JOURNAL OF THE EXPERIMENTAL ANALYSIS OF BEHAVIOR
1997, 68, 47–66
NUMBER 1 (JULY)
FOOD AND AMPHETAMINE SELF-ADMINISTRATION BY
BABOONS: EFFECTS OF ALTERNATIVES
RICHARD W. FOLTIN
NEW YORK STATE PSYCHIATRIC INSTITUTE
AND COLUMBIA UNIVERSIT Y
The effects of the availability of an alternative reinforcer on responding maintained by food pellets
or ?uid solutions were examined in 6 adult male baboons (Papio cynocephalus anubis). During daily
23-hr experimental sessions, baboons had concurrent access to both food pellets and ?uid, with
responding maintained under ?xed-ratio schedules of reinforcement that varied between the two
commodities. The ?xed-ratio requirement, or cost, for pellets was increased when (a) no ?uid, (b)
a dilute dextrose vehicle, (c) 0.002 mg/kg d-amphetamine, or (d) 0.004 mg/kg d-amphetamine was
available. When given nonrestricted concurrent access to food pellets and amphetamine at minimal
cost (FR 2), baboons self-administered suf?cient amphetamine to decrease pellet intake. Increasing
the response requirement for pellets decreased pellet intake at a similar rate regardless of the avail-
able ?uid and increased ?uid intake in a variable manner among baboons such that there were no
statistically signi?cant increases in ?uid intake. In contrast, when access to pellets was restricted to
70% of maximal intake under nonrestricted conditions, increasing pellet cost decreased pellet intake
and increased ?uid intake more rapidly when the high amphetamine dose was available. Thus,
amphetamine was more effective as an economic substitute for pellets when access to pellets was
restricted. The response cost for vehicle and both amphetamine concentrations was increased when
baboons had nonrestricted and restricted access to pellets. Increasing the response requirement for
?uid delivery decreased intake of all three ?uids similarly under both pellet-access conditions. The
results indicate that substitution between commodities with minimal commonalities can be studied
under controlled laboratory conditions and is dependent upon reinforcement schedule and com-
modity restrictions.
Key words: food intake, amphetamine, ratio schedules, self-administration, behavioral economics,
demand, baboon
Drugs of abuse, as well as reinforcers nec-
Stotz, 1983), (c) sedatives and anxiolytics with
essary for life, such as food and water, main-
food-intake-increasing effects (Carroll, Stotz,
tain responding under a wide range of ex-
Kliner, & Meisch, 1984), and (d) hallucino-
perimental conditions (Johanson & Schuster,
gens (Carroll & Stotz, 1984).
1981). Research with laboratory animals has
The interaction between drug and food
examined the effects of food availability on
self-administration is complex, with changes
drug self-administration (e.g., Carroll & Ro-
possible at different points in the chain of
defer, 1993). Among the most robust proce-
behavior leading to drug use. For example,
dures for increasing drug self-administration,
the availability of an alternative reinforcer re-
for example, is to maintain laborator y
duces the rate of acquisition, or initiation, of
animals at reduced body weight (see review
cocaine self-administration by rats (Carroll,
by Carroll & Meisch, 1984). Food depriva-
Lac, & Nygaard, 1989). The interactive effects
tion, resulting in decreased body weight, in-
of a nondrug reinforcer on drug self-admin-
creases the oral and intravenous self-admin-
istration are also in?uenced by the cost of
istration of (a) ethanol, which contains
both commodities and by income (Carroll &
calories (Meisch & Thompson, 1973), (b)
Rodefer, 1993). Several authors have turned
stimulants with anorectic effects (Carroll &
to the ?eld of behavioral economics to pro-
vide a framework for interpreting the effects
This research was supported by Grant DA-04130 from
of nondrug reinforcers on drug intake (Bick-
The National Institute on Drug Abuse and was approved
by the New York State Psychiatric Institute Animal Care
el, DeGrandpre, & Higgins, 1995).
and Use Committee. The assistance of Julian Perez, Ni-
The application of economic principles to
cole Hamilton, Meg Haney, Sandra Comer, and Moha-
the experimental analysis of behavior empha-
med Osman is gratefully acknowledged.
sizes the importance of studying responding
Reprints may be obtained from the author at New York
State Psychiatric Institute, 722 West 168th St., Unit 54,
maintained by a commodity at more than one
New York, New York 10032.
response requirement or cost (Allison, 1981,
47

