I-. *V Foo and Agiul
I Unvriyo heVri sad Agrclua Exeim n Staio V~ 6, 199
2 New Ornamental Pot Crops
3 Rapid and Uniform Papaya Emergence with Primed Seeds
Thomas W. Zimmerman
6 Evaluation of Three Indigenous Caribbean Finfish for Culture Potential
Bill Cole and Kurt Shultz
10 Environmental Influences on Reproduction and Milk Production of Holstein Cows in St Croix
Robert W. Godfrey and Peter J. Hansen
13 Mahogany Response to Water Stress
15 Preservative Characteristics of Guineagrass/Leucaena Silage Compared to Sorghum Silage
Martin B. Adjei and S. Josephat
18 Improving Guineagrass Forage Feeding Value by Urea Treatment
Martin B. Adjei and Terry J. Gentry
22 Strategies for Increasing Yam Production in the Virgin Islands
S.M.A. Crossman, C.D. Collingwood, M.C. Palada and J.A. Kowalski
25 AES Personnel and Current Research Projects
26 Recent Publications
President, University of the Virgin Islands
Darshan S. Padda,
Vice President for Research and Land-Grant Affairs
and Experiment Station Director
James E. Rakocy,
Robin Stems, Editor
UVI Research is published annually by the University of the Virgin Islands Agricultural Experiment Station. RR 2,
Box 10,000, Kingshill, St. Croix, USVI, 00850.
Contents of this publication constitute public property. The written material may be reprinted if no endorsement of a
commercial product is stated or implied. Please credit the authors, the University of the Virgin Islands Agricultural Experiment
Station and UVI Research.
To avoid overuse of technical terms, trade names of products are occasionally used. No endorsement of these products
or firms is intended, nor is criticism implied of those not mentioned.
The University of the Virgin Islands is committed to the policy that all persons shall have equal access to its programs.
facilities and employment without regard to race, religion, color, sex, national origin, age or veteran status.
A number of colorful plants can be found growing
throughout the landscape in the Caribbean, like Hibiscus, Bird
Pepper and Christmas Snowflake (See UVI Research 4:17-
19). Others are relatively new, like Mussaenda from the
Philippines and Africa and the Lipstick plant from Central
and Tropical America.
But these plants all have one thing in common. Each can
now be produced as an ornamental pot crop for the Virgin
Islands and the Caribbean employing techniques such as
pruning and the use of plant growth regulators (PGR).
Recent research at the Fruit and Ornamental section of UVI-
AES has demonstrated that many common and not-so-common
garden species can now be tailored for pot crop production, adding
to the list ofpotted plants for the home and augmenting the number
of crops local nurseries can now offer to the consumer.
Most tropical species actively grow throughout the year
due to the absence of the climatic changes that limit temperate
species. This year-round growth often makes it difficult to
grow ornamental species in pots where they may have added
attraction by being easily moved around the land- or interiorscape.
Until now, continuous pruning was the only technique for
adapting such plants to a pot environment. But this also meant
continually destroying the stem tips where blooms are formed
as in Hibiscus and Christmas Snowflake or colorful and edible
fruits as in the Bird Pepper. A unique group of PGRs, called
plant growth retardants (GR), have now made it possible to
overcome this problem. By physiologically slowing growth
through the inhibition of natural growth promoters, GRs (when
applied in very controlled amounts to many potted plants) inhibit
vegetative growth while either maintaining or promoting flower
production. Hence, a much smaller but more floriferous plant
adapted to a pot environment often results from GR treatments.
An example is Mussaenda, also called 'Satin Plant,' a
member of the Coffee family Rubiaceae. It is a spectacular
flowering shrub grown for its showy bract-like sepals ranging
from the deep 'Ashanti Red' of Africa to the 'Snow White' of
the Philippines and the pink of the hybrid Dona Luz cultivar.
The plant is covered with a delicate pubescence that imparts
a soft satiny texture and has shiny deep green leaves, hence
its common name.
Mussaenda grows well in pots larger than two inches in
diameter and can be kept as a patio plant in cool to moderate
climates but grows well as a small shrub in warmer climates.
In Southeast Asia, Hawaii and south Florida Mussaenda is
an established outdoor ornamental. It has an indoor postharvest
life of three to four weeks or longer.
Untreated plants bloom in one to two years in the
landscape, after which they must be pruned back for further
growth and flowering. Mussaenda will not tolerate high-pH
soils, thus severely limiting its culture in calcareous islands
like St Croix. However, started from a soft wood cutting,
Mussoenda can be forced to flower in a six-inch pot within
five months through the application of 2500 ppm sprays of
the GR Alar or B9. After about two months of flowering,
the plant can be pruned, repotted and similarly forced again
within the next six months, thus producing its attractive
blooms twice in one year. Because they are non-
photoperiodic, they make an excellent flowering pot crop
at virtually any time of the year. Additionally, they are
almost pest-free and could therefore become a good
substitute for poinsettias, which are highly seasonal and
very susceptible to pests.
At UVI-AES, Mussaenda is just one of the many native
and exotic flora species being investigated for their potential
as new pot crops. Many will augment the arsenal of the
nurseryman's range of crops, add color and variety to the
local landscape and make potentially valuable import
substitutes to help improve the economies of many
Caribbean Basin island states.
This research was supported in part by the Caribbean Basin
Advisory Group (CBAG) Project No. 91-34135-6173.
Rapid And '1 PT !'"I
Thomas W. Zimmerman
Papaya seeds have been attributed with poor
germination in the propagation of papaya. Poor germination
has been associated with growth inhibitors present in the
sarcotesta, the gelatinous membrane surrounding the seed,
as well as the seed coat itself. Seed priming is a system of
soaking seeds in a solution for a given period of time prior
Seed priming is commercially used to reduce
germination time and increase the uniformity of the seedling
stand (see the home recipe for priming salts on page 6).
Uniform plant emergence allows for the full growth
potential of every seedling. When seedlings emerge unevenly
over a period of time, the first to emerge shade the latter
emerging seedlings and develop ahead of them.
Seed germination is divided into three phases:
imbibition, lag phase and radicle emergence. Imbibition is
the physical uptake of water by a seed. Imbibition is not an
indication of seed viability because even dead seeds will
draw up water and swell. The lag phase is the time during
which the cell membranes are repaired and the metabolic
processes in the cell are reinitiated. Radicle emergence
occurs when the root protrudes from the seed due to cell
elongation and cell division.
The purpose of seed priming is to promote the first two
phases of seed germination, imbibition of water and
restoration of cellular biochemical activity-but to inhibit
the final phase- radicle emergence. Once the radicle has
emerged, seeds are considered germinated and dehydration
is lethal. Either salt solutions or water can be used for
priming. When using water to prime seeds, priming should
be terminated before radicle emergence. The osmotic
potential is the key factor in developing salt solutions for
seed priming. Osmotic potential,measured in megapascals
(MPa), is an indication of how strongly a solution will hold
water and inhibit it from being absorbed by the seeds.
This is different from molar concentration which
indicates the amount of salt dissolved in a solution. An
osmotic potential of-I MPa will allow proper seed priming
and yet inhibit radicle emergence. The objective of this
study was to use K or Ca salt solutions to develop a system
for reducing the time required for papaya seed germination
and seedling emergence. Seedling emergence was
determined when both the hypocotyl and the seed cotyledons
broke through the potting mix surface.
Seeds were collected from mature fruits of four open-
pollinated papaya varieties, 'Cariflora', 'Puerto Rico Red',
'Solo 64' and 'Waimanalo 162'. Seeds were washed to
remove the gelatinous sarcotesta and any floating seeds
were discarded. Floating seeds often contain aborted
embryos or under developed embryos that are nonviable.
Cleaned seeds were air-dried and stored in a refrigerator at
56C until used. The priming solution treatments were
developed to be -1 MPa as determined by a Decagon
The salts chosen for priming solutions were: CaCI, (173
mM), Ca(NO,), (173 mM), KCI (232 mM) or KNO3 (232
mM). A distilled HO and a control (no priming treatment)
were also included. Twenty-five seeds were placed in each
half of 100 x 15 mm divided petri dishes to which 8 ml of
priming solution was added. Each treatment was replicated
four times. Seeds were primed for five days at room
temperature (200C) with a change of fresh solution after
so- X cU
6 7 5 9 10 11 12 13 14 15
Figure 1. Papaya plant emergence over time, oombining rCaora, S-64,'and
Wainanalo.' as influenced by e salt solutions used In seed pming. The 1PR
Red data was signitfcanydlfemt (P 0.05) fom the otdr three vadeties arnd
6 7 8 9 10 11 12 13 14 15
Figure 3. The percentage of total plant ei genmce per day for the four papaya
Seed priming was shown in this study to enhance the total seed
germination and seedling stand establishment for slow germinating varieties
or varieties with reduced viability. The benefit of seed priming was to reduce
the average seedling emergence time and Increase the uniformity of plant
48 hours. After five days of priming, seeds were rinsed
with distilled water and air dried two days prior to planting
under greenhouse conditions in mid April at a I cm depth
in 1:1 (v/v) Pro Mix : sand potting mix. Emergence was
recorded when the hypocotyls forced the seed cotyledons
above the potting mix surface.
During imbibition the cell membranes have not yet been
repaired from the damage incurred during dehydration of
the mature ripening seed. The cells are therefore leaky and
cell contents which dissolve with the uptake of water can
be lost from the cell until the membrane is repaired. The
loss of cell contents or electrolytes can be measured by
Conductivity readings provide an indication of the seeds
passing from imbibition to the lag phase. The conductivity
from all the priming solutions were the highest after the
first two days and decreased by the fifth day of priming
(data not presented). This indicated that the seeds were
viable and able to repair the cell membranes upon imbibition
of water. Dead seeds are unable to repair damaged
membranes caused by dehydration and would have
continuous electrolyte leakage over time. The conductivity
readings indicate that the priming solutions did allow
imbibition and activation of cellular membrane repair.
The effect of priming solution on total plant emergence
over time was similar for all salt solution and significantly
different from the control over time. The salt solutions used
for seed priming provided better emergence over time than
the water treatment or the control. While plant emergence
began on the seventh day for all priming treatments, the
start of plant emergence was delayed until the tenth day
for the controls (Figure 1). All priming treatments
significantly reduced the average emergence time over the
controls for all four varieties. Priming solutions containing
either KNO, or Ca(NO,), had the greatest overall effect of
reducing the average emergence time (Figure 2).
