CASH CROP WJITH ANIMAL PRODUCTION SYSTEM:
COFFEE, SUGARCANE 67ITH DUAL-PURPOSE CATTLE
Centro Agronomico Tropical de Investigacion y Ensenanza
Turrialba, Costa Rica
Cofftee sugarcane and cattle production are tr adi ti o nal
activities on a large number of farms in Costa Rica and rep-
resent, together with cocoa and bananas, the country's main
income from agricultural exports (Table 1)and total exports
Farms of 50 ha or less represent 85% of the total number of
farms, but occupy only 20.3% of the country's total area in
use. In spite of this, they produce a considerable portion
of the cash-crop exports, especially coffee (Table 3).
Some Comparisons Between Cattle Production, Pere;nnial, and
Annual Crop Farming Systems
Guillen and Avila (1981) studied the relative importance of
the different crop-anim~al activrities on 230 farms in Costa
Rica. All of the farms had less than 50 ha, or less than 25
head of cattle (minimum of one); this represents 60% of the
total number of farms in Costa Rica.
Table 4 shows the proportion of land used and the production
value (PV) for these farms' main components. Pasture occu-
pies 80% of the productive area of the farms but generates
only 34% of the total PV. In spite of this, farmers still
keep a great part of their farm for livestock to reduce
risk, solve cash problems, assure daily consumption of basic
food by their families, and for efficient resource use. Us-
ing linear correlation techniques, Guillen and Avila (1931)
found that when farm size increases, cattle activity on the
farm also increases; however, the total PV is not correlated
with pasture area, but is related to the area dedicated to
crops. They also found no correlation between farm size and
As farm size increases, the role played by the annual crops
in the generation of income is constant, while the cattle
gain importance at the expense of perennial crops, despite
the fact these generate more income per surface area. The
use of land also seems to be affected by the farm size; as
farm size increases, there is a tendency to diminish the
proportion of land dedicated to perennial and annual crops,
keeping the proportion in pasture land more or less con-
stant. There is also an increase in the proportion of un-
used land, ranging from 4% on farms with less than 5 ha to
40% on the larger farms. The economic pr od uc tiv ity per
total area and per crop and pasture area, diminishes as farm
size increases. This is partially explained by the fact
that the small farms, having excess labor dedicate mnore
day-labor per production unit, and this labor is reflected
in greater productivity.
ACTI V ITY EXTENlSION! PRUDUCT ION- EXPORT- EXPORT VALU~ES-
(Thiousands of ha) (Thousands of metric tons) (Thousands of metric tons) (Millions of dollars
1972 1976 1976 1976
Table 1. EXTENSION CULTIVATED, PRODUCTION AND EXPORT OF
COSTA RICA's HOST IMP'ORTANT AGRICUL.TURAL ACTIVITIES
271.7 (milk, millions of It)
4.3 (life ca
2 4 (m lk)
uinder uIsL 3122.4 (100) 352.7 (100.0)
- Censo Aytupceturioo 1973 (Costa Rica, 1974).
utlicina~ du P~ljlificaii iIn Sectorial Agropecua~ria (OPSA) 1977 (Costa Rica, 1977).
3Anularms) de CI)mer.cia Exterior In: Oficina de Planificaci~n Sectorial Agrapecuaria, Diagn~stico del Sector
Agrrcpcuairio de Costa Rica, 1962-1976 (Costa Rica, 1976).
1 Percentage of totial
VAXLE VOLLME PERCENTAGE OF
(.UEllion USS) (Thousiands MT) TOTAL EXYPORTS
Coffee 201.: 80.4 23.0
Sugar 43.1 73.4 4.9
Fresh meat 55.6 24.1 6.3
Banana 212.8 864.7 24.3
Cacao 1.5 1.0 0.2
All ocher inaustrial or agricultural
products 361.7 -41.3
-Direcei~n Ceneral de EstudisticJ y Censos In: Oficina del Cat6 (1982).
rable 2. FOB VALUE IN USS, VOLUPE AND RELATIVE LMIPORTANCE
OF COSTA RICA's MA~IN EXPORTS, 1981-
%( OF TOTAL COFFEE SCCAR-CANE CATTL
Number of farms 93 86.5 77
Surface covered 66 34 23-
Production 61 27
From: Censo Agroprcuario 1973 (Costa Rica, 1974).
Number of heads
fable 3. RELATIVE IMPORTANCE OF FARMIS 50 RA. OR LEss,
ACCORDI::G TO NUMBER OF FARMS, SURFACE: COVERED
AND PRODUCTION. (PERCENTAGE OF TOTALS)
COMPONENT LAND USED ZN P.V.
Grassland 58 34
Perennial crops 10 61
Annual crops 5 5
Area not used for productive
Table 4. LAND USE AND PRODUCTION VALUE ACCORDING
TO THE: DIFFERENT FARM COMPONENTS
From: Guillen and Avila (1981).
Thre DiFffrentt Farming Sy~stems
Av~ita anmd Serpa (1931~), working with the same data used by
T;uillen7 andJ viiila (191), report th-at 35% of the farms were
compo~tsed of cattle and perennial crops (CP), 26;% cattle,
perennial and annual crops (C PA) 13% cattle and annual
crops (CA) and 26% cattle alone (C). Using records from 38
of these farms (April 1978 to M~arch 1979), they found that
the Total Production Value (TPV), Family Net Income (FN~I),
Net Income (NI), NI/ha and FNI/ha were higher for the CP
systems. They concluded that the mixed farming systems ,
specifically those with perennial crops, are economically
more efficient because they show a complementary effect in
the use of labor and income and output flows.