---

48
RICHARD W. FOLTIN
1983; Hursh, 1980, 1984; Hursh & Bauman,
within this latter category of dif?cult-to-pre-
1987; Lea, 1978). When only one commodity
dict substitution. The economic construct of
is available, increasing the cost for that com-
substitution, based on empirical data, may
modity will increase responding for that com-
provide a useful framework for describing
modity until a cost is reached that maximizes
such complex interactions among reinforcers
responding. Increasing the cost above that as-
(Bickel et al., 1995; Carroll, 1996).
sociated with maximal responding will cause
Most studies involving drug self-administra-
responding to decrease below maximal levels.
tion and alternative reinforcers have been de-
A demand curve can be used to represent the
signed to examine how the alternative rein-
relationship between consumption and cost
forcer, most often the essential commodity,
of the commodity. When responding increas-
food, affects drug consumption. In the cur-
es with increasing cost, demand is said to be
rent study, we were interested in how an al-
inelastic, and when responding fails to in-
ternative drug reinforcer affects responding
crease with increasing cost, demand is said to
that is maintained by food pellets. The pur-
be elastic. The availability of a second com-
pose of the present study was to examine how
modity can alter the pattern of responding
responding for food pellets would be affected
when cost increases for the primary commod-
by the concurrent availability of an oral am-
ity: (a) The second commodity may have no
phetamine solution. The interactions be-
effect on responding for the primary com-
tween these two commodities were examined
modity, indicating that the commodities are
when baboons had nonrestricted access to
independent; (b) consumption of the second
pellets and when access to pellets was restrict-
commodity may decrease as consumption of
ed. Amphetamine was chosen because it de-
the primary commodity decreases, indicating
creases the food intake of laboratory animals
that the two commodities are complements;
and humans (Foltin, 1993; Foltin & Fisch-
or (c) consumption of the second commodity
man, 1988; Foltin, Kelly, & Fischman, 1990),
may increase as consumption of the primary
and it also maintains oral self-administration
commodity decreases, indicating that the two
(Carroll & Stotz, 1983; de la Garza & Johan-
commodities are substitutes (Green & Freed,
son, 1987; Ja¨nicke & Coper, 1984). This study
1993; Hursh & Bauman, 1987).
provided a parametric evaluation of the abil-
Research from this laboratory has recently
ity of self-administered drug and food pellets
been directed toward examining the factors
to function as economic substitutes for each
that affect responding maintained by food in
other, and it provided data on the extent to
baboons. These studies have shown that in-
which such substitution was dependent upon
take of a caloric alternative (dextrose solution
food-access conditions.
or identical pellets) increases when the cost
for food pellets increases. In addition, in-
creases in responding maintained by pellets
METHOD
were signi?cantly smaller when an alternative
Animals and Apparatus
source of calories was available, that is, de-
mand for food was more elastic in the pres-
Six adult male baboons (Papio cynocephalus
ence of a caloric alternative (Foltin, 1992,
anubis), initially weighing 24.4 to 47.0 kg, were
1994).
housed in standard primate cages (0.94 by
Although substitution among commodities
1.21 by 1.52 m high) at The New York State
can often be predicted by physical character-
Psychiatric Institute. The baboons had 3 to 5
istics (e.g., different brands of socks, white
years experience responding under ?xed-ratio
and whole wheat bread, pens and pencils),
(FR) schedules of food and ?uid delivery. The
even in these cases it is dif?cult to predict the
room was illuminated with ?uorescent lighting
extent to which one commodity will substi-
from 6:00 a.m. to 6:00 p.m. daily. In addition
tute for another. In many cases, however,
to food and ?uid earned during experimental
even partial substitution among commodities
sessions, two chewable vitamins (Kiddy Chews,
cannot be predicted based upon physical
Schein Pharmaceutical, Inc.), two pieces of
characteristics (e.g., bread and cole slaw).
fresh fruit (80 to 100 kcal each), and a dog
Clearly, most studies involving drug self-ad-
biscuit (150 kcal, Old Mother Hubbard, Inc.)
ministration and an alternative reinforcer fall
were also given daily. Water was available ad