The varieties selected did provide a range of response
for plant emergence. Seedling emergence began on the
seventh day and leveled off for 'Cariflora', 'Solo 64' and
'Waimanalo' by the twelfth day. Total plant emergence on
the fifteenth day among these varieties ranged between 85-
95 percent and was not significantly different. Total 'PR
Red' emergence was lower at 70 percent but stabilized by
day 15. The greatest total daily emergence for 'Cariflora',
'S-64' and 'Waimanalo' was on day 8, while 'PR Red'
was similar for both day 10. The greatest average daily
emergence for 'Cariflora', 'Solo 64', 'Waimanalo' and 'PR
Red' was 31.4, 29.2, 26.4 and 17 respectively (Figure 3).
These data presented are averages of the priming
solution treatments without the control. Seed priming with
one of the four salt solutions increased the total plant
emergence for 'Cariflora', 'Waimanalo' and 'PR Red' than
was possible with priming in water or no priming at all.
The priming solutions had no effect on the total germination
of 'Solo 64' (Figure 4).
Seed priming was shown in this study to enhance the
total seed germination and seedling stand establishment for
slow germinating varieties or varieties with reduced
carltflka 844 Walmmnlo PR Red
Figure 2. Average plant emergnce as influenced by papaya vadely and seed
priming beatnwt DW0 mnt etten Indicate significant diff ces between
bement within a variety. Mean aeparaion test based on LSDP-0.05
844 Cardfon Wasnum lo PR Red
Figure. Total papaya plant emergence by variety teen days aferplantgr as
Mnutenod by seed pricing mtamnt
viability. The benefit of seed priming
was to reduce the average seedling
emergence time and increase the
uniformity of plant emergence.
Priming papaya seeds in water or
one of the four salt solutions
examined benefited seedling stand
establishment.One of the beneficial
effects of seed priming may be to
leach some of the plant growth
inhibitors from the seed coat and
internal seed tissues.
Using a salt solution for seed
priming of some papaya varieties
can enhance the performance of
seedling stand establishment over
that obtained by water priming or
The uptake of the nutrient salt
during priming may increase
metabolic activity during the priming
process which can stimulate low
vigor seeds to germinate. Pretreating
papaya seeds with a salt solution can
be used to enhance seed germination
and provide uniform plant emergence
in a shorter length of time than
Priming papaya seeds in a KNO,
solution resulted in the greatest
seedling emergence in the shortest
length of time. Seed priming can be
effectively used to promote better
papaya seedling stand establishment.
This research was supported in part
by Caribbean Basin Advisory Group
(CBAG) Grant No. 94-34135-0280.
From the kitchen of...
A Home Recipe for Starting Papaya Seeds
Seed priming can easily be done in a home or small-scale gardening situation:
/Put up to 50 mature and dried papaya seeds in a sandwich bag (Ziploc Is best) or a small
/Add 114 cup of bottled water or make a fertilizer solution made of 1 teaspoon CaCI, or Ca(Noj),
per cup of water or 1/2 teaspoon of KC1 or KN0, per cup of water). These salts are found in the
commercial "Miracle Grow" plant food.
/This should be enough solution to keep the seeds wet without submerging them. Seeds are living
and require air, so be sure not to completely submerge them or they will drown.
/Seal the bag but leave as much air in the bag as possible.
/The solution will become dark from the tannins In the seed coat.
/After two days, pour out the old solution and replace it with fresh, clear solution.
/Allow the seeds to soak an additional three days. Now they are ready for planting.
/The seeds may be planted 1/4 to 1/2 inch deep directly Into pots or trays containing a loose
potting soil with good drainage. Water the planted seeds and keep them in an area above 80' F
during the day (you'll get very little or no plant emergence at temperatures below 75' F).
/Seeds should start coming up in a week.
Bill Cole and
Worldwide fish harvests increased
dramatically in the years following
World War II, primarily due to
improved fishing technology. Annual
yields began to taper off in the late
1970s and it became evident that the
ocean's fisheries resources were not
inexhaustible. For the past several
years the annual catch has hovered
around 100 million metric tons (mt), a
figure which many fishery biologists
consider to be the ocean's maximum
sustainable yield (MSY). MSY is the
harvest of a particular stock or stocks
of fish over time without causing a
decrease in abundance. Today 13 out
of 17 of the world's major fisheries are
depleted or in serious decline. The
remaining four are either fully utilized
or over-exploited. Total catch has
remained constant because of shifts in
species composition that are harvested.
The reasons for the decline in stocks
include over-fishing, habitat
destruction, pollution and climatic
The majority of the Caribbean Sea
consists of deep, nutrient-poor waters
that are characterized as a biological
desert. Many islands are surrounded
by shallow, narrow shelves, that, while
productive, are extremely sensitive to
over-fishing. Landings in the U.S.
Virgin Islands have been declining
since the late 1970s in spite of
increased fishing effort. Despite the
small human population relative to
surrounding sea, Caribbean islands
rely on imports to meet demand. A
study published in 1984 found that
annual seafood demand in the
Table 1. Composition of diets fed to juvenile white grunts for ten weeks. Protein, fat fiber and ash are expressed as a percent of dry weight.
CIET %PROTEIN %FAT %FIBER %ASH %MOISTURE
Dry, salmonidk 43 15 4 12 10
Semi-moist. salmonid 43 15 3 11 18
Madrine finfh 55 11 2 11 7
Non-formulated" 82 8 2 7 80
Table 2. Composition of diets fed to juvenile schoolmasterfor ten weeks. Protein, fat. fiber and ash are expressed as a percent of dry weight.
DIET %PROTEIN %FAT %FIBER %ASH %MOISTURE
Dry, salmonid 44 15 3 7 11
Semi-moist, salmonid' 43 15 2 7 21
Marne ffiisl 55 11 2 11 7
Non-formulated* 72 5 2 13 78
Table 3, Composition of diets fed to palometa for 16 weeks. Protein, fat, fiber and ash ar expressed as a percent of dry weight.
DIET %PROTEIN %PA %FIBER %ASH %MOISTURE
Drysalmonid 48 16 2 8 10
Semi-moist, salmonkld 44 16 3 9 21
Floatng, salmonid' 41 11 2 13 10
Marinefinfisht 61 15 1 8 10
*Moor-Clark Co., Ic., LaConner, WA.
'Formulated by John Tucker. Harbor Branch Oceanographic Institution. Fort Pierce, FL, and manufactured by ZelgerBrothers, Inc., Gardners, PA.
'Composed of 45% fish, 45% shrimp and 10% squid. The Ingredients were ground, mied and supplemented with vitamins and minerals.
OBioProducts, Inc., Warrenton. OR.
*Composed of 70% fish, 20% shrimp and 10% squid. The Ingredients were ground, mixed and supplemented with vitamins and minerals.
Integral Fish Foods, inc., Grand Junction, CO.
Caribbean was approximately 775,000
metric tons (mt) which greatly
exceeded not only the fish landings
(88,947 mt), but also the estimated
potential yield (193,785 mt). Local
fishery landings account for
approximately 30 percent of seafood
consumed in the U.S. Virgin Islands.
If seafood supply is to be increased,
other methods of production must be
Mariculture, which refers to the
culture or farming of marine organisms,
has the potential to increase seafood
production, provide economic
diversification, create employment,
reduce dependence on fish imports and
generate foreign exchange through
exports. With careful planning and
management, stock enhancement
programs may be used to reestablish
fisheries where over-fishing has
occurred. The Caribbean has excellent
possibilities for mariculture, but
development is restricted by a number
of factors including the lack of technical
and biological information. For example,
insufficient information has been
gathered to describe growth, feed
conversion and survival of indigenous
finfish under culture conditions.
The University of the Virgin Islands
Agricultural Experiment Station (UVI-
AES) initiated a program to evaluate
the culture potential of several near-
shore Caribbean marine finfishes. In
the first phase of the program, feeding
trials are used to evaluate growth
performance, feed conversion
efficiency and survival ofjuveniles fed
Three species have recently been
evaluated: white grunts (Haemulon
plumieri), schoolmaster snapper
(Lutjanus apodus) and palometa
(Trachinotus goodei). The studies
were conducted in a flow-through
facility consisting of 12 2-m3
fiberglass tanks. Water was pumped
directly from the ocean using a 1-
horsepower (hp) pump. Each tank was
covered with 80 percent shade cloth
and a 1/2-hp air blower provided
Fish captured from the wild using
traps and seine nets were trained to eat
a formulated (pelleted) diet by
incorporating the formulated feed with
a fish/shrimp/squid mixture. Over
several days the percent of formulated
feed was increased in the daily ration
until the fish consumed 100 percent
pellets. The fish were stocked in the
tanks at a rate of 5 fish/m3 and fed
four diets for 10 to 16 weeks.
Treatments were randomly assigned to
tanks and replicated three times. Diet
types included commercially-available
salmonid diets, experimental diets
formulated specifically for warm-
water marine finfish, and non-
formulated diets consisting of fish,
shrimp and squid plus vitamins and
At the conclusion of each
experiment growth rates, feed
conversion ratios (FCR), survival and
condition factors (CF) of fish in each
diet treatment were compared.
Absolute growth rate (g/d) was
calculated by dividing weight gain (g)
by time (days). Specific growth rate
was determined by the formula:
Specific growth rate (%/d) =
(Ln W2 Ln WI / T) X 100
where WI was the initial mean weight
(g), W2 was the final mean weight (g)
and T was time (days). FCR values
were obtained by dividing the total
amount of feed administered (g) by the
increase in fish weight (g). CF was
calculated by the formula:
CF = 105 X W2 / TL23
where TL2 was the final mean total
Mean values for growth rates,
FCR. survival and CF were compared
by one way analysis of variance and
pairwise multiple comparisons were
made using the Student-Newman-
Keuls method. Differences among
treatment means were considered
significant at the 0.05 level of
Juvenile white grunts (mean initial
weight = 53 g) and schoolmaster (mean
initial weight 53 g) were fed three
formulated, sinking diets and a non-
formulated diet (Tables I and 2) for
10 weeks. Feed was administered once
daily at rates of 10 percent of biomass
for grunts and 3-4 percent of biomass
for schoolmaster. Fish were sampled
and feed rations adjusted accordingly
by weighing all fish from each tank at
two-week intervals. A second eight-
week experiment was conducted using
the semi-moist diet to compare growth
of schoolmaster (mean initial weight
= 78 g) fed once per day with those
fed continuously during daylight
hours. Continuous feeding was
achieved using spring-operated belt
feeders. Palometa (mean initial weight
= 85 g) were fed four formulated diets
(Table 3) for 16 weeks. Feed was
administered four times daily, initially
at a rate of 5 percent of biomass which
was gradually reduced to 2 percent of
biomass. Sampling occurred at four-
Grunts fed the dry salmonid diet
had significantly lower growth rates
and a significantly higher FCR than
fish fed the other diets (Table 4). The
FCR of grunts fed the non-formulated
diet was significantly lower than fish
of the other treatments. There were no
significant differences in condition
factors for grunts fed each diet. In the
first schoolmaster experiment, fish fed
the non-formulated diet had the highest
growth rates and the lowest FCR
followed by fish fed the marine finfish
diet, the semi-moist salmonid diet and
the dry salmonid diet (Table 5).