Using the same data, Avila and Treminio (1981) reported that
a typical farm in the Turrialba county has 14. 35 ha, 8 53
labor-days (LD) and $2,778 in operational capital per year.
These resources are allocated to 2.8 ha of coffee, 4.55 ha
of sugarcane, and 7 ha in a dual-purpose cattle system, and
produce a FN~I per year of $4,115. The economic optimization
obtained through linear programming resulted in 3.75 ha of
coffee and 2.49 ha of sugarcane, with a PNI of $4,3:::: The
limiting factor was lack of labor from July to October (cof-
fee harvest time). On the basis of these results, they con-
cluded that it is generally not possible to significantly
raise FNI by changing the farm activities and that the far-
mer combines his activities in an efficient way.
Analyzing the same data, Sancho (1981) found a higher labor
use on the CP farms, with 68 LD monthly ( coe f fic ient of
variation, CV = 30), followed by the CPA farms with 42 LD
(CV = 25), C farms with 28 LD (CV = 25) and CA farms wiith 21
LD (CV = 26).
The average monthly flow of cash income, cash output, and
cash balance was higher for the CP farms than any other farm
(Table 5). The coefficients of variation (CV) suggest that
the CP systems are more stable in the monthly cash flows
throughout the year, a characteristic desired by many far-
CATTLE/CASH-CROP FARMING SYSTEMS
A typical region, the Tuis district in Turrialba, was
selected to study the cattle/coff~ee/sugarcane farming
systems. This area was chosen because it was near CATIE
(Figure 1), information wras available from some farms in the
region, and the major products are of interest.
o~ so7C 7o
Fig. Stu y sie oro m p ( nte fro Ibrra, 1979
PhausSl O ES M in lwns x Alulde ( osi
IOT :I~ 01( 4 9I ol *3 ;itge l wahr o
o i I r s
;I3 = 10 19 9
4 2 -400 i '
5 4 rai ars E oOt: C
The physical environment. The average yearly rainfall
is around 2500 mm (La Suiza, 616 m) distributed more or less
evenly throughout the year with a relatively dry period
from January to April. March is the driest month with an
average of 70 mm and December. shows the largest rainfall
The average relative humidity (CATIE) is 87.5% (Beer, 19)79;
Rockenbach, 1931). The average yearly temperature is 22.30
C with an average maximum of 270 C and an average minimum of
17.60 C (Rockenbach, 1981).
According to Holdridge's Life Zones, the study area corre-
sponds to Premontane Rain Forest on the higher altitudes and
Premnontane W~et Forest in-the lower areas.
The farms studied are at an altitude between 900 and 1000 m
with very irregular topography and slopes ranging from 10 to
40%, although some areas have slopes over 40% (Figure 1).
The soils are predominantly of the Ultisoles, Typic Tropohu-
mult associated with Typic Hlumitropept. They are mainly of
a reddish color, deep, and clayish, with a pH varying from 5
to 6 (farms' soil analysis, CIID-CATIE Project (CATIE,
1979]). Darker, less developed soils are found on the upper
ridges. According to Ibarra et al (1979) the soils are
well-drained, with medium apparent fertility.
The primary sector. This is considered the productive
sector of the region; its components are the production
units (farms) with their socioeconomic components, agricul-
tural and animal agroecosystemns, natural ecosystems and the
physical resources. The flow of money agricultural and
animal production inputs, technological inputs, etc. inter-
act with the farm components, thus accounting for the
region's productive process.
In Tuis, there are 175 farms spread over 2542 ha, an average
of 14.5 ha per farm. A large number of very small farms
(less than 5 ha) are found near the higher population con-
centrations (near churches or stores). The larger farms are
more distant (only a couple of farms are greater than 100
The small farms comprise most of the area in use and usually
have a labor surplus that is sold to the larger farms. The
latter usually have the level areas planted with sugarcane,
the steeper slopes planted with coffee or in pastures. The
upper ridges usually are composed of secondary forest on the
highest parts and thickets on the forest edges, even though
these are commonly found in patches in the pastures and
alongside the roads and watersheds.
Of the total number of farms, j6% have coffee, 63% have cat-
tie, and52% sugarcane (Tables 6 and 7) (Costa Rica, 1974)
which aare by far the most important crops in the area. Only
8% of~ the far-~ms have banana monoculture and 8% have corn
mnono c ult u re. AllP other activities are found only in a
small number of farms.
The secondary sector Hsregional sector includes
all of the processing or redevelopment of products corning from
the Primnary Sector. Very little processing is done to the
products in Tuis, but in La Suiza (5 km) and Turrialba (20
km) (see Figure 1), many processing plants can be found
(sawmills, butchers, slaughterhouses, etc).
Regional service sector (third sec tor ). This sector
includes all services such as banks stores, mnar kets,
schools, etc. Pr od uct-marke t ing facilities are widespread
over the area. Coffee is sold mainly to Coopesuiza, a coop-
erative organization that takes charge of selling the whole
fruit to the processing plants
Sugar mills are abundant in Tulrrialba county. Most of the
sugarcane is transported by private truckers, many of whom
live in the district and have their own farms. Some sugar
mills send tractors with wagons to haul the sugarcane from
the farms to the mills.
Cattle are sold mainly on the farm to truckers who resell
them at Turrialba's slaughterhouse or who sell them in other
areas as feeder cattle.
There are many schools, stores, and churches in the district
or nearby areas. Banks, markets, warehouses, hospitals, and
other services are found in Tutrialba.