---

SUBSTITUTION BETWEEN FOOD AND DRUG
49
libitum from a spout located at the back of
below when pellet intake was restricted. The
each cage. A response panel holding, from
remaining hour of the day was used for cage
bottom to top, a food hopper, two Lindsley
and animal maintenance. During mainte-
levers spaced 0.3 m apart (Gerbrands), four
nance periods food and ?uid spillage was not-
stimulus lights (two above each lever), a ?uid
ed. Spillage is rare with baboons, who nearly
spout, and a pellet dispenser (BRS-LVE Model
always consume delivered items.
PDC-005) was attached to the front of each
Initially, the FR requirement for both com-
cage. Resting on a shelf atop each cage were
modities was two responses, resulting in the
a 4-l bottle for ?uid solutions and a peristaltic
delivery of a 1-g food pellet or 5 ml of ?uid.
pump (7543-06 with pump head 7016 result-
The response requirement was increased, in
ing in a ?ow rate of 10 ml/min; Cole Parmer
ascending order, for one commodity on Mon-
Instrument Co.). All schedule contingencies
days, Wednesdays, and Fridays, while the oth-
were programmed using Apple IIgs comput-
er commodity was available under a constant
ers located in an adjacent room.
FR 2 schedule. Thus each cost was in effect
for 2 or 3 days under each condition. The
Reinforcement Schedule
response cost for pellets was increased until
Responding on the lever to the baboon’s
total daily pellet intake decreased to about
right was maintained by food delivery (Noyes
190 g (mean pellet intake at minimal cost
Formula L banana-?avored 1-g food pellets
527 g; range, 330 to 768). The response cost
containing 3.7 kcal/g, 21.0% protein, 4.7%
for ?uid was increased until total daily ?uid
fat, 62.0% carbohydrate, 5.3% ash, 3.1% mois-
deliveries decreased to about 10 (mean num-
ture, and 3.0% ?ber) under an FR schedule.
ber of deliveries at minimal cost
121;
Responding on the lever to the baboon’s left
range, 20 to 246). Responding for pellets was
was maintained by delivery of a dilute dextrose
maintained under FR 2, 16, 32, 64, 128, and
solution (D-( )-glucose, 0.625 or 1.25 kcal per
160 schedules, and responding for ?uid was
delivery; Sigma) that in some conditions con-
maintained under FR 2, 8, 16, 32, 64, and 128
tained d-amphetamine sulfate (0.002 or 0.004
schedules. Given that the maximal FR value
mg/kg per delivery; Sigma). Illumination of a
was dependent upon each baboon’s behavior,
red stimulus light above each lever indicated
the maximal values varied slightly among ba-
the availability of the commodity associated
boons (Table 1).
with that lever. Under the FR schedule, a re-
inforcer was delivered upon completion of a
Procedure
set number of responses (i.e., lever pulls). A
Responding of each baboon was studied
30-s timeout followed each reinforcer delivery:
under two pellet-availability conditions. In the
No stimulus lights were illuminated, and re-
?rst condition, baboons had nonrestricted ac-
sponding, though recorded, had no pro-
cess to pellets, and during the second con-
grammed consequences. Although baboons
dition, the maximum number of pellets that
could respond on either lever at any time, re-
could be earned each day was limited to 70%
sponding did not initiate a new FR until the
of pellet intake under the initial nonrestrict-
timeout had elapsed. The ?rst response of a
ed condition. Under each pellet-availability
new FR started a limited-hold timer and elim-
condition, seven pellet-?uid access manipu-
inated the schedule opportunity on the alter-
lations were completed. Responding was re-
nate lever. These changes were indicated by
corded when baboons had access to (a) only
stimulus changes on the response panel: The
pellets, and the response requirement for
red lights above both levers were extinguished
pellets was increased; (b) pellets and a dilute
and a green stimulus light was illuminated
dextrose vehicle solution, and the response
above the chosen lever. If the response re-
requirement for pellets was increased; (c)
quirement was not completed during the lim-
pellets and a low-dose amphetamine solution
ited-hold period (which varied with the re-
(0.002 mg/kg), and the response require-
sponse requirement: FR value
10 s), the 30-s
ment for pellets was increased; (d) pellets
timeout was initiated and the ratio was reset.
and a high-dose amphetamine solution
The schedule was in effect 23 hr per day, 7
(0.004 mg/kg), and the response require-
days per week, from 10:00 a.m. to 9:00 a.m.
ment for pellets was increased; (e) pellets and
the following morning, except as described
a vehicle solution, and the response require-