Pairwise comparisons of growth rates
and FCRs indicated significant
differences among all treatment means
except between the two salmonid diets.
The final condition factors were nearly
identical for fish fed each diet. In the
second schoolmaster study, there were
no significant differences in growth
rates among fish fed once per day at 4
percent of biomass (0.24 g/d, 0.26 %/
d), fish fed continuously at 4 percent
of biomass (0.33 g/d, 0.38 %/d), and
fish fed continuously at 8 percent of
biomass (0.27 g/d, 0.32 %/d). Absolute
growth rate, specific growth rate, FCR
Table 4. Mean values for absolute growth rate, specific growth rate, feed conversion ratio and condition factor of juvenile white grunts fed three
formulated diets and a non-formulated diet for ton weeks. For each column, values followed by the san letter are not sigrnAlcanty different (40.05).
DIET ABS1OLUTMGROWH SPECFICGROW FEED CONVERSION CONDITION SURYWVALN
RAtTEE" (RATE(A RATIO FACTOR
Dry, saknonid 0.20a 076a 10.9a 1.84a 100a
Seni-moist salmonid 0.27bc 0.94b 8.6b 1.93a 100a
Marinefinftsh 0.29b 1.02b 8,4b 1.92a 10Oa
Non-formulated 0,28c 0.92b 2.0c 2.13a 10D
Table S. Mean values for absolute growth rate, specific growth rate, feed conversion ratio and condition factor of juvenIle schoolmaster fed three
formulated dalts and a non-formulated diet forten weeks. For each column. values followed by the same letter are not significantly different (p<0.05).
DIET ABBOLUTEGROWTH SPECFICGROWN FEEDCON EROSION CONDITON SURVIVALf)
RATEWgl) ,RATE(%M, RATIO FACTOR
O y. sakurnid 0.2Ba 0,45a 7,9a 1.67a 100a
Semi-mot, salmonid 0.30a 0.48a 7,4* 1.6Ia 100s
Marine finfh 0.48b 0.71b 5.1b 1.74b 100a
Non4ornulted 0.658 0.81c 3,5c 1.74b 100a
Table t Mean values for absolute growth rate, speflc growth rate, feed conversion ratio and condition factorof plomet fed four formulated diets
for 16 weeks. Foreach column, values followed by the same letter are not significantly difrent (p<0.05),
oDE ABSouruEGFwm SPECFICGROWHN IE CONVERSION CONDmON SURVIVAL.(%
RATE(g) RATE(%AI RA11D FACTOR
ODy.saluonid 2,7a 1.36a 215e 2.93ab 93a
SemWnloit, aalmonid 2.85a 1.39a 2.13a 3.04a 1000
Floating, sanmonid 2.7a 1.83a 2.1a 2,83b 97a
Marinetfirsh 2.Bla 1,390 2.16a 3.0Sa 100a
,looreClsk Co., Inc., LaConner. WA.
Fonmlated by John Tucker. Harbor Branch Oceanograptic Inluition. FortPree, FL, and manufactured by Zigler Brother, Inc., Gardners, PA.
*Composed af45% fish, 45% shrimp and 10% sqid.The ingredients were ground, mixed and supplemented with vitamins and minerals.
BioProducts Inc., Warenton, OR.
*Cmposed of 70%fih, 20% shrimp and 10% squd. The ingredients weaground, miad and supplemented with vamins and mineral.
Integral Fish Foods, Inc., Grand Junction, CO.
0 Z 4 S I 10 t t14 16
Thn t C1a
Figure 1. Mean weights of white grunts,
schoolmaster and palometa fed semi-moist
salmnid diets for 10 or 16 weeks.
Although both white grunts and
schoolmaster snapper responded well to
handling and no diseases were observed,
growth rates obtained in these studies
were lower than the range desired for
commercially-cultured finfish. This is
also true for the f&ed conversion ratios
obtained for both species, with the
exception of grunts fed the non-
formulated diet Continuous feeding of
schoolmaster did not significantly
improve growth. Absolute growth rates,
which tend to increase with fish size,
decreased throughout each schoolmaster
study. A better understanding of these
species nutritional requirements and
behavior in confined conditions may
improve growth and feed conversion.
These studies indicate that white grunts
and schoolmaster are not suitable
candidates for food fish culture under
these experimental conditions,
Although direct comparison between
species can not be made due to variation
in diets and initial weights, Figure 1
illustrates increases in mean weight for
grunts, schoolmaster and palometa fed
semi-moist diets of similar composition.
Growth rates and feed conversion ratios
exhibited by palometa fed each diet
compare favorably to commercially-
cultured finfishes. Palometa fed the
lower protein salmonid diets grew as
rapidly and had similar feed conversion
as those fed the high-protein marine
finfish diet The floating diet was readily
accepted. Mortalities were due to
parasitosis (Neobenedenia melleni).
Parasites were controlled by a one-hour
bath treatment of 167 ppm formalin and
subsequent three-minute dips in
freshwater at four-week intervals. This
parasite has been problematic with
other cultured warm water marine
species and practical methods of
control will have to be developed for
Juvenile feeding experiments are only
one step in UVI's program to evaluate
the mariculture potential of indigenous
finfish Species, such as the palometa,
which appear to have mariculture
potential, require further research.
Grow-out systems (i.e., tanks and cages)
need to be evaluated as well as spawning
and larval rearing techniques. Nutritional
and environmental requirements of
potential species have to be determined.
Practical disease prevention and control
methods must be developed before
commercial mariculture is possible.
Cultured seafood production will
increase in the future because of the
shortfall of ocean fisheries, an increasing
world population and developments in
marculture technology. Total aquaculture
(freshwater and marine) production in
1990 was 12.1 millionmetrictons (nt)and
the projected aquaculture productioninthe
year 2000 is between 20 and 22 million
mnit More than 50 percut of aquaculture
production comes from brackish and
marine environments. The Caribbean,
which has excellent potential for
mariculture, could play a role in future
seafood production. To help meet these
needs, UVI-AES will continue to pursue
This research was supported in part
by the U.S. Department of Agriculture
under Hatch Project No. 0220530.
Of Holstein Cows
On St. Croix
Robert W. Godfrey and
Peter J. Hansen -
The dairy industry is one of the major animal-related
agribusinesses in the U.S. Virgin Islands, but the industry
has been unable to consistently meet local demand. A
consistent supply of fresh milk for consumption or
processing into dairy food items could minimize the use of
imported or reconstituted milk for human consumption in
Several factors, including nutrition, genetics and the
environment can influence milk production and reproductive
efficiency of dairy cattle. Dairy farmers in the Caribbean
have the least amount of control over the environment when
compared with nutrition or genetics.
Two components of the Caribbean environment have
high potential to influence dairy cattle. The first is ambient
temperature. The ability of cattle to maintain
thermoneutrality is more difficult when they are exposed
to elevated ambient temperatures. When ambient
temperature rises above the upper limit of the thermoneutral
zone (80 F) it is more difficult for cattle to maintain their
normal body temperature (10 10F). When this occurs the
animal will experience some degree of heat stress.
The second environmental component is solar radiation.
Studies have shown that dairy cattle in sub-tropical and
tropical environments have depressed reproductive and
productive traits. This suppression was most noticeable
during the warmer months of the year. The amount of solar
radiation plays a major role in determining the ambient
temperature, and may also be involved in suppressing the
reproduction and production of dairy cattle.
It has been known for many years that darker colors
absorb more solar radiation which leads to elevated
temperatures of dark colored objects. Recently, there has
been interest in determining the relationship between hair
coat color and production and reproduction in Holstein cows
in hot, humid climates with high amounts of solar radiation,
based on the theory that darker cows will absorb more solar
radiation which could lead to elevated body temperatures
and increase the incidence of heat stress.
The objectives of this study were: 1) to evaluate if there
are seasonal influences on reproduction and milk production
of Holstein cows on St. Croix, and 2) to determine if there
is a relationship between percent black hair coat and milk
production or reproductive traits of Holstein cows on St.
Croix. The project was conducted in retrospect by
examining herd records from a commercial dairy farm
covering the period from 1960 through 1986. The herd
consisted of both registered and commercial grade
Holsteins. Information obtained directly from the herd
records included milk production during the first lactation
(lb), length of first lactation (days) and calving dates.
Calving interval was calculated as the number of days
between consecutive calving dates for an individual cow.
Cows were bred by natural service using Holstein bulls
and were exposed to the bulls at all times of the year during
the time frame that the records cover. Cows were milked
two times a day during lactation and maintained on guinea
Percent black hair coat (BHC) was determined by using
the identification pictures on the record sheets for the cows.
Table 1. Number of obnrvatiors collected from dairy hed records of 520
Hotedin cows covering a 25-year period.
Calves born 1293
Calving interva 787
Percent black hair coat 462
FRt lactation levels 456
* y-. &94+o.05xK-2.3+32.0.0oix4
a I I I I I I I I I II V I
JAM F MARAPR MAY JUN JUL AUG WP OCT NOV OeC
Month of year
iF .79 P .0001
I I I I I 1 I I I I I
JA FS MAR APR MAY JUN JUL AUG SEP OCT NO OW
Month of birth of calf
Figue 1. Dlributlon of caVings throughout the year In Holtein cows (n Figure 2. Reationship of calving intemal and month of calving in Holtein
12930caV). cows (n -767).
Cows that calve and begin lactating during the late part of the year when
grass Is abundant and the temperature is decreasing have more efficient
reproduction and higher milk production than cows that calve during the
summer when grass is sparse and the temperature is elevated. "
A transparent grid was placed over the
left and right profile drawings of the
cow and the number of points that fell
on areas of black hair were counted.