The population. In 1973, the Tuis district had 1422
inhabitants, dis tribute in 262 dwellings. Of these, 214
were occupied by a one family nucleus, with an average of
5.2 people; 37 were occupied by two families with an average
of 6.2 occupants per dwelling; and 11 were inhabited by
three or four families, with an average of 7.5 people (Costa
Rica, 1974). Usually, when morie than one family occupies a
dwelling, the second, third, and fourth families are descen-
dants from the first, and are recently married, or have one
Table 8 shows the distribution of the population 12 years or
older Table 9 shows the education level of population 6
years of age or older, with 17%j having no formal education.
The cash cr op/cattle farming systems' components strongly
interact with the economic process of the region, with
TABLE 5. AVERAGE MONTHLY FLOWS OF CASH INCOMES, OUTPUTS,
AND BALANCE IN THE DIFFERENT FARMS SYSTEMS.
(US$) CP (CV) C (CV) CPA (CV) CA (CV)
Cash incomes 610 (26) 389 (60) 237 (47) 210 (112)
Cash outputs 28 2 (37) 219 (79) 86 (23) 45 (193)
Cash balance 327 (39) 170 (166) 149 (74) 166 (155)
From: Sancho (1981).
TABLE 6. NUMBER OF COFFEE AND SUGARCANE PRODUCING FARMS,
SURFACE PLANTED ANNUAL PRODUCTION AND ON-FARM
CONSUMPTION, TUIS DISTRICT, 1973
Number Crop Production Consumption
of Farms Area (ha) (tons) (tons)
Coffee 133 218 630.4 0.66
Sugarcane 91 234 9215.0 4.00
From: Costa Rica (1974).
OBS: Total number of farms in Tuis District is 175.
TABLE 7. NUMBER OF FARMS WITH CATTLE, NUMBER OF HEADS AND
SEXS, TUIS DISTRICT, 1973.
of Farms Heads Females Males Oxen
Cattle 110 1,147 734 374 39
From: Costa Rica (1974).
OCCUPATION NO. %
Agriculture 347 85.0
Manufacture 2 0.5
Construction 8 2.0
Store 3 0.7
Transportation 23 5.6
Community, social, or
personal service 15 3.7
Not specified 10 2.5
From: Costa Rica (1974).
TABLE 9. DEGREE OF SCHOLAR INSTRUCTION IN POPULATION 6
YEARS OF AGE OR OLDER.
EDUCATIONAL LEVEL NO. %
Primary school 895 79.5
High school 35 3.1
University 4 0.4
No instruction 192 17.0
From: Costa Rica (1974).
ACTIVE POPULATION, 12
YEARS OF AGE OR OLDER
little diversity in the production process. H~owJever, the
multicropping/small livestock farm systems are usually rela-
tively independent of the regional systems, wJith a very
strong interaction among its components.
The Socioeconomic Component
A c omnpl ete family nucleus usually is composed of 6 to 8
people (4 to 6 children). The 5.2 person average mentioned
previously includes young couples with a few children or
elderly couples who usually hire the necessary labor.
The father and the sons take charge of all the hard field
work, while the women and younger children do the housework
and take care of the home garden and small livestock. Labor
for off-farm work is often hired from the smaller farms.
The male labor is hired out mostly for sugarcane harvesting,
fence fixing, or pasture weeding. Women and children are
hired out for the coffee harvest.
Most farms have a good house built of wood and frequently,
the "wet" areas (kitchen, bathroom, etc.) are made of con-
crete Almost all farms have the basic work implements
(backpack-pump sprayers, tools, etc). A few have oxen to
pull carts or ploughs, to rent, or to move home-sugarmills.
There are twelve trucks in the area that belong to wealthier
farmers who usually use the ir t ime in tran spor ta tion and
contract labor for their farms. Almost~no mechanization is
found in the region.
In the labor distribution throughout the year (Figure 2),
there is a complementary effect in the use of labor for cof-
fee and sugarcane with a nearly constant use in the cattle/
pasture system. If the pooled variances are compared with
the respective sum of individual variances (Table 10), we
see that there is an 85% reduction in the pooled variance
for the three systems. When two component-pooled variances
are compared with their respective variance sum, there is a
2% and an 88% reduction of the pooled variances for the cat-
tle-sugarcane and coffee-sugarcane componen ts respective-
ly. This finding suggests a positive or complementary
interaction between these components.
If the coffee and cattle pooled variance is compared with
the sum of the individual variances a 6.1% increase is
found in the pooled variance; in other words, there is a
-slight competition for labor between these two components.
Due to the out-of-phase harvest periods of coffee and sugar-
cane, and the steadiness of the cattle component, a relative
stability is seen in the monthly input and output flows ;
this _is contrary to the flows usually observed on farms with
only one of these perennial crops.
1 Sugar cone
:I : so I"'
J M AJ J AO
Fig 2 Monthly us flbro ofesgrcn/otepouto
overag loor( '1'110 tG1 C' DODM 00 Oer1e Cmpoe
size)Basedon dto frm CAIE C D
Cattle + S.Cane--------L0.7
Cattle + Coffee--------67, 1
Coffee + S.Cane--------63.9
Cattle + Cof fee +S.Cane!75 .9
From: CATIE (1979) .
Table 10. MEAN AND VARIANCE OF LABOR USE ACCORDING; TO
THE DIFFERENT COMPONENTS AND THEIR COMBINATIONS
Cattle/pasture, Thie-re are 1147 head of cattle spread
over 63: of '0.8? os, giving an average of 10.5 head per
farm. Of these, 734 (69,.) are females, 374 (32.6%) are in-
tact males, and 39 (3.4%) are oxen (Costa Rica, 1974).
The animals are a result of 'several crosses be twee n I3os
indicus and dairy Bos taurus, in a variety of combinations
and proportions. The approximate weight of the mature cows
varies from 350 to 400 kg.