---

50
RICHARD W. FOLTIN
Table 1
Schedule of conditions and response requirements for each baboon.
Nonrestricted access to pellets
Restricted access to pellets
Fluid
Fluid
Baboon
Pellet cost
Type
Cost
Pellet cost
Type
Cost
1
FR 2–128
None
FR 2–160
None
FR 2
1.25 Da
FR 2–64
FR 2–160
1.25 D
FR 2
FR 2–128
1.25 D
FR 2
FR 2
1.25 D
FR 2–128
FR 2–128
0.004 Ab
FR 2
FR 2–160
0.002 A
FR 2
FR 2
0.004 A
FR 2–128
FR 2
0.002 A
FR 2–128
FR 2
0.002 A
FR 2–128
FR 2–160
0.004 A
FR 2
FR 2–128
0.002 A
FR 2
FR 2
0.004 A
FR 2–128
2
FR 2–48
None
FR 2–128
None
FR 2–48
1.25 D
FR 2
FR 2
0.625 D
FR 2–64
FR 2
1.25 D
FR 2–48
FR 2–128
0.625 D
FR 2
FR 2
0.004 A
FR 2–48
FR 2–128
0.004 A
FR 2
FR 2–64
0.004 A
FR 2
FR 2
0.004 A
FR 2–128
FR 2–48
0.002 A
FR 2
FR 2–128
0.002 A
FR 2
FR 2
0.002 A
FR 2–48
FR 2
0.002 A
FR 2–128
3
FR 2–128
None
FR 2–160
None
FR 2–128
1.25 D
FR 2
FR 2–128
0.625 D
FR 2
FR 2
1.25 D
FR 2–128
FR 2
0.625 D
FR 2–128
FR 2
0.004 A
FR 2–128
FR 2
0.002 A
FR 2–128
FR 2–128
0.004 A
FR 2
FR 2–160
0.002 A
FR 2
FR 2–128
0.002 A
FR 2
FR 2
0.004 A
FR 2–128
FR 2
0.002 A
FR 2–32
FR 2–160
0.004 A
FR 2
4
FR 2–128
None
FR 2–160
None
FR 2
1.25 D
FR 2–64
FR 2–160
1.25 D
FR 2
FR 2–128
1.25 D
FR 2
FR 2
1.25 D
FR 2–128
FR 2–128
0.004 A
FR 2
FR 2–160
0.002 A
FR 2
FR 2
0.004 A
FR 2–128
FR 2
0.002 A
FR 2–128
FR 2
0.002 A
FR 2–64
FR 2–160
0.004 A
FR 2
FR 2–128
0.002 A
FR 2
FR 2
0.004 A
FR 2–128
5
FR 2–128
None
FR 2–160
None
FR 2
1.25 D
FR 2–128
FR 2
0.625 D
FR 2–128
FR 2–128
1.25 D
FR 2
FR 2–160
0.625 D
FR 2
FR 2–128
0.004 A
FR 2
FR 2
0.002 A
FR 2–128
FR 2
0.004 A
FR 2–128
FR 2–160
0.002 A
FR 2
FR 2
0.002 A
FR 2–128
FR 2
0.004 A
FR 2–128
FR 2–128
0.002 A
FR 2
FR 2–160
0.004 A
FR 2
6
FR 2–128
None
FR 2–128
None
FR 2
1.25 D
FR 2–64
FR 2
1.25 D
FR 2–64
FR 2–128
1.25 D
FR 2
FR 2–128
1.25 D
FR 2
FR 2–128
0.004 A
FR 2
FR 2
0.002 A
FR 2–48
FR 2
0.004 A
FR 2–128
FR 2–128
0.002 A
FR 2
FR 2–64
0.002 A
FR 2
FR 2
0.004 A
FR 2–64
FR 2
0.002 A
FR 2–128
FR 2–128
0.004 A
FR 2
a D, dextrose (kcal/delivery).
b A, d-amphetamine sulfate (mg/kg/delivery).
ment for ?uid was increased; (f) pellets and
Testing order was systematically varied among
a low-dose amphetamine solution, and the re-
baboons. When baboons were ?rst given ac-
sponse requirement for ?uid was increased;
cess to the dextrose vehicle, and whenever the
and (g) pellets and a high-dose amphetamine
?uid commodity was changed, responding was
solution, and the response requirement for
maintained under an FR 2 schedule of pellet
?uid was increased.
delivery and an FR 2 schedule of ?uid delivery
Table 1 lists the order of testing each con-
for 7 to 10 days to allow ?uid intake to stabi-
dition for each baboon and the range of FR
lize. Following completion of the seven con-
values that were tested under each condition.
ditions with nonrestricted access to pellets, re-