Percent BHC was calculated by
dividing the number of points that were
over areas of black hair on both the
left and right sides of the cow by the
total number of points over the entire
left and right sides of the cow. The
number of cows and observations for
the various traits analyzed are shown
in Table 1.
To evaluate the effect of time of
year on reproduction and milk
production, the year was divided into
four seasons based on rainfall on St.
Croix. The first dry season included
January through April (DRY-1). May
is considered a rainy period (WET- 1),
and June through September is the
second dry season (DRY-2). The
second rainy season (WET-2) occurs
from October through December.
Based on data collected from 1987
through mid-1994, the average
monthly rainfall during DRY-1, WET-
1, DRY-2 and WET-2 was 2.1, 4.9,
3.6 and 5.1 inches, respectively, and
the average daily high temperature was
86.7, 88.7, 90.3 and 89.9 F,
The calving distribution
throughout the year exhibited a
definitive seasonal pattern (Figure 1).
Greater than 50 percent of the calves
were born during the winter months
(October through January) while less
than 8 percent were born during the
summer (June and July). By back
calculating from calving dates, it was
determined that the majority of
conceptions occurred during the
months of January through April. This
coincides with the first dry season
(DRY- I) which is also the coolest time
of the year on St. Croix. The
combination of relatively cooler
temperatures and adequate levels of
forage during the first two months of
the year, due to the previous rainy
season (WET-2), enhanced the
conception rate of the cows during the
early months of the year. Conception
rate is related to both temperature and
humidity in a negative fashion, which
may explain why the majority of the
conceptions occurred during a cooler,
dry period of the year (DRY-1).
Nutritional status may also have been
improved since the cows were
receiving adequate nutrition from the
forage available in the pastures.
Calving interval was influenced by
the time of year when the calf was
born. Cows that gave birth early in the
year had'a longer calving interval than
cows that gave birth during the later
months of the year (Figure 2). A
similar pattern was observed when the
data were analyzed using the wet and
dry seasons as a time frame (Table 2).
Cows that gave birth during DRY-1
had a 61-day longer calving interval
than cows that gave birth during WET-
2. The cows that gave birth during the
last three months of the year (WET-2)
were exposed to decreasing
temperatures, higher rainfall and
higher levels of forage availability than
at other times of the year. This
combination may have improved
conception rate at first service after
calving and shortened the postpartum
interval due to the increased plane of
nutrition provided by the forage. This
may also explain why the majority of
conceptions occurred during the early
part of the year. The cows had shorter
postpartum intervals and were bred
sooner than cows that calved during
other times of the year, both of which
led to a shorter calving interval.
Milk production during the first
lactation exhibited a pattern similar to
that of calving distributions (Figure 3).
Cows that began lactating in the middle
of the summer (July) had the lowest
milk production compared to cows that
began lactating in either the early or
late part of the year. The length of
lactation was not different among cows
Table 2 CaMng interval (Cl), first lactation mik production and length of
first lactation (DAYS) of Holtein cows on St. Croix hat either gav birth or
began their first location in a dry orwet season.
Season n CI, days n M01, Ib DAYS
DRY-1 260 5244 179 10.7501 294
WET-1 33 504" 22 10.121d 28
DRY-2 136 485W 113 9,877d 291
WET-2 338 486d 142 11.576c 308
WET 371 469' 164 10.849 296
DRY 386 51it 292 10,314 293
*ODRY-1 a Jnuary-Apdrl, WET-1 a June-September, WET-2 a October-
NWET WET-1 + WET-2, DRY = DRY-1 DRY-2.
Vakes within a tra it w diffeet supensripts diffr (P <.0).
that began lactating at different times of the year (Table
2), which means that the pattern of milk production is most
likely caused by other factors, such as the environment
and/or nutrition. Cows that began lactating during WET-
2 produced approximately 1700 lbs. more milk than those
that began in DRY-2 (Table 2). Cows that began lactating
in July (DRY-2) did so at a time of year when the
temperature was increasing and rainfall was low, while
those that began in DRY-1 or WET-2 were exposed to
either cooler temperatures or increased rainfall,
respectively. Holstein cattle spend less time grazing and
more time seeking shade during the hottest time of the day.
This could mean that the difference in milk production is
partially due to nutrition, since lactation increases the
nutrient requirements of cattle. There are higher levels of
forage available during the rainy season of WET-2 than
during the dry season of DRY-2 and the cows are in a
better nutritional status during the rainy season.
The population of cows was skewed towards a darker-
colored animal with an average percent BHC of 74.1
percent from a range of 12 to 100 percent. This description
of the population of cows is in agreement with results of
other researchers who described populations of cows in
Florida. There was no relationship found between BHC
and calving interval or milk production in the present study.
It may take analysis of larger numbers of cows before any
correlation between hair coat color and either production
or reproduction can be established: Since the average high
temperature on St Croix is 89F, cattle may be constantly in
a state of heat stress of varying degrees throughout the year.
Any influence of coat color and heat absorption on
reproduction/production traits may be masked by the effects
of the consistently high ambient temperature on St. Croix.
The results of the present study indicate that
reproductive efficiency and milk production of Holstein
cows on St. Croix are influenced by season of the year,
when determined by rainfall and temperature. Both
reproduction and milk production are negatively influenced
by the hot and dry times of the year when forage availability
is low. Cows that calve and begin lactating during the late
part of the year when grass is abundant and the temperature
is decreasing have more efficient reproduction (shorter
y 6802 +4428x- 1456 W* 171x'-6.Sx4
r= .5P< .00
JAN FEB MAN APR MAY JUN JUL AUG SBP OCT NOV DEC
Month of sted office lactaon
igur 3 Mik production during fkt laatios that be n atdffiMWt monthly
of the year (n -456).
calving intervals) and higher milk production than cows
that calve during the summer when grass is sparse and the
temperature is elevated.
The authors would like to thank Mrs. Caroline Gasped
of Castle Nugent Farms. St Croix, for providing access to
the animal records. This project was supported by CBAG
Project 9204572 in collaboration with Dr. Peter J. Hansen,
Professor, Department of Dairy and Poultry Sciences,
University of Florida, Gainesville.
0 0 a e g g
Mahogany is a precious wood,
highly prized by woodworkers and
furniture makers for its color, strength,
working qualities and durability. The
mahogany genus includes three species
(Swietenia mahagoni, S. macrophylla,
and S. humilis) and at least one
Small-leaf or West Indies mahogany
(S mahagoni) is native to the Greater
Antilles (except Puerto Rico), the
Bahamas and southem Florida. It has since
been introducedto other islands in the West
Indies and is now naturalized throughout
the Caribbean. Small-leaf mahogany
ranges in habitat from lowland, tropical
moist to tropical dry forest formations. It
is noted for its ability to grow on dry sites
with shallow soils.
The big-leaf or Honduras mahogany
(S. macrophylla) is a lowland, moist to
humid tropical species that has a native
range from southern Mexico through
Central America to the Amazon basin
of Brazil, Peru and Bolivia. It has
become the most widely used mahogany
due to the depletion of the small-leaf
mahogany wood supply.
As a result of adjacent plantings of
these two mahogany species a natural
hybrid was produced. This hybrid is
referred to as medium-leaf or hybrid
mahogany (S macrophylla x mahagoni).
Medium-leaf mahogany, as the name
implies, is intermediate between its two
parents in leaf and seed capsule size. It is
thought to grow faster than the small-leaf
mahogany and be more drought tolerant
than big-leaf mahogany. Its wood is of
better quality than big-leaf mahogany
and produces a straighter stem than the
"Both the big-leaf and medium-leaf mahoganies
exhibited severe defoliation at the 7- and 21-day
watering Intervals. 9
Small-leaf mahogany was probably
brought to the U.S. Virgin Islands as early
as 1770. It was widely planted as a shade
tree along roads and driveways and was
grown for use in furniture making. It has
become naturalized on St. Croix and can
now be found in nearly all parts of the
island. Big-leaf mahogany was introduced
much later on St. Croix for use in species
trials. Although not as common as small-
leaf, it is found planted along roadsides and
in public areas in the Virgin Islands. The
medium-leaf mahogany began to appear
in St Croix as a result ofcross-pollination
of big-leaf and small-leaf mahoganies in
adjacent plantings. After a number of
growth studies, the U.S. Forest Service
recommended that the medium-leaf
mahogany be used in forest plantations for
timber production on St Croix. This has
resulted in increased planting of the
Studies of survival and growth of the
three mahoganies common to St. Croix
have indicated that these species are
susceptible to drought conditions,
especially in the seedling and sapling
stages. Since many potential areas for
forest plantings on St. Cmix are marginal
sites with shallow or rocky soils, water
stress can be a recurring problem. In order
to evaluate the effect of water stress on
mahogany an experiment was conducted
at UVI-AES. The three mahogany species
canmcntDStci.Ck marophylk S
mahagoni, and S. macrophylla x
mahWn) were subjected to four watering
intervals (1,3,7, and 2l days) for period
of 37 weeks. Two-year old trees were
grown in 28-cm plasticpots(1I -L volume)
with a growing medium composed of
potting media, field soil and river sand in a
2:1:1 ratio. All trees received fertilizer and
pots maintained at field capacity prior to
the initiation ofthe experiment The study
was conducted in a greenhouse with a plant
spacing of 0.5 m within rows and 1 m
tree height and stem diameter were
measured. Tree height was measured from
the soil surface to the top of the terminal
bud with a measuring pole. Stem diameter
was measured at the soil surface with
calipers. After imposition of watering
regimes, observations were made on leaf
abscision, changes in leaflet color and
branching, Leafwater potential (LWP) was
measured every 21 days with a J-14 leaf
press (Decagon Devices, Inc., Pullman,
WA). Three 2-cm leafdiscs taken from the
most recent fully expanded leaves were
used for leaf waterpotential measurement
0 3 6 9 12 15 18 21
Watering Interval (days)
1 3 6 9 12 15 18 21
Watering Interval (days)
NO. of Branches 0-2 + 0.58x
2 R 20.37
/ Flgurm 4
0 3 6 9 12 15 16 21 2
Watering Interval (days)
R2 = 0,75
Figure 1. Rels onship between mahogany height
growth and watedag interval
FIgure 2. Relationship between mahogany
diameter growth and watering interval.
Figure 3. Relationship between mahogany leaf
water potenal and waring interval.
Figure 4. Influence of watering interval on
branching in big-leaf mahogany.