Most farms have a dual-purpose system with milking only once
a day when calves are with the cows. The calves are sepa-
rated fromn their mothers in the early afternoon until the
next morning at milking time. The estimated milk production
above that consumed by the calf is between 500 and 300 kg
per lactation, which normally varies fromn 130 to 300 days,
depending on the cow and thle farmer's needs. M~any farmers
milk their cows during only the first few months to obtain
better weight gains on the calves. Farmers usually wean the
calves at 6 or 7 months. Male calves are sometimes sold at
this time but usually they are kept for 2 to 3 years,
depending on cash needs and pasture availability.
There is very little information about reproductive para-
meters. Apparently, the cattle are relatively fertile--far-
mers indicate that the cows have calves each year. The far-
mers who do0 not own a bull rent or borrow one from their
neighbors when a cow is in heat.
Most of the pastureland is covered with native species, but
a substantial area is in star-grass (Cynodon nlenfuensis) or
Almost all pastures are associated with trees, especially
Guave (Psidium guajava) for shade and firewood, or Laurel
(Cordia allidora) for shade and timber Miost fences are
strung on living fence-posts, the most common trees being
Gliricidia sepium. Firewood and new posts are the main pro-
ducts obtained from the pruning of these trees, which is
done every two to three years. A few farms have cut-and-
carry systems for feeding the calves, using mainly elephant
grass (Pennisetum pupureum).
A. Inputs : The farmers have an empirical but
steady, sanitary control. They apply products once a month,
to fight external parasites, mainly ticks (Boophilus micro-
plus) or larvae (Dermatobia hominis); and twice a year for
internal parasites, especially on the young animals. Very
few farmers vaccinate their cattle.
Animals receive common salt every two weeks throughout the
year. Some farmers feed the cattle sugarcane tops and
banana stems. Supplementation with molasses is scarce, due
Labor applied to this system is usually surplus from the
farm, mainly on Saturdays. Some work (such as weeding) is
done on weekdays once or twice a year (usually near the mid-
dle and end of the rainy season). There is a small, but con-
stant, labor use in milking and calf care.
B. Ou tpu ts : The most con stan t output is milk,
basically used for home consumption. Other occasional out-
puts are calves or steers sold when there is a need for
cash. The oxen power is another output, which is used main-
ly for carrying cane to the trucks in inaccessible areas, or
water used for spraying herbicides on the cane or coffee
Sugarcane. Sugarcane is widespread in the area; 52% of
the farms -have sugarcane over a total area of 223.5 ha.
Approximately 9200 tons are produced annually, for an aver-
age of 43 tons/ha/year (Costa Rica, 1974).
Sugarcane plantations are harvested over a period of about 4
to 7 years, after which a complete replanting process is
necessary. After every harvest, some replanting is done on
the spots where resprouting is weak, using the cut sugarcane
tops. This is one of the main reasons for observing a high-
er yield in the third and fourth years after establishment.
Harvest time is from January to September This period
includes the drier months, favoring the saccarose content of
the stem; labor is also more abundant since there is no cof-
fee harvested. Some farmers burn the stubble, a practice
which is slowly being discarded because of the greater use
A. Inputs: Herbicides are applied 3 months following
harvest. Afterwards, weeds are controlled only on the round
edges and roadsides--usually by hand in the rainy season or
with herbicides during the dry months. The plantation is
Oxen power is another input for hauling the cane from the
areas that are not accessible to tractors or trucks and also
to carry water to use with herbicides.
The cane harvest requires much labor (approximately 40
LD/ha) (Avila, n.d.), obligating farmers with larger areas
to hire labor or to harvest the whole cane by contract; that
is to pay a group of men to cut and transport all the cane.
Planting or replanting fields also takes much labor but
usually the farmer does this work during the harvest months
between one cut and the next (not all the cane ripens at the
B. Output: The main output is the cut and defoliated
cane Only 0.042% of this crop is consumed (Costa Rica,
1974) or sold in the nearby towns in a form of brown sugar
processed on the farm.
Coffee. The most important crop in the region is cof-
fee; 76% of the farms have coffee over a total area of 219
ha, producing 3.17 tons/ha/year.
Coffee plantations are always associated with shade trees.
Usually, in young plantations, banana (Mlusa sapient um) or
plantain (Musa parad isiaca) plants are the most commonly
used shade tree. Poro (Erythrina poeppigiana), a leguminous
tree, is widely used i n young and ol-er plran ta t ion s. In
some cases, a more complex, three-s tra ta system is used ,
with the lower coffee stratus (1.5-2m), the mid-level
banana, p~oro stratus (4-5 m), and an upper stratus composed
mainly of existing timber trees, mainly Cordia alliodora
( 10O- 15m ). Th is c om pl ex agroforestry system has b'een
described in detail by Beer (1979).
A. I npu ts : Al though h coffee is highly profitable,
labor requirements during harvest time limit the area that
can be properly managed by a farmer. Estimations based on
data collected by CATIE-CIID (1979) and interviews with far-
mers indicate that labor needs vary from 90 to 250 LD/ha
during harvest time which starts in August and ends in
January. Peak production occurs between September and
November. Harvesting is done by a group of people (family
or hired labor) who go through the whole plantation once
every 15 days picking the ripe fruits.
Pruning of the coffee trees and the shade trees also
requires labor. Coffee trees are pruned by removing all the
unpr oduc t ive old branches right after the end of harvest
time--an important source of firewood. At this time, almost
all of the foliage is cut off the Erythrina trees, and the
old banana stems are cut down. This allows greater light
intensity, facilitating the resprouting and flowc~ering of the
coffee. It also reduces the competition for water, since
these are the drier months of the year. The Cordia tree
drops its leaves, providing the same effects on systems
including this component.