---

SUBSTITUTION BETWEEN FOOD AND DRUG
51
sponding was maintained under an FR 2
casions per day (an occasion was de?ned as
schedule of pellet delivery without ?uid (ex-
beginning with the ?rst response for a com-
cept water) available for 2 weeks, providing
modity and ending when there was a pause
the baseline intake, on an individual basis, that
longer than 10 min between reinforcement
was used to determine pellet intake under
and the initiation of responding under anoth-
nonrestricted access conditions. The maxi-
er FR).
mum number of pellets that could be earned
Dependent measures were analyzed using
each day was set to 80% of baseline for 1
repeated measures analyses of variance with
month and was then reduced to 70% for 1
two within-subject factors. The ?rst factor was
month prior to retesting the seven conditions.
experimental condition (no alternate, where
Once a baboon had earned all his pellets dur-
appropriate; dextrose; low and high amphet-
ing a session, the red stimulus light above the
amine doses), and the second factor was food
food lever was no longer illuminated, and re-
or ?uid cost (each baboon provided data for
sponding on the pellet lever had no pro-
?ve response costs). Separate analyses were
grammed consequences. Preliminary data ob-
conducted under each pellet-access condition
tained with a higher dose of amphetamine
for responding when the cost for pellets was
than used here (0.008 mg/kg), indicated that
increased and for responding when the cost
some baboons that self-administered this dose
for ?uid was increased. Given the large indi-
would stop eating altogether. In order to in-
vidual differences in pellet and ?uid intake,
sure the baboons’ safety, the number of ?uid
the analyses of the number of reinforcers
deliveries was initially limited to 120 per day.
were also accomplished using data that were
This was increased to 240 shortly into the
converted to proportion of baseline intake
study for most baboons, but a protocol error
under that condition. Results were consid-
inadvertently left the ?uid maximum at 120
ered statistically signi?cant at p
.05.
for Baboons 4 and 5 under the nonrestricted
access condition when the high amphetamine
dose was available. Although not reported,
RESULTS
three experimental conditions were also ex-
Effects of Increasing Pellet Cost on
amined under the 80% access conditions for
Pellet and Fluid Intake
Baboons 1, 3, 4, and 5, such that they were
maintained under 80% access conditions for
The top panels of Figure 1 present total
about 6 weeks longer than the remaining 2
daily pellet intake as a function of pellet cost
baboons. Responding was recorded when
when baboons had restricted and nonrestrict-
these baboons had access to (a) only pellets,
ed access to pellets and concurrent access to
and the response requirement for pellets was
each of the three ?uids. Providing baboons
increased; (b) pellets and a dilute dextrose ve-
with a ?uid alternative signi?cantly decreased
hicle solution, and the response requirement
total pellet intake under nonrestricted access
for pellets was increased; and (c) pellets and
conditions, F(3, 15)
20.14, p
.0015. In-
a vehicle solution, and the response require-
creasing the cost of each pellet signi?cantly
ment for ?uid was increased.
decreased total pellet intake, F(4, 20)
20.14, p
.0001. When access to pellets was
Data Analysis
restricted, total pellet intake was identical un-
Data collected on the 2nd day of each FR
der the FR 2 condition. Increasing the cost
condition (3rd day if the FR was changed on
for pellets signi?cantly decreased total pellet
Friday) were included in the analysis. Re-
intake, F(4, 20)
16.17, p
.0001. There was
sponding and reinforcement were recorded
a signi?cant main effect of ?uid condition on
throughout the day, providing each of the fol-
total pellet intake, F(3, 15)
27.31, p
.003,
lowing dependent measures under all FR
and a signi?cant Condition
Cost interac-
conditions: cumulative intake throughout the
tion, F(12, 60)
2.00, p
.039. Contrasts
day; running rate (responses per second dur-
indicated that total pellet intake at maximum
ing the time from the ?rst to the last response
cost when the high amphetamine dose was
in the FR); number of started ratios that were
available was less than total pellet intake un-
not completed within the limited hold; and
der either the low amphetamine dose, F(1,
mean size and number of consumption oc-
60)
8.34, p
.011, or dextrose conditions,