I a I I d I i
Pressure was applied via the leaf press
until a color change was observed, and
the pressure noted. Measurements of
LWP were made between 10:00 a.m. and
24 2:00 p.m. At the end of the experiment
total tree height and stem diameter were
There were no differences in total
height or stem diameter growth between
the three types of mahogany; however,
watering interval did have a significant
effecton both height and diametergrowth.
Greatest height (Figure 1) and diameter
(Figure 2) growth occurred at the three-
day watering interval and lowest growth
for both parameters at the 21-day interval.
Height growth ranged from 137.2 cm at
the three-day watering interval to 30.5 cm
24 atthe 21-day interval. The 21-daywatering
interval resulted in a diameter growth of
0.05 cm as compared to 1.24 cm of
diameter growth at the three-day interval.
The 21-day watering interval resulted in
dieback of the terminal bud in 50 percent
of both the big-leaf and medium-leaf
mahoganies. Fifty percent of the big-leaf
mahoganies also had terminal bud
dieback at the seven-day watering
interval. In addition, one small-leaf and
one medium-leaf mahogany died in the
21 -day watering treatment.
There were no differences in LWP
between species, but watering interval did
24 affect overall LWP. Leaf water potential
responded linearly to watering interval,
with highest LWP at the one-day watering
interval and lowest LWP at the 21-day
interval (Figure 3). The absence of
variation in LWP between species may
indicate that maintenance of LWP is not a
specific mechanism of drought tolerance
in mahogany. Although no differences in
LWP were found, distinct responses to
water stress were observed. Both the big-
severe defoliation at the seven-and 21-day
watering intervals. Leaf abscision is a
4 droughtavoidance mechanism. Since both
of these species have much larger leaves
than the small-leaf mahogany, their
tmnspirational losses would be greater,
thus necessitating leafabscision to reduce
water loes duringtinmesofstress. The small-
leaf on the other hand, did notdemonstre
any large scale leaf abscision, but did
reorient its leaves from a horizontal to a
vertical position at the higher watering
intervals. This is also a drought avoidance
technique for reducing transpirational
losses by decreasing the interception of
solar radiation. In both cases, the reduction
of leaf area for the interception of solar
radiation and the subsequent decrease in
photosynthesis could have been a factor in
the reduced growth of mahogany at the
higher watering intervals.
Branching was much more
pronounced in the small-leaf mahogany
(2ree) than theothertwo species (/Gbn)
although there was no correlation between
branching and watering interval for the
small-leaf. Small-leaf mahogany has a
tendency to produce mom branches and to
branch at an earlier age than either big-
leaf or medium-leafmahogany. Branching
in big-leaf mahogany was affected by
watering interval with a linear increase in
branchingwith increasing watering interval
(Figure 4). The increased branching in big-
leaf mahogany was probably due to the
dieback of the terminal bud and loss of
Under the environmental conditions
encountered on St Croix-shallow soils,
extended dry periods and high solar
radiation-water stress may be an
important factor in the survival and growth
of mahogany. Although older trees with
well-established root systems can
withstand extended periods of low water
availability,trees in theseedling and sapling
stage may be more susceptible to water
stress. This study demonstrates the effect
water stores has on mahogany height and
diameter growth. Additionally, the
increased branching exhibited by big-leaf
mahogany, and the dieback ofthe terminal
bud in both medium-leaf and big-leaf
mahogany at the higher watering intervals
suggest thatwater sss may also influence
stem growth and bole formation and,
ultimately, lumber quality.
This research was supported in part
by the U.S. Department of Agriculture
under Mclntire-Stennis Grant No.
Martin B. Adjei and
The dairy cattle industry on the Virgin Islands requires
high quality feed year-round to sustain production of milk and
milk products for the public. The industry is largely supported
by native pastures dominated by guineagrass and leucaena (tan-
tan) which grow well on the islands and produce large quantities
of highly nutritious forage during the rainy season. However, in
the dry season, the mature guineagrass stops growing and loses
quality. Stockpiling guineagrass in the field during the growing
season for use in the dry months can help overcome the lack of
forage caused by dry weather. When stockpiled, however,
guineagrass and leucaena lose digestibility and, in turn, lower the
animal's intake of essential nutrients. As an alternative to
stockpiling, nutritive value of native forage can be preserved by
ensilage. Indeed, some dairy farmers on the Virgin Islands have
contemplated harvesting surplus native guineagrass on a large
scale during the growing season for conservation as silage. As
part of the on-going evaluation of forage conservation systems in
the Caribbean, a pilot experiment was conducted at UVI-AES to
evaluate the preservative characteristics ofguineagrass/leucaena
forage when ensiled withoutor with molasses addition at3 percent
offorage biomass. Leucaena leaf meal inclusion was varied at 0,
5, 10 and 20 percent of forage mass before ensiling. In a second
experiment, silage characteristics ofguineagrass were compared
with those of sorghum and millet-elephantgrass hybrids which
are known to be good silage crops.
Knowledge ofthe ensilage process is an essential first step
towards adoption of good ensilage practices. During ensilage,
soluble carbohydrates, such as sugars and starch that are
present in forage material, get fermented to fatty acids. The
two essential acids produced are lactic and acetic. They reduce
the pH of the forage material which kills micro-organisms
present and thereby pickles the forage for long term
preservation. However, as long as oxygen is present, plant
enzymes and oxygen-utilizing (aerobic)micro-organisms will
use up the sugars through respiration and generate carbon
dioxide and heat which is capable of causing a considerable
rise in temperature of the material. Therefore, if forage is not
finely chopped and well packed in the silo after filling, air will
remain in the material and cause excessive overheating with a
consequent depletion of the same soluble carbohydrates needed
for conversion into acids.
In addition to the possible breakdown of carbohydrates to
carbon dioxide through respiration, approximately 60 percent of
the proteins in forage are broken down even in well preserved
material during ensilage. In well preserved silage, where a rapid
fatty acid type of fermentation occurs and a satisfactory degree
of acidity is produced, the end-products of protein breakdown
are mainly amino acids. However, in badly preserved material,
the amino acids are further degraded by micro-organisms into
ammonia gas which may be lost from the silo. As a result of
these chemical changes, gaseous losses (mainlyofcarbon dioxide
and ammonia) occur during ensilage. The amount of dry matter
lost in gaseous form may vary from 2-30 percent of the original
forage dependingon the type ofchemical changes induced. Since
these losses are caused by a breakdown of soluble and highly
digestible nutrients, it follows that the higher the gaseous losses,
the lower will be the feeding value ofthe silage.
Two problems associated with ensiling forage grasses are
their low concentrations of dry matter (DM) and fermentable
carbohydrates. Wilting offorage and the addition ofcarbohydrates
such as cane molasses to forage material before ensilage may be
beneficial. The objectives of our study were to characterize
guineagrass/ecana silage as influenced by wilting and molasses
additive, and to compare it with sorghum silage.
The field ofguineagrass used forour experimentwas mowed
back to a 6-inch stubble on May 13,1993, and allowed to regrow
for 40 days. Leucaena leaves, including young green branches,
were pruned from an established field on the 17th of June 1993,
chopped through a shredding machine into a meal and partially
dried in a forced-air oven at 1400F (60n) for 48 hours. The 40-
day guineagrass regrowth was approximately four feet in height
and atthe early flowering(boot) stage ofmaturity. It was harvested
with a sickle bar mower on the 22nd ofJune, 1993, and chopped.
Approximately 13 lbs ofchopped guineagrass forage was mixed
SA Fresh forage- molasses PH
34 0 Frsh forage + molasses I DM 5
33 Wiled forge + molesses I 1
Figure 1. Temperature
changes during ensiling of 21
on St Croix.
with all possible combinations ofO, 5, 10
or20 percentofthe partially dried leucaena
leaf meal and 0 or 3 percent molasses, on
fresh weight basis. The molasses was
diluted with an equal weight of water to
facilitate mixing. Treated forage was
thoroughlymixed, hwd-packed in 5-gallon
plastic buckets and weighed. There were
three replicates (buckets) for each
treatment. Fresh forage subsamples were
dried inoven at60C (140F)to determine
dry matter (DM) content Filled plastic
buckets were covered first with a 6 mil
plastic sheet and then with a lid that had a
rubber seal around the rim to keep out air.
In addition to ensiling fresh grass, some
guinea grass was leftto wilt in the field for
approximately three hours before it was
subjectedtosimilarchopping, leucaena and
molasses treatments and ensiled. A 3-ft
copper-constantan thermocouple was
inserted though a hole drilled in the center
of the plastic lid into the bucket to help
monitor temperature changes during the
fermentation process. The hole in the lid
around the thermocouple was sealed with
an epoxy resin (diethylene triamine).
Temperatures were read at 24 hour
intervals on a digital thermometer for 10
days during ensilage. Filled buckets were
stored atroom temperature (approximately
27C or 80F) in a shed for 28 days after
which period, they were reopened and
reweighed.Following visual determination
of color, subsamples ofensiled forage were
removed for DM, pH and volatile fatty acid
(VFA)determinations. Color was rated on
a scale of 1 to 5 (green=l greenish
brownf2.5; brown and moldy=5).
In the other experiment, fresh and
wilted guineagrass, Puerto Rico 5BR
0 24 41 723 9 120 144 iM 10S2 211 240
Time of ensuing, hrs
forage sorghum (PR 5BR) and a millet-
elephantgrass hybrid (M-E) forages were
chopped,mixed withOor lOpent leumcana
leafmeal andOor3 pant molassesadditive
and ensiled as described above.
experiment whereas follows.Themaximum
observed temperatures occurred within the
first few hours of ensiling and ranged
between 30 and 32 (86-90F) depending
on treatment (Figure 1).
44 For proper
fermentation and long
term preservation of
silage, the results
suggest that molasses or
other energy additive are
Although minimal, the rise in
temperature was affected (P<0.01) by
molasses additive and the condition of
guineagrass (fresh vs. wilted). Maximum
temperature was greater in forage without
molasses additive and for wilted over fresh
guineagrass (Figure 1). This might have
been caused by differences in rates of fatty
acid production and hence rates of
containment ofaerobic microbial activity
under the different treatments. Following
the initial increase, temperatures declined
sharply as entrapped air became depleted.