In June or July, a second, less-seve~re, pruning is done on
the shade trees, especially on the Erythrina. Some farmers,
with greater labor limitations, do not prune until August or
early September; this delays the ripening of the fruit until
laborers are more available.
Some farmers fertilize their coffee trees one to three times
a year M~any farmers apply herbicides when pruning in
January. This is done, in spite of its high price, because
of the intensive use of labor on the pruning activities.
VPery few farmers use insecticides or fungicides, which are
usually applied during the drier season.
In some farms where the coffee area is being increased ,
plastic bags are an important input, since some tree nurser-
ies require them. Most farmers plant the trees on a small
piece of land and when transplanting cut the soil in
blocks around the root system and wrap it with banana
Oxen are used mainly to haul water when herbicides are used
and to transport the firewood collected during the pruning.
B. Outputs: The main output in this system is the
coffee itself. In spite -of the small area usually dedicated
to this crop, it represents an important part of the far-
mers' income. About 5 kg of the harvested coffee are con-
sumed on the farm.
Other outputs are bananas (mainly for human consumption, but
on some farms they are given to the hogs), plantains, banana
stems (given to the cattle), and firewood.
For e st a nd thi ck~ets Many of the larger farms (25 ha
or more) have great portion of land covered by thickets
consisting of bushlike vegetation from 2 to 3 m high. Some
farms have secondary forest covering the higher and steeper
areas, and especially the water sheds.
This component does not have well-defined limits, frequently
being found in patches inside the pastures. Its area is
very elastic through the years; when the farmer needs more
land for pasture or crops, he cuts the thickets or forests
(Guillen and Avila, 1981); when he sells the cattle for a
large investment, he abandons a section that soon transforms
itself into a thicket. Many times, however, the opposite is
true when the farmer cuts part of the forest to use the
timber for a new home and sells the cattle to collect enough
funds to build. To be able to use the timber trees, a per-
mit from the Forestry Department must be obtained. When the
wood is for on-the-farm use, the permit is easily obtained.
This agroecosystem also is an important firewood supplier,
usually requiring surplus labor.
Home gardens. This agroecosystem plays an important
part in almost all the farms; it supplies the family with a
constant amount of food and requires very few inputs.
Al though bound ar i es are not well-defined gardens usually
are found near the house.
Gardens consist of a large variety of food crops (cassava,
bananas, sweet potato, squash, chayote), fruit trees (cit-
rus, papaya, mangos), and medicinal plants. Chickens and
swine are sometimes kept in the area.
:ce -.n and .cilr supply most of the labor, except when
heavy weeding is required.
Crop components are usually managed under fixed amount of
labor or other impor tant inputs. Quantifying is a rather
difficult process, due to the complexity and the relatively
little importance given to it by the farmer.
Swine are not commonly found in the area, probably as a con-
sequence of the lack of~ annual crops that are an important
feed source for these animals. In Tuis, 48 farms (27%) had
a total of 137 swine (of these animals, 554 were less than 6
months old) (Costa Rica, 1974). In this area, green bananas
coming from the coffee agroecosystem are the basic part of
the feed. Usually, the small pigs are raised so that they
can be consumed during the Christmas holidays.
A CASE STUDY
The following is a description of a typical cash crop/dual-
purpose cattle production system. Most of the quantitative
data was collected from April 1978 to March 1979 as part of
a CIID-CATIE Project (CATIE, 1979). Figure 3 is a schematic
semiquantitative diagram of the farm.
The farm~ consists of 45.5 ha of which 7 ha are in pasture,
4.9 ha are in sugarcane, 0.53 ha in coffee, and with the
remainder mainly in thickets with a small area in forests.
A small area around the house is dedicated to the home
garden. Sugarcane and coffee are the farm's major outputs,
representing 78.9% and 10.6% of the TPV, respectively.
Cattle represent 9.9% of the TPV; poultry, 0.6%.
The family. The family consists of six members: the
farmer, his wife, three sons, and one daughter. The farmer
is 38 years old and is the principal decision-maker. All
family members have had basic school education. The sons
help the father on the farm, working occasionally in other
activities. The daughter helps the mother with the domestic
Installations: I house with 72 m2 built in 1953,
1 house with 64 m2 built in 1970, I warehouse
with 9 m2 built in 1977, 1 small building for
436 hl TRACTOR VORh 43sks ( 1685
Fig. 3 FARM OlAGRAM PARTIALLY CUANTIFIED.
Quantitotive dolo bonrd <0 CATIE-CllD
mi'lk.n-n withh 21 m2 built in 1973. There are
about ~'."00 m in fences--1200 m around pas-
tures and 800 m around sugarcane.
Equipment: Backpack-pump sprayer, Ox-cart, metal
barrel, milking pails, tools.
Animals: 4 cows, 1 heifer, 2 steers, 3 calves, 2
oxen, 9 poultry.
Labor distribution. Table 11 shows the annual labor
use in the farm's main activities. Coffee has the greatest
labor requirement per unit area (387 LD/ha), while sugarcane
uses 64.8% of the farm's labor over the year. Figure 4
indic ates that labor required for sugarcane production is
complementary with that used for coffee and cattle systems;
however, the two latter_ systems compete for this resource.
When the pooled variance of all the components is compared
to the sum of the individual variances, a reduction of 63%
is noted. In other words, the total variance is 63% smaller
than the sum of the component variances; there is a strong
complementary effect in the use of labor on thie different
components resulting in a more constant use of labor
throughout the year.