---

52
RICHARD W. FOLTIN
Fig. 1.
Mean total daily pellet intake as a function of pellet cost and type of concurrently available ?uid when
baboons had nonrestricted access to pellets (top left). Mean total daily pellet intake as a function of pellet cost and
concurrently available ?uid when access to pellets was restricted (top right). Data presented in top panels converted
to proportion of intake at minimal cost (FR 2) (bottom). Error bars, representing 1 standard error of the mean
(SEM), are presented for the no ?uid and high amphetamine dose only.
F(1, 60)
7.12, p
.017. Clearly, pellet in-
7.34, p
.003, decreased with increasing
take decreased more rapidly as cost increased
cost, F(4, 20)
20.89, p
.0001, and was
when amphetamine was available and pellet
lower at maximum cost when the high am-
intake was restricted.
phetamine dose was available compared to
Given the variability in baseline pellet in-
dextrose, F(1, 60)
10.16, p
.002.
take when access to pellets was not restricted,
Figures 2 and 3 present total daily pellet
the data are regraphed in the bottom panels
(g) and amphetamine (mg/kg) intake as a
of Figure 1 as a proportion of pellet intake
function of ?uid condition (dextrose vehicle
under minimal cost. Converting pellet intake
and 0.002 and 0.004 mg/kg amphetamine)
to a proportion of baseline produced curves
and pellet cost for each baboon. The data
that overlapped substantially when pellet in-
shown in Figure 2 were obtained when ba-
take was unrestricted. Although total pellet
boons had nonrestricted access to pellets,
intake still decreased as cost increased, F(4,
and the data shown in Figure 3 were obtained
20)
50.67, p
.0001, there were no differ-
when baboons had restricted access to pellets.
ences among ?uid conditions. In contrast,
Because logarithmic scales are commonly
converting pellet intake to a proportion of
used in studies that manipulate response cost,
baseline when pellet intake was restricted did
the data in Figures 2 and 3 are graphed using
not alter the pattern of results: Pellet intake
logarithmic scales. Baboons 2, 3, and 5 in-
differed among the four conditions, F(3, 15)
creased the number of ?uid deliveries under

---

SUBSTITUTION BETWEEN FOOD AND DRUG
53
Fig. 2.
Total daily amphetamine (milligrams per kilogram; open symbols) and pellet intake (in grams; closed
symbols) as a function of pellet cost and amphetamine dose for each baboon under the nonrestricted access con-
dition. A protocol error inadvertently left the ?uid maximum at 120, indicated by the dotted line, for Baboons 4 and
5 under the nonrestricted access condition when the high amphetamine dose was available. Note that axes are
logarithmic.