Apparently, fermentation wascompleted in
8 days (192 hours) beyond which time a
further decline in forage temperatures was
observed for all treatments (Figure 1). Itis
0 3 a 3
Molasses Additive, %
Figure 2. Effect of molasses additive on the final
pH and dry matter losses of guineagrassfitucaena
silage onSt Croik Average overfreh and willed
generally accepted that an early
temperature rise above 55-C (130F) of
forage material in the silo is undesirable.
In our experiment, temperature rise caused
by the initial aerobic respiration was
minimal; probably because forage was
firmly packed to ecludeairand sealed in the
plastic bucket silos.
The initial pH of fresh or wilted guinea
grass forage before ensiling was 6.1 and
that for the partially dried kucaena meal
was 6.2. The pH ofsilage was significantly
affected by molasses additive and leucaena
inclusion (Figures 2 and 3). The final pH
of ensiled forage averaged 4.5 when
molasses was added but 5.1 without
molasses (Figure 2). A pH of about 4.2 is
generally recommended for long term
preservation of silage. The final pH of
silage was generally elevated by the
inclusion of leucaena meal (Figure 3)
probably because of the high buffering
capacity of protein from the legume.
Dry matter content of the fresh and
wilted guinea grass before ensilage
averaged 20 and 25 percent, respectively.
The dried leucaena leaf meal was 83
percent DM. Thus, inclusion of leucaena
at the 0, 5, 10 and 20 percent of fresh
weight on an ensiled basis corresponded
to 0, 18, 32 and 51 percent DM inclusion
for fresh guinea grass and 0, 13,25 and
43 percent DM inclusion for wilted guinea
In our experiment, DM losses during
the 28 day storage ranged from I to 6.8
percent and was significantly influenced
by molasses addition and leucaena
inclusion (Figures 2 and 4). The DM
losses, when averaged over leucaena
treatments, were 2.4 and 5.2 percent for
Figure 3. E tof level oflaucama inlusion (fresh
wight basis) on the final pH of gulneagras/
laucnam silage on St. Crolx.
0 I 10 20
Level of Laucaena, %
Figu & Effec of lvl of aucMan Inclusion (feeh
weight basis) on the final color of gulneagrass/
leuceena silage on St. Crokx.
forage ensiled with and without molasses,
respectively (Figure 2). However, theeffect
ofleucaena inclusion on DM losses (Figure
4), was inconclusive. Dry matter losses
during ensilage was highest at 0 percent
level of leucaena inclusion and lowest at
the 5 and 20 percent levels of inclusion.
In well preserved silage, where the
temperature has not risen to any
appreciable extent, the carotene(pigment)
content should be similar to that of the
original crop. Large amounts of carotene
can be lost, however, in overheated silage
changing the color to brown. Although
color is sometimes used to classify the
quality of silage, the color of silage
produced in our experiment was largely
controlled by leucaena inclusion. The
addition of dried leucaena meal (which was
brownish green incolor)generallymodified
the greenish color of guinea grass silage
towards a brownish color (Figure 5).
In the second experiment, the level of
leucaena inclusion (O0vs. 10%)did nothave
any marked effect on measured traits. The
final silage DM for the sorghum, wilted
guineagrass and M-E hybrid ranged from
28 to 30 percent but only 23 percent for
the fresh guineagrass (Figure 6). The
sorghum and M-E hybrid silage contained
lower (PO.01) acetic acid but higher levels
of lactic acid than guineagrass (Figure 7).
A higher level of lactic acid is the ideal
situation since acetic acid is less stable. It
has been suggested elsewhere that a
deficiency of lactic acid bacteria may lead
to high concentration of acetic acid in
silages from tropical grasses. The final pH
of M-E hybrid and sorghum silages were
4.04 and 3.85, respectively. These levels
of pH were not affected (P>0.05) by
molasses additive (Figure 8). Apparently,
grains in the sorghum and sugars in the
M-E hybrid forage provided sufficient
frmentablecarbohydrates during ensilage.
On the contrary, molasses additive was
essential to lowering(P0.01) the final silage
pH ofguineagrass from 5.0 to 4.5, whether
ornot the forage was wilted initially.
Our preliminary experiment has
demonstrated the possibility of preserving
guineagrass/leucaena forage as silage in
the Caribbean for useduringthe dry period.
Temperature rise, DM losses and final
silage pH were minimal when molasses
was added to forage prior to ensilage.
Therefore, for proper fermentation and
long term preservation of guineagrass/
leucaena silage, the results suggest that
molasses or other energy additive will be
necessary. The concentration ofacetic acid
in guineagrass silagewas greaterthan lactic
acid which suggests a possible lactic acid
bacteria deficiency. Future studies will
evaluate the efficacy ofbacteria inoculant
and the level of molasses additive required
to obtain a pH in the range of 3.8 to 4.2.
We also intend to assess the effect of
leucaena inclusion on the nutritive value
This research was supported in part
by Hatch Project No. 0156333.
FHG WDO M-E PRSaR
Type of Gras. %
Figure Effectofgrass type n the final dy mater
cornent of slage. FG ftmh guineagrass WDG =
wedguineagrass M E= m st-lephantgreashytmd
and PR5BR Puertc 5BR foragesoughum.
FMG WOO M-E PRSBR
Type of Grass
Figure 7. Effect of grass type on the final acetic
acid and lactic acid content of sllge,
A, comparison within acid at P- 0.05
FHS WDOO M-E PRSR
Type of Greas
Figure 8. Effect of grass type and molasses
addition on the final pH of silage.
Grass x molasses Interaction P < 0.01.
0 5 10 20
Level of Leucaena. %
FgurM 4. effect of vel of lb n Minclusion (fresh
weight base) on dry matterloses otguinagnrei
eucena silage on SL Croix.
2I griwnAti btwomn
0 5 10 20
Martin B. Adjel and
Terry J. Gentry
Efficiency of ruminant livestock
production in the Caribbean Basin is
severely hampered by seasonal deficiencies
in the quantity and quality of available
forage, resulting in recurrent livestock body
weight losses. The inability of native
pastures to produce quality forage during
cyclic dry periods has been known for
many years since the work done by A.J.
Oakes in 1966 and 1969 on St. Croix. He
reported that crude protein content of
predominant pasture grasses on St Croix
normally ranges between 3 and 6 percent
ofdry matter (DM) during the dry season,
with digestibilities of 36 to 43 percent.
According to the Animal Nutritional
experts, any forage crude protein content
below 8 percent of DM will reduce forage
intake and adversely affect livestock
In 1989, a survey conducted by
personnel of the Cooperative Extension
Service of the University of the Virgin
Islands also showed that hay produced for
dairy farms on St. Croix was low in crude
protein (5.2 to 6.5% DM) and total
digestible nutrients (51 to 59%). This poor
feeding value of hay necessitated the use
of dairy rations that contained more than
60, percent of DM from imported
concentrates at an estimated monthly cost
of $24,000 to the three major dairies in St.
Croix alone during the severe 1989-90 dry
season. The huge expense of importing
grains and oilseed products, which compete
with human and nonruminant livestock
consumption, limits the widespread use of
concentrate supplements in the region.
Forage conservation either through hay or
silage production provides a viable option
for many Caribbean Islands; however,
advanced maturity of many stored forages
and crop residues results in low feeding
Chemical treatment to improve the
feeding value of low quality forage offers
another opportunity to utilize large
amounts of low quality grasses and crop
residues. There is some evidence that alkali
treatment of forages results in increased
forage digestibility, feed intake and animal
performance. Among the chemicals,
ammonia, in either the aqueous or
anhydrous form, has received the greatest
attention because of its dual capability to
increase both crude protein and fiber
digestibility in forages. Ammoniation of
hay has been adopted rapidly by farmers
in North America and Europe because of
its high economic return. However, limited
availability of ammonia, lack of
pressurized equipment to handle it, high
cost and increased Federal regulation of
transportation ofanhydrous ammonia have
prevented its use in the tropics including
The use offered grade urea as a source
ofammoniation to improve forage quality
has greater application in the tropics where
it is widely available. For it to be effective,
urea must first be converted to ammonia
under air-tight conditions such as a plastic
cover. This conversion requires the
presence ofurease enzyme and water. Most
urea treatment studies have been conducted
with wilted or reconstituted cereal straws
stored by ensiling. Very little information
is available concerning urea treatment of
tropical grass hay. The objectives of our
study were to investigate the effectiveness
of urea-ammoniation for improving
guineagrass hay and measuresheepgrowth
performance when fed urea-treated hay.
Factors including final moisture content,
urease addition, urea application method
and urea treatment level were studied.
Experiments were conducted at the
Agricultural Experiment Station of the
University ofthe Virgin Islands (UVI-AES)
on St. Croix, under tropical forage-
livestock production system. In each
experiment, hay was purchased from the
local Department of Agriculture which
supplies most of the hay used by local
farmers. The hay was composed ofgreater
than 90 percent ofguineagrass (Paicum
Figurm 1 I
a 1 4 6 N
Urea Level, %OM
0 a 4 I U
Usre Level, %OM
Figure 1. Effet of urea treatment level on
aude protein content of guneagrs hay in
Experiment I I
Fgum 2. Effct of urea trealmnt evel on
in vitro organic matter digestion of
guinagra hay in experiment 1.
maximum), with small quantities of leucaena (Leucaena
lewoceplrda),johnsongrass (Sorghm lepense), cash (Acaci
spp.) and hurricane grass (Bothriochloapertusa). Initial hay DM
content was consistent across experiments and averaged 87
Two laboratory-scale experiments were conducted using 5
kg portions of hay to determine the effects of final forage moisture
content, urea treatment levels and grease application on hay
Treatment factors and levels for Experiment I were as follows:
1) Forage final moisture content 25% or 40%
2) Urea treatment level 0, 4, 6 or 8% DM
3) Urease With or without
All possible combinations of factor levels were studied. Urea and
urease were mixed with the appropriate volume of water to
reconstitute the hay to the desired final moisture concentration,
and then sprayed onto the hay. There were three replicates of
each treatment combination. Forage was then packed into
individual plastic bags and stored air-tight at room temperature
(approximately 78,F) for 60 days.
In Experiment 2, urea treatment level was deliberately
confounded with final forage moisture content by using a 15%
(w/v) urea solution to obtain all urea application levels.
Treatment factors and levels for Experiment 2 consisted of;
) Urea treatment level 0, 4, 6 or 8% DM
2)urease With or without
Ureasewas added to an appropriate quantityof the 15 percent
urea solution toyield the appropriate urea treatment level, sprayed
onto the hay and stored as described above.
After 60 days storage, bags in Experiments I and 2 were
opened and the contents of each bag were thoroughly mixed.