Wh en the pooled variances of coffee and sugarcane, cattle
and coffee, and cattle and sugarcane are compared with the
sum of the respective individual variances, the pooled vari-
ance is reduced 36% and 26% for the labor use in coffee and
sugarcane and cattle and sugarcane, respectively. However,
there is a 23% increase in the pooled variance of the cattle
and coffee labor use. These findings provide a quantitative
me sure of complementation in labor use in the coffee and
sugarcane and cattle and sugarcane components, as well as
for the competition for the labor resource between the cat-
tle and coffee components.
Economic Analysis. Table 12 provides a complete cost/
b e nefit analy s is wiFt h several prod uc tion i ndi c es based on
the information contained in the CATIE-CIID study (CATIE,
1979). The farm's NI is only 10.8% of the TPV; this is due
mainly to the noncash costs (labor, etc.) and to the fixed
costs (depreciation, interest, etc) If the money input
from outside the farm is not taken into account, the FNI
would be US$5,239 representing 60% of the TPV. The highest
variable costs (VC) are associated with sugarcane production
(70% of the total VC) and are due to the large labor
requirements. In spite of the cattle having the lowest NIJ
(US$299),r it gives the best return in terms of FNI per LD
invested--due mainly to the high fixed costs involved in
Except for the animal production, most of the farm's pro-
ducts (93.5%) are sold. This is typical of cash-crop farm-
ing system-s and accounts for the high level of interaction
of its components with the regional productive process. The
Table 11. AMOUNT OF LABOR USED EACFL; YEAR FOR \':UTDUS FARM~ ACTIVITIES.
TOTAL DAYS STANDARD
DAYS fl.:;S WORKED/ S: OF DEVIATION~
ACTIVITY WORKED WORKED HA TOTAL
Cattle FL1 107
CL/ 0 107 15 11.7 7.5
Coffee FL 193
CL 10 203 387 22.2 15.9
Sugarcane FL 393
CL 199 592 121 64.8 41.5
Poultry FL 12
CL 0 12 1.3 0.0
TOTAL 914 20 100.0 27.5
Source: CATIE (1979)
1/ Family Labor
2/ Contracted labor
Table 12. CO)ST-BENEFIT ANALtYSIS OF' THE6 FARM~S AN~D ITS COMPONENIiS, INCLUDING SUME PRUDUCTI~ON PARAMETERS
Coffee + banana
0.525 has (1.2\)
7 ha (15.41) 4.9 ha (10.89)
Total cash costs
Out of farm income
Production sold (%)
Percentage of TPV
Proportion sold (%)
Family labor (LD)
Hired labor (LD)
Total labor (LD) 107 (11.71)
LD ha year 15
fail'scash income amo un ts to US$9,817 of wh i ch 82.3%
comaies from.farm activities.
Cattle/pasture agroecosystem. Cattle play an important
role in the farm's economy and use of resources. Although
cattle represent only 9.9% of the farm's TPV and 3.6% of the
farm's sales (Table 12), they provide a source of readily
available cash throughout the year. In this extensive hus-
bandry system, the amount of fixed labor required is very
little (2.54 LD/month, s2 0.008) and does not represent a
great expense for the farmer. The labor demanded for all
other activities is variable and fairly independent. That
is, the work has to be done but there are no specific dates
for completion giving the farmer ease in programming these
activities, usually when he has some surplus labor (see
Figures 4 and 5).
The genetic constitution and basic husbandry of the cattle
are as described earlier; the animals originated from seve-
ral crosses between Holstein, Jersey, Guernsey, Brahman, and
Criollo. A dual-purpose system is used with one daily
milking in the morning, mostly with the calf's help (stimu-
lation). The calves stay with their mothers from milking
time until the early afternoon, when they are put in a sepa-
rate pasture Three groups are managed separately: the
oxen, the dry c~ows, and milking cows. The calves, steers,
and heifers are kept with the dry cows.
Seven hectares, representing 15% of the farmland and 56% of
the land in use, are employed as grassland. All of this
area is covered with low-grazing native species and is
divided into four parts. The fences are strung on live
fence- pos ts, consisting mainly of Madero negro (Gliricida
sp ium ); however Ery thr ina trees, Robusta coffee and a
variety of other species are used. The pastures include a
large amount of guava trees (Psidium guajava) that are used
for shade and firewood. Stocking rate averaged 0.92 animal
units/ha throughout the year Weeding took place once a
year, requiring 9.5 LD/ha.
A. Inputs: A total of US$34 was spent on veterinary
products. Most of this was used on external parasite con-
trol, which was done once a month for killing ticks (Boophi-
lus sp.) and larvae (Dermatobia sp). Internal parasite con-
trol is done twice a year, usually on all animals.
Common salt was given every 15 days giving a total of
US$2.10. Another input was the sugarcane tops, approxi-
mately 5 kg/day, which were consumed by the oxen during the
All labor comes from the family, totaling 107 LD, with an
opportunity cost equal to US$402 (see Figure 6).
J F M A M J J A S O N D
Fig. 4 Labor distribution throughout
main activities. Based on data
th~e year oc~cording to the forms
from CATIE -CllD
II Animal core
7 Milking an
J M M J S N
Fig.S~~~~~~~~~I Loo s hoghu h er01ctl/ot
agroecosystem. Bae ndt rmCTECl
Common salt U'SS2.oo
Vec.orodcts SS34oo AGROECOSYSTE1
Suear-cane tons 5kc/day during harvest time
Calves & nitss USS293.oo
Milk 19s4ke U~SS35.oo
Ueicth usin, 200k- USSl40.oo
Ox-work 22Shours USSL45.oo
Fig. 6. Schematic diagram of the cattle/pasture agiroecosystem. Based on
data from CATIE-CIID.