---

54
RICHARD W. FOLTIN
Fig. 3.
Total daily amphetamine (milligrams per kilogram; open symbols) and pellet intake (in grams; closed
symbols) as a function of pellet cost and amphetamine dose for each baboon when access to pellets was restricted
to 70% of baseline under the nonrestricted access condition. Note that axes are logarithmic.

---

SUBSTITUTION BETWEEN FOOD AND DRUG
55
the restricted access condition to the maxi-
amphetamine dose increased by 9%. Given
mum 240 (300 kcal per day, or the equivalent
the variability in baseline ?uid intake, the
of 81 pellets; data not shown). Because this
data are regraphed in the bottom panels of
left no room for possible increases in ?uid
Figure 4 as a proportion of ?uid intake under
deliveries under amphetamine conditions,
minimal cost. Although ?uid intake increased
the dextrose concentration was halved (0.625
by about 100% to 250% under both pellet-
kcal per delivery), and the maximum number
access conditions, there were no signi?cant
of ?uid deliveries was increased to 360. Al-
differences related to either ?uid condition
though this change reduced the number of
or pellet cost.
vehicle deliveries, it complicates comparisons
between the two food-access conditions. Ex-
Effects of Increasing Fluid Cost on
amination of the individual data (Figures 2
Fluid and Pellet Intake
and 3) indicates that ?uid intake increased
Figure 5 compares total daily amphetamine
with increasing pellet cost for some baboons
intake as a function of unit price (i.e., num-
and under some ?uid conditions, but the ef-
ber of responses emitted to obtain 0.002
fects were not consistent. In summary, (a) in-
mg/kg amphetamine) for each baboon when
creasing the cost for pellets decreased pellet
baboons had restricted and nonrestricted ac-
intake under both food-access conditions, (b)
cess to pellets. Because the same FR values
pellet intake decreased more rapidly with in-
were used for both the low and high am-
creasing cost when amphetamine was avail-
phetamine doses, the unit price for the low
able and food intake was restricted, and (c)
dose at each FR value was doubled (i.e., half
?uid intake increased for some baboons un-
the amount of drug was delivered following
der some ?uid conditions, but the effects var-
the same number of responses), shifting the
ied greatly among baboons.
curves for the low amphetamine dose to the
In order to provide estimates of stability,
right of the curves for the high amphetamine
the range, mean, and median of the number
dose. The data in Figure 5 are graphed on
of reinforcers obtained under each FR 2 con-
logarithmic axes and are presented as a func-
dition are presented in Table 2. In nearly ev-
tion of unit price because such presentations
ery case, the mean and median of the distri-
are often used in studies that vary response
butions are within 10 reinforcers. The
requirements.
minimum number of reinforcers obtained
Figure 6 presents the number of ?uid de-
under each FR condition almost always oc-
liveries as a function of ?uid cost when ba-
curred on the 1st day of that condition, ex-
boons had restricted and nonrestricted access
cept under FR 2 conditions, when the maxi-
to pellets and concurrent access to each of
mum responding usually occurred on the 1st
the three ?uids. These data are presented as
day. Because FR values were tested in increas-
number of deliveries per FR value, rather
ing order, FR 2 conditions always occurred
than drug dose each day as a function of am-
after the largest FR tested for each baboon.
phetamine unit price, because dextrose con-
For this reason, data presented in the ?gures
tained no amphetamine.
were obtained on the 2nd day of 2-day con-
When baboons had nonrestricted access to
ditions and the 3rd day of 3-day conditions.
pellets (top middle panel of Figure 6), there
The top panels of Figure 4 present total
were no differences among ?uid conditions
number of ?uid deliveries as a function of
in intake as a function of cost. Increasing the
pellet cost and ?uid condition. The total
cost per ?uid delivery signi?cantly decreased
number of ?uid deliveries increased slightly,
intake of all three ?uids, F(4, 20)
10.55, p
although not signi?cantly, with increasing
.0001. With one exception (Baboon 6),
pellet cost when baboons had nonrestricted
there was no evidence for increased amphet-
access to pellets: Dextrose intake increased by
amine intake under the restricted access con-
13%, and intake of both amphetamine doses
ditions compared to the nonrestricted con-
increased by 25%. Intake of the high am-
dition (although the sweetness of the vehicle
phetamine dose increased by 63% (p
.06)
was decreased for Baboons 2, 3, and 5).
with increasing pellet cost, when baboons had
When access to pellets was restricted (top left
restricted access to pellets, whereas dextrose
panel of Figure 6), increasing the cost per
intake increased by 1%, and intake of the low
?uid delivery signi?cantly decreased intake of