Subsamples were dried, ground and analyzed for dry matter,
organic matter (OM) and crude protein according to the
Association of Official Analytical Chemists method. Cell wall
fibercomponents including hemicellulose(HC), cellulose, neutral
detergent fibre (NDF), acid detergent fibre (ADF) and acid
detergent lignin (ADL) were detennined by the procedure of
Goering and Van Soest. In vitro OM digestion (IVOMD) was
determined by the modified Tilley and Tenry procedure described
by Moore and Mott.
Two field-scale experiments were also conducted using large
round bales of approximately 32 kg (700 pounds) each.
Experiments 3. Treatment factors and levels follow:
1) Final moisture content 25% or 40%
2) Urea treatment level 0, 4 or 6% DM
3) Application method Sprayed-on or injected
Urea was dissolved in an appropriate volume of water to
reconstitute each bale of hay to 25 or 40 percent final moisture.
The urea solution was applied by either spraying from watering
cans onto both flat surfaces of the bales, or by low pressure (10
psi) injection at five sites on each flat surface. Each treatment
combination was applied to three bales. Each bale was stored
separately in slip-on, 6-mi plastic bags for60 days. Afterstorage,
each bale was sampled with a core sampler at approximately 20
sites. Forage samples were processed and analyzed for quality
measures as described above.
The second field experiment (Experiment 4) was designed to
confound moisture content and urea treatment levels by using a
15 percent (w/v) urea solution. The appropriate quantity of the
15 percent urea solution was either sprayed onto or injected into
bales and stored as described. Bales were sampled and analyzed
for quality as described.
Large round bales described above fir Experiment 3 from
the 0,4 and 6percentureatreatment levelswhere the measolution
was sprayed on the hay at 25 percent forage moisture content
were used to determine the effect of urea-treatment level on the
digestibility and growth performance of sheep fed guineagrass
In the growth trial (Experiment 5), 30 St. Croix white hair
weaned lambs ofapproximately 18 kg (40 pounds) initial body
weight were assigned to six pens, resulting in five head per pen
and fed for 50 days. Two pens were assigned to each 0, 4 and 6
percent urea-treated hay. Sufficient feed grade urea was applied
at feeding time to the 0 and 4 percent urea-treated hay to equal
Table 1. The effect of urease addition on the chemical compoeltlon and in
Table 1. The effect of urease addition on the chemical composition and in
vitro digestion of guineagrass hay In Experiment 1.
Uase CP1 NDF ADF Cellulose HC ADL IVOMD
Without 11.4 75.1 48.1 26.8 38.1 10.1 38.9
With 12.0 74.9 48.1 26,8 38.1 10.0 371
1CP = crude protein; NDF neutral detergent bre: ADF acid detegent
fibre: HC hernicellulose ADL = acid detergent lignin; IVOMD -= In vitro
Table 2. The interactive effects of method of application, forage moistum
and urea treatment levels on the crude protein content guineagrass hay In
Application Moisture 0 4 6 Mean U Effect
- - CP,%DM - -
Spray 25 5.3 7,8 10,5 7.4 L"
40 5.9 6.8 8.1
Inject 25 5.3 6.7 8.1
40 5.7 5.7 6.4 -
Application x moitre x urea interaction P<0.05.
L, NS a Linear or non-sig nicant urea effect
" = Significance at P<0.01.
the crude protein content of the 6 percent
uwea-treated hay. Dehydrated alfalfa pellets
(14% CP) were fed to all sheep at a daily
rate of 1 percent of body weight Water
was provided free-choice. Total feed
offered was recorded and total feed refused
was collected and weighed. Dried and
ground subsamples of feed and refusals
were analyzed as described above.
In the digestion trial (Experiment 6),
six wether lambs, similar to those used in
the growth trial, were used to determine
the digestibility ofthe same diets as those
in the growth trial. Sheep were fitted with
fecal collection bags and housed in
individual crates. Each of the three feeding
periods consisted of a 10-day dietary
adjustment phase followed by a 5-day
collection phase. Two animals were
assigned to each treatment diet during a
feeding period and animals were rotated
among diets between periods. Following
total recording of feed or collection and
weighingof refusals and feces, subsamples
were processed for quality analyses as
1) Did urease addition, forage moisture
content and urea treatment level affect
quality of small hay samples in the
Results in Table I indicate that urease
addition did not influence any forage
quality measure nor did it interfere with
the quality response to urea treatment level.
This was probably due to the presence of
adequate amounts of urease enzyme in or
on the hay that was treated. However, as
urea treatment level was increased from 0
to 8 percent of forage DM, crude protein
(CP) content of forage increased in a linear
manner from 4.3 to 18 percent as shown
in Figure 1. The improvement in forage
CP due to urea treatment was greater for
the 25 percent than for the 40 percent
Additionally, urea treatment improved
organic matter digestibility of forage by
an average of 35 percent as shown in
Figure 2. Also, acid detergent lignin
concentration in Experiment I and
hemicellulose concentration in experiment
2 decreased in a linear manner to increasing
urea treatment level (data not shown).
6 Our results have
provided evidence that
urea treatment will
forage crude protein
and overall feeding
value of locally
2) Did forage moisture content, urea
treatment level and application method
affect the quality of guineagrass bales of
The response of forage quality to urea
treatment level depended on forage
moisture and application method. In both
Experiments 3 and 4, greatest increase in
forage crude protein content due to urea
treatment was found when the urea solution
was sprayed onto the hay and when the
final forage moisture content of 25 percent
was used as shown in Tables 2 and 3. Also
in both experiments, forage organic matter
digestibility was only moderately improved
when the solution was injected into the
bales. When the urea solution was sprayed
onto the hay, forage digestibility was
increased by approximately 15 percent in
both Experiments 3 and 4 (Tables 3 and
4). In addition, cell wall components such
as ADL and HC were reduced when the
solution was sprayed onto the hay butwere
not affected when the solution was injected
into the hay.
3) Did urea treatment level affect hay
digestibility and growth performance in
In the digestion trial, hay intake
increased in a quadratic manner with
increasing urea treatment level (Table 5).
Apparent organic matter digestibility was
not affected by urea treatment, but, due to
increased hay intake, digestible organic
matter intake increased in a quadratic
manner with increasing urea treatment
level. Apparent digestible fiber fractions
increased in a linear manner due to urea
In the growth trial, hay intake also
increased in a quadratic manner (Table 6).
The results show that daily gain (17, 51,
48 g) and feed efficiency (.013, .032, .033)
increased in a linear manner with increasing
urea treatment level from 0, 4 to 6 percent
Our results have provided evidence to
support the fact that urea treatment will
significantly improve forage crude protein
content, digestibility and overall feeding
value oflocally produced hay. The fact that
urease addition had no effect on urea-
ammoniation has an important practical
implication. Since synthetic urease is an
expensive input item, the need for it would
adversely affect he cost/benefit ratio ofthe
In the laboratory experiments where
the quantity of water required to
reconstitute the hay to the appropriate
forage moisture content was thoroughly
Table. The Interactive effects of method of apploalton and urea treabnent
level, using a common strength (15%) urea solution, on the chemical
composition and ri vitro digestion oguineagrss hay in Experknent 4.
Ure, %DM (U)
Method 0 4 6 BE U Eiect
Spray CP* 54A 8.0 8.2 0.27 L"
NDF 75.6 73.0 71.6 0.39 L'
IVOMD" 41.6 48.0 48.0 0.78 L" Q*
Inject CP 5.1 6.7 6.9 0.27 L*
NDF 74.2 74.9 74.5 0.39 NS
IVOND 40.4 45.4 43.0 0.78 Q"
CP" = Significant method effect on CP (P<0.01);
L, Q, NS w linear, quadratic or non-significant urea effect.
Table 4. The efcts of method of application, forage moisture and urpe
treatment level on In virom digetion of gulneagrass hay in Experiment 3.
AfolaBon Mtelsture 0 4 6 Memi U Eflect
- - IVNOMD ----
Spray 25 42.6 48.2 49.3
40 42.4 45.9 46.2
InJect 25 42.0 44.0 43.5
40 43.8 44.3 46.7
mixed with the hay by hand and the entire
contents were sealed in plastic bags. we
obtained improvement in forage crude
protein content in the order of 400 percent
and organic matter digestibility of 30
percent. A small farmer with a few square
bales that can be completely soaked with
urea solution inside a container could
approach laboratory conditions and obtain
similar high levels of quality improvement.
different fmn those in the field where larger
amounts of round-baled forage aretreated.
With large round bales (320 kg or 700
Ibs) of approximately 87 percent DM, we
used approximately 64 kg(140 lbs)or 164
kg (360 Ibs) of water as urea solution to
reconstitute balesto 25 or40 percent forage
moisture contents, respectively. The
solution could not be mixed into hay by
hand but rather applied by spraying or
injection. Large quantities ofurea solution
was observed to seep out of bales atthe 40
percent moisture level. This considerably
reduced the overall effective meatreatment
level at the 40 percent moisture content
Nevertheless, we obtained an improvement
of 100 percent in crude protein (from 5 to
10% CP) and of 14 percent (from 43 to
49% IVOMD) in organic matter
digestibility by spraying urea solution to
the flat surfaces of round bales of hay to a
final moisture content of 25 percent The
injection method was less effective
probably because the urea solution became
localized in the bales.
When the 15 percent strength urea
solution was used to obtain urea treatment
levels, final forage moisture contents of
treated bales were approximately 26 and
31 percent for the 4 and 6 percent urea
treatment levels, respectively. Therefore,
increasing urea treatment level produced
intermediate effects on hay quality as
described above for 25 and 40 percent
moisture contents. When dealing with dry
hay (87% DM), fmers will have to select
appropriate urea solution strength (15 to
22% w/v) thatwill allow reconstituted final
forage moisture to be restricted between
25 and 30 percent in order to avoid
excessive urea seepage losses.
Urea treated hay when fed to sheep
improved daily gain by as much as 150
percent Such large improvements in the
feeding value of conserved forage via urea
treatment will aid considerably in
sustaining livestock production in the
Caribbean during the dry season if the
technology is adopted. The new technology
is ready to be tested on selected fanns on
St. Croix- Several scientists from other
Caribbean islands have expressed interest
in the technology as well. The next phase
ofthe project will have to focus on the need
to conduct feeding trials on dairy cattle and
measure residual levels of ammonia in the
milk product before technology can be
extended to the regional dairy industry.
diets based on urea-treated hay could be
formulated to overcome many of the
perennial, dry-season feed constraints to
ruminant livestock production in the
This research was funded In part by
the Caribbean Basin Advisory Group
(CBAG) Project No. 92-34135-7300.