B. Outputs: During the year, 1944 kg of milk were
consumed by the family, representing US$435, equivalent to
50% of the system's PV. Four calves were sold, giving 34% of
the PV (US$293). Ox work, totaling 228 hours was another-
output, with an opportunity cost of USS445.
Weight gain, which is not considered a true output, is con-
sidered as a change of inventory. It was estimated at 200
kg with a value of US$140, representing 16% of the PV;
Sugarcane agroecosystem. Covering 4.9 ha (which repre-
sents 10.7% of the total area and 39% of the area actually
in use) sugarcane represents the farm' s largest activity
and generates 79% of the TPV (Table 12).
A. Inputs: A total of 592 LD was used; 393 family
and 199 hired LD. This represents 65% of the total labor-
used on the farm during the year. The labor distribution is
shown in Fig ure 7. Harvest time extends from January to
August, requiring an overall average of 25 LD/ha/month, withi
up to 62 LD/ha/month in the peak production month All
labor required for harvesting (199 LD) was hired because the
family labor was being employed for other activities on the
farm, including sugarcane replanting.
The herbicides were applied separately on the different
plots, three months after each was harvested. A total of
US$290 was spent on herbicide.
Other inputs included 194 hours of oxen work, with an oppor-
tunity cost of USS378, and 436 hours of tractor work costing
B. Outputs: The farm produced 436.4 tons of sugar-
cane (89 tons/ha), which totaled US$6,893. All production
was sold to the sugarmills. Another small output was the~
sugarcane tops that are eaten by the oxen (Figure 8).
Coffee agroecosystem. This farm has only 0.53 ha ini
coffee (1.15% of the total land); however, coffee generates
10.6% of the TPV and 17% of the NFI. The Cash Net Incore
(CNI) per ha is US$1,510, twice the amount generated by the
sugarcane and 40 times the amount generated by the cattle.
On the other hand, it occupies three times the labor used in-
sugarcane and 25 times that used for cattle per unit area,--
thus labor is the main constraint to the area cultivated.
Regional labor during harvest time is scarce and expensi-;e,
even though it is paid at a fixed rate according to M
amount of product harvested. The farmer was establishing a
nursery to expand his coffee plantation.
C~All other activities( (replanting, weeding etc. )
J F MA MJ JAS ON D
Fig. 7 Use of labor in lhe sugar coner ag~roecosystem
throughout the year.
Based on data from CATIE-CIID
Labor 592LD USS2 925 oo
Herbicides. 50kmr- USS290.oo
Ox-work, 190hours USS378.oo.
Tractor 436bours USS1,684.00
Suepr-cane ;36.4T. USS6.893.oo
Fuelwood from live fence costs
Fig. 8. Schematic diagram of the Sugar-cane agroecosystem. Based on data
from CATIE-CT ID
AP. Inputs: The most important input: is labor; a
total ?E-203 LD were required--193 family LD and 10 hired
LD, with aiE total cost of USS762 (Figures 9 and 10). Other
inmputs wer~e 3 kg~9 fertilizer (US$4.50), 2500 plastic bags for
gnrowPing n~ew~ trees (US~-~344) 5.4 kg pesticides (US$18), and 34
hours ox work (US$66).
B. Ou tpu ts: T rhe main output con~sists of 9 fanegas
(one fanega = 258 kg) coffee beans, totaling UBS$893. Other
outputs are 52 banana bunches (US$30) and firewood, collec-
ted after the coffee bush pruning done in February.
Forest and thickets. A large part (33 ha or 73%) of
the farmland is in thickets, with a small proportion in for-
est. Probably the largest constraint in working this land
is labor, even though capital may play a decisive role. The
thickets provide firewood but do not have any other apparent
There is no precise quantitative data about this component
of the system since it was not included in the CATIE-CIID
report (CATIE, 1979), and the farmer does not consider it an
active component of his farm.
Home garden. As described previously, this component
is managed mainly by the females of the family and does not
require major inputs even though it provides small but con-
stant food outputs.
Poultry Poul try pr imar ily prov ide eggs and ,
occasionally, meat. A total of 12 LD were
invested throughout the year (US$45 opportun-
ity cost). The feed consists basically of
leftovers and somne of the home-garden crops.
The poultry run loose so that they can com-
plete their diet with insects, worms, etc. A
total of 730 eggs were produced in one year
(US$51), providing a net income of US$3.99.
Crop component. The area dedicated to this com-
ponent has irregular boundaries; basically it
consists of the area around the house. There
is a variety of food crops (cassava, chayote,
squash, malanga dasheenn], bananas, plan-
tains) and fruit trees (citrus, avocado,
coconut, papaya, guavas).
Very little labor is allocated to this component (mostly for
weeding and planting). As with the forest/thicket agroeco-
system, this component was not included in the data collec-
ted by CATIE-CIID, and is very hard to quantify because of
the informal nature of the system, its complexity, and the
relatively little importance given to it by the farmer.
~3Soll preporotion for nursery
SPr uning, fertilizng, weeding~, others
J FMA MJ J A S ON 0
Fig.9 Labor distribution throughout the year on
the coffee crop activities.
Bosed on doto f rorn CATIE -CIID
Labor 203LD USS762.oo
Fertilizer 8kg USSA.50
Plastic baes, 25000. USS4L.oo
Ox-work 34hours USS66.oo
Coffee 9.03 fancas1 gSS 899.0
Bananas, 52 bunches USS30.oo
Pesticides 5.4ke USSIS.oo
Fig. 10. Schematic diagram of the coffe/banana agroecosyst-em
1 faneca = 258 kg
Interaction Among Crop and Animal Activities
In summary, there is very little interaction between compo-
nents on this type of farm. Labor availability is probably
the most limiting factor and represents the area of greatest
comnpet i tion Competition for land is, in theory another
important constraint; however, considering that close to 70%
of the farm's land 'is not in use, it can be concluded that
for this particular farm and almost all of the medium and
large ones in this region, land is not a major constraint.
