcreasingly

of

systematic

analysis

REFERENCES Allan, R W and Matthews, and water induced vention, Balsam,

Baltimore,

Killeen,

Assoctatrve

Paper presented

and ehctted

factors m the modulatton

at Eastern

Psychologtcal

of food

Assocration

Con-

Maryland

P D. and Tomte,

Hdlsdale,

T J (1982)

stereotyptes.

A (1985)

Context

and Learmng

Lawrence

Erlbaum

Assoctates,

New Jersey.

P. (1975).

Pear, J J (1988)

On the temporal Behavioral

control of behavior.

stereotypy

Psychologzcoi Revrew, 82,89-l

and the generalized

matching

equatron.

15.

Journal of the

Expenmental AnalysLs of Behavror, 50,87-95. Reberg,

D , Inms, N K., Mann, B. and Etzenga,

periodic

response-independent

Skmner, B.F. (1948). Superstttton Staddon,

J.E.R

(1977)

Staddon,

J E.R., and Simmelhag,

tts rmphcattons

behavior

resultmg

from

m the pigeon.

Journal of Expenmental Psychology, 38,168-172.

behavior.

of adaptive

26,507-S 19

In W.K Homg and J E R. Staddon

Appleton-Century-Crofts,

V.L (1971)

for the prmctples

‘Superstttious”

of food or water. AnzmalBehavlour,

Schedule-induced

Handbook of Operant Behavior.

C. (1978)

presentattons

The ‘superstttton’ behavior.

(Eds ),

New York experiment:

Psychologxal

A reexammatton

Revrew, 78,3-43.

of