Table S. Influence of urea treatment level on the digestibility by sheep of
gu eagrass haey
Ursa (U) Treatenmt Level
ItemI 0 4 8E U Effect
Intake, g OM
Hay 510.9 614.4 572.6 20.4 Q*
Pellet 166.3 153.5 157.7 4.5 NS
Total 677.1 767.9 730.3 22.2 0Q
OM igmibilty, % 62.6 64,9 65.3 1.2 NS
DigemtbleMOitake.g 424.4 496.5 475.6 14.3 Q*
NDF digestibillly,% 65.9 69.2 70.5 1.1 L*
ADF dignatillty,% 62.5 68.7 67.1 1.3 L*
HCdigetWbiltWy.% 70.9 73.4 76.3 1.1 L"
10M organic mater, NDF neutral detergent fibre, ADF = acki detergent
fibre. HCG hemiellulose
2 L, Q = Lnearorquadraic uea treatment effects ("= P<0.05." = P<0.01).
Table 6. Influence of urea treatment level on the growth performance by
sheep fed gulneagrass hay.
Uise (U) Treatment Level
Iem' a 4 6 SE U Effe?
Intake, g OM
Hay 1025 1294 1156 50.21 L" Q"
Pellets 306 317 309 3.6 NS
Total 1330 1610 1465 52.2 L" Q"
Daily gain,g 17.3 50.8 47.7 7.5 L*
Gain~feed 0.013 0.032 0.033 0.005 L'
1OM = organic matter, NOF neutral detergent fibre, ADF = acid detergent
fibre, HC hemicaluose.
2 L 0- Unear or quadrat icuream tramt effects = P<0.05. -P
Production In The
M.C. Palada and
Yam, plants of the genus Dioscorea, is an important
crop in the Caribbean, where production ranks second only
to West Africa. Prices for this crop are higher than for
other tropical root crops. However, production of this
popular food crop declined dramatically in the Virgin
Islands between 1960 and 1987 (production in 1987 was
20 percent of production in 1960). This decline was
primarily due to a concurrent reduction in the available
agricultural cropland during the period (acreage in 1987
was 15 percent of acreage for 1960). If this trendof reduced
harvested agricultural acreage continues, it is critical that
production per unit land area be increased to at least
maintain the present level of local yam production. The
incidence of the fungal disease anthracnose (Colletotrichum
gloeosporioides) has also played a role in reducing the yam
yield per unit area of land.
The University of the Virgin Islands Agricultural
Experiment Station (UVI-AES) is conducting research
investigating methods to increase the production of tropical
root crops, including yam. This research has focused
primarily on D. alata because it is the only locally grown
species of economic importance.
In the Virgin Islands, yams are usually grown without
pest management practices, proper fertilization or
irrigation, and weeded only when infestations are severe.
Yam, however, requires the most intensive management
compared to other tropical root crops for a high yield of
good quality tubers. This includes proper plant nutrition
and optimum soil moisture levels. Most soils in the Virgin
Islands pose some stress conditions for plant growth-high
soil pH, deficiencies in phosphorous and micronutrients,
heavy soils and low annual rainfall.
The studies summarized in this paper were conducted
at UVI-AES on St. Croix. The soil is a Fredensborg
loamy, fine, carbonatic, isohyperthermic, shallow, typic
Initially, a series ofgermplasm screening and evaluation
studies were conducted over a number of years. Germplasm
were obtained from USDA-TARS (Mayaguez,PR),
Barbados, the Dominican Republic and local farmers.
Anthracnose tolerance was given high priority in these
evaluations. These trials have identified cultivars which
consistently produced high yields of good quality tubers,
while exhibiting acceptable disease tolerance. Selected
cultivars include Binugas, Seal top, Gunung, PR-PI 15580
Subsequent trials were conducted to determine the
optimum plant density and tuber piece size for planting. It
was found that closer spacing resulted in a necessary
reduction in tuber size, a larger proportion of better-shaped
yams and increased yields. Closer spacing also reduced
the amount of time needed for the plant foliage to cover the
soil, thereby reducing weed competition and the number of
times the crop needs weeding. The results of these trials
indicated that combining a yam tuber piece weighing
approximately 115-125 g (1/4 lb) and a planting density of
33,333 plants/ha (.3m x lm) were optimum. This
combination increased yields due to the increased number
of tubers produced. The tubers were smaller and better
ToM UAblMe DIVy MN~r
50 100 1
Nitrogen application rate (kg ta)
Figure 1. Effect of applied nitrogen on yield of 'Gunung'yam.
0Ju Aug Sep Oct Nov Dec Jan
NMrogen application rate (kgm)
Figure 2. Monthly rainfall amountsduring 1992 and 1993 yam growing usonmm.
If this trend of reduced harvested agricultural
acreage continues, it is critical that production per
unit land area be increased to at least maintain the
present level of local yam production. ??
shaped (with fewer 'toes') than when
larger-sized planting material and
lower plant densities were used. The
benefits of smaller tubers are that they
are 1) easier to harvest, 2) less prone
to damage during harvesting, 3) easier
to handle (for both farmer and
consumer), 4) more marketable and 5)
easier for consumers to utilize.
Even though many trials have been
conducted in Africa and the Caribbean
to determine the response of yams to
fertilizer, little research of this nature
has been conducted in the Virgin
Research in other Caribbean
countries has indicated that different
varieties respond differently to the
same fertilizer application. Fertilizer
recommendations will vary based upon
soil type and local climatic conditions.
Varying levels of positive yield
responses have been obtained when
fertilizer was applied to yams. The
level of response was affected by other
factors including cultural practices.
Nitrogen has been reported to be the
most important nutrient because its
application was found to significantly
Two trials were conducted at UVI-
AES to evaluate the effect of varying
rates of nitrogen on the production of
D. alata cultivars. In the first trial,
varying levels of nitrogen were
evaluated for their effect on the yield
of two cultivars, 'Binugas' and
'Gunung'. Ammonium sulfate was
split-applied to provide the plots with
nitrogen at rates of 0, 100, 200 and
300 kg/ha. The first halfof the nitrogen
and all of the phosphorous (100 kg/ha
using triple super phosphate) were
applied one month after planting. The
second half of the nitrogen and all of
the potassium (100 kg/ha using
potassium sulfate) were applied three
months after planting. The initial
soil pH was 7.8 for both trials and
the soil nitrogen were 116 and 95
ppm for 'Binugas' and 'Gunung'
tuber pieces weighing approximately
I 15g astheplanting material. Plots were
Table 1. Effect of irrigation on 'Binugas' yam production (1992). Tablet. Effect of irrigation on 'Hinuga' yam production (1903).
Irrigation Tuber Total Yield Marketabl Dry Matlar Total Dry Irrigation Tuber Total Yield Marketable Dry Matter Total Day
Rate (Pa) Sze (g) (ttha) Yield (tha) (%) Matter (i1l) Rate (kPa) S3e (0) (Utha) Yield (that) (%) Matter (titha)
20 500 252 23.5 21 5.3 20 308 14 10.1 18.3 2.5
40 480 23-8 20.7 20.9 5 40 367 16.2 9.8 17.4 2.8
00 775 23.8 21.6 21.8 5.2 00 336 13.2 6.7 18.9 2.5
Rain 357 20.2 17.1 22.6 4.6 Rain 373 14.8 10.4 19.7 3.1
Signilicance NS NS NS Q" NS Signihcance NS NS NS Q" NS
N8.* Nonsignificant or significant at P = 0.05. NS.* Nonsignificant or significant at P = 0,05,
Quadratic (Q) response. Quadratic (Q) response.
3 m x 3.7 m and consisted of 3 rows (ridges), spaced I m
apart. Plants were spaced 0.3 m within rows. The
experimental design was a randomized complete block with
four replications. A drip irrigation system was installed
consisting of 1.27 cm poly-hose as the sub-mains and Drip
Strip Plus (Hardie Irrigation) tubing with laser-drilled
orifices 0.3 m apart (Hardie Irrigation) as the laterals. Each
plot was harvested at six months after planting. Tubers
from 10 plants in the center row of each plot were harvested.
The weight of marketable tubers was recorded.
In a second trial, cultivar 'Gunung' was evaluated for
the effect of nitrogen applied at varying rates. Ammonium
sulfate was applied to provide nitrogen at rates of 0, 50,
100 and 150 kg/ha. Phosphorous and potassium were both
applied at rates of 75 kg/ha using triple super phosphate
and potassium sulfate, respectively. The initial soil pH was
7.9 with a nitrogen content of 90 ppm. The experimental
design, plot size, layout, establishment and harvesting
method were similar to those described, All of the fertilizer
was applied two months after planting. At harvest (seven
months after planting), the total and marketable weight of
tubers were recorded. Tuber sub-samples from each
replication were peeled, sliced, then dried at 700C to obtain
the dry matter content.
The application of nitrogen in the first trial up to 300
kg/ha did not affect production of the two cultivars
('Binugas' or 'Gunung') utilized in the study. Yields were
very similar for both cultivars at all application rates, and
there was a trend that higher yields were obtained from the
0 nitrogen plots.
Production of 'Gunung' yams was influenced by the
nitrogen fertilization treatments in the second trial. Tuber
size was largest from the 0 and 100 kg N/ha treatments.
There was a negative linear response of both total and
marketable yields to the rates of applied nitrogen (Figure
1). Yields decreased as the nitrogen application rate
increased. The percent dry matter of the tubers was not
affected by the treatments but total dry matter production
had a linear response (Figure 1). Tuber size was not
significantly affected by the application of nitrogen.
Yams do not tolerate prolonged periods of drought
without a yield reduction, especially during the critical two
to three month period when all of the food reserves of the
seed piece have been depleted. Moisture stress also delays
tuber initiation. Tubers develop best when rainfall is
frequent and the soil is well-drained.
Two trials were conducted in 1992 and 1993 to evaluate
the effect of varying irrigation rates on the production of
D.alata yam. Cultivar 'Binugas' was grown in field plots
and irrigated to maintain soil moisture levels of 20,40 and
60 cb. A rain-fed treatment was also included. Nitrogen
was applied at 100 kg/ha using ammonium sulfate, and P
and K were both applied at 75 kg/ha (using triple super
phosphate and