Most of the farm's better areas are now in use, and not all
the land is potentially useful.
Capital is a much needed resource; all of the farm's activi-
ties could increase outputs if larger amounts of inputs were
put into these systems.
Basically, all components have research opportunities, but
the small farmers have shown that they are, generally speak-
ing, making "maximum" use of their resources. This "maxi-
mum" is not specifically an economical or productive maxi-
mum, but includes a series of other factors, such as per-
sonal preference or needs, social belief, etc. This finding
does not suggest that a great deal cannot be done,
especially with the introduction of "new" crop/animal pro-
Probably very little can be done to improve the actual farm-
ing systems unless new concepts are used and the farm is
visualized as a complex unit of production.
New activities that will not compete greatly with use of
labor or capital would probably be accepted by most farmers
and slowly could change a whole r e gion 's prod u ct ion pro-
cess. However, the introduction of more profitable systems
that would require the farm's main resources must be well
tested before farmers try them out on a commercial scale.
Avila M. (n.d.) La importancia y productividad de cultivos
perennes en Centroamerica y especificamente en cuatro
cantones de Costa Rica. Unpublished paper. 18 p.
Avila, M. and R. Serpa. 1981. El papel comnplementario de
los cultivos en los sistemas de pequenas fincas
ganaderas en Costa Rica In : Programa Cooperativo
Centroamericano para el Me joram ie n to de Cultivos
Al imenticios, Resumenes, 27 a. Reunion Anual M~arch,
1981. pp 163-164. Santo Domingo, Dominican Republic.
Avila, M. and R. Tremini~o. 1981. Optimizacion economic de
sistemas mixtos tipicos en cuatro regions de Costa
Rica. In Pr ogr ama Cooperativo Centroamnericano de
Cultivos Alimenticios; Resumenes, 27 a. Reunion Anual,
March, 1931. pp 226-227. Santo Domingo Dominican
Beer, J. W. et al. 1979. A case study of tr additional
agroforestry practices in a wet, tropical zone; the La
Suiza Project. CATIE, Turrialba, Costa Rica.
CATIE. 1981. Caracterizacion de sistemas agricolas; Inti-
buca, Honduras Serie Materiales de Ensenanza No 6.
CATIE, Turrialba, Costa Rica. 84 p.
CATIE. 1979. Proyecto CIID-CATIE. Unpublished working
Costa Rica. 1974. Direccion General de Estadistica y Cen-
sos. Censo Agropecuario 1973. San Jose, Costa Rica.
Costa Rica 1977. Information basica del sector agro-
pecuario de Costa Rica. Oficina de Planificacion
Sec tor ial Agrope cuar ia I ICA San Jose Costa Rica .
Costa Rica. 1978. Asociacion de subgroups de suelos (M~apa
preliminar); Limon. Scale 1:2000,000. Color. Oficina
de Planification Sectorial Agropecuaria, San Jose,
Costa Rica. 1976. Diagnostico del sector agropecuario de
Costa Rica; 1962-1976. Oficina de Planification
Sectorial Agropecuaria, San Jose, Costa Rica.
Guillen, R. and M. Avila. 1981. Relacion economic entire
los components agricola y pecuario en las pequenas
fincas ganaderas de Costa Rica. Desarrollo Rural en
las Americas 13(1):11-20. 1981.
Ibarra, A. E., et al. 1979. Inventario de recursos; area
del Programa de Diversificacion de Turrialba. Turrial-
ba, Costa Rica. IICA. Publicacion miiscelanea no. 62.
Locatelli, E. 1980O. Al terna tivas te cnolog icas para el
mejoramiento de la production y productividad de los
sistemas usados por el productor de escasos recursos.
In: Seminario Latinoamericano sobre Mejoramiento de la
Productividad de Pequenos Productores en el Desarrollo
Rural. San Jose, Costa Rica. 10 p.
Oficina del Cafe. 1982. Informes sobre la actividad cafe-
talera de Costa Rica. San Jose, Costa Rica. 98 p.
Rockenbach, O. G. A. Enriquez, and R. D. Hlart. 1981.
Estudio del componente de plants perennes en un sis-
tema tipicamente lechero de Turrialba, Costa Rica. In:
programna Cooperativo Centroamericano de Cultivos Ali-
menticios, Resumenes, 27 a. Reunion Anual, March,
1981. pp 216-217. Santo Domingo, Dominican Republic.
Rockenbach. 1981i. Analisis dinamico de dos sistemas de
finca predominantes en el canton de Turrialba, Costa
Rica. M.S. Thesis. Universidad de Costa Rica-CATIE,
Turrialba, Costa Rica. 175 p.
Sancho Mora, F. 1981. Unpublished paper. Turrialba, Costa
Serpa Valles, R. and M. Avila. (n.d.) Evaluacion economic
de algunas fincas de Costa Rica con diferentes sistemas
de production. Unpublished paper. 27 p.
Soley Monge, A. 1 978 Adm i ni str a cion de explotaciones
ganaderas en Costa Rica Editorial Costa Rica, San
Jose. 162 p.
Tremninio, R. and M. Avila. 1980. Analisis comparativo y
optimizacion de fincas en Costa Rica. Unpublished
paper. Turrialba, Costa Rica. 17 p.
Ugalde A., L. A. 1979. Descripcion y evaluacion de las
practices agroforestales en las Cuenca Piloto de La
Suiza, canton de Turrialba. UNU-CATIE, Turrialba,
Costa Rica. 31 p.