---

56
RICHARD W. FOLTIN
Table 2
Indexes indicating variability in commodity intake when commodities were available under a
?xed-ratio 2 schedule of reinforcement (?ve to eight sessions). Values indicate the number
of deliveries per session.
Nonrestricted access to pellets
Restricted access to pellets
Baboon
Commodity
Min
Max
Mean
Median
Min
Max
Mean
Median
1
Pellets
597
688
641
652
449
449
449
449
1.25 kcal/ml dex
38
181
77
59
55
183
134
144
Amph
47
171
110
117
77
162
125
125
Amph
2
25
120
76
70
55
97
77
80
2
Pellets
319
354
332
329
233
233
233
233
1.25 kcal/ml dex
111
202
166
166
0.625 kcal/ml dex
102
141
125
128
Amph
115
160
136
137
55
100
68
61
Amph
2
86
120
111
119
44
75
57
59
3
Pellets
246
306
279
281
191
191
191
191
1.25 kcal/ml dex
121
198
154
148
0.625 kcal/ml dex
88
175
120
108
Amph
50
66
61
64
89
148
119
116
Amph
2
49
104
66
60
72
128
100
100
4
Pellets
331
387
364
364
254
254
254
254
1.25 kcal/ml dex
151
221
177
180
125
195
151
146
Amph
31
78
54
52
57
139
113
118
Amph
2
92
120
109
110
66
134
87
84
5
Pellets
496
645
580
573
406
406
406
406
1.25 kcal/ml dex
129
264
217
241
0.625 kcal/ml dex
69
124
91
85
Amph
47
171
110
117
35
210
107
85
Amph
2
118
120
119
120
35
103
58
56
6
Pellets
302
379
338
338
237
237
237
237
1.25 kcal/ml dex
20
67
42
43
25
71
59
63
Amph
31
80
51
39
34
76
56
54
Amph
2
36
65
47
45
32
58
46
48
all three ?uids, F(4, 20)
60.05, p
.0001.
?uid cost when pellet access was restricted,
There was a signi?cant interaction between
F(8, 40)
2.61, p
.021.
?uid condition and ?uid cost, F(8, 40)
The top right panel of Figure 6 compares
2.38, p
.033: The number of dextrose de-
the total daily number of pellet deliveries
liveries was greater than the number of am-
among the three ?uid conditions as a func-
phetamine deliveries under the three lower
tion of increasing ?uid cost when baboons
costs, but was lower than the number of am-
had nonrestricted access to pellets. Contrasts
phetamine deliveries under the two higher
calculated for the Fluid Condition
Cost in-
costs. The bottom panels of Figure 6 present
teraction indicated that pellet intake signi?-
the data contained in the top panels after
cantly increased when the cost of the high
they were converted and analyzed as propor-
amphetamine dose was increased, F(1, 40)
tions of baseline. When analyzed as propor-
4.34, p
.044. Converting and analyzing the
tions of baseline, the data con?rm the results
data as proportion of baseline, as shown in
based on absolute intake: increasing the cost
the lower right panel, yielded a borderline
per ?uid delivery signi?cantly decreased in-
nonsigni?cant effect of cost (p
.09) and
take of all three ?uids when pellet access was
con?rmed that pellet intake increased when
not restricted, F(4, 20)
13.13, p
.0001,
the cost of the high amphetamine dose was
and when pellet access was restricted, F(4, 20)
increased, F(1, 40)
8.57, p
.0056. When
123.48, p
.0001, and there was a signi?-
access to pellets was restricted, pellet intake
cant interaction between ?uid condition and
remained at maximum levels regardless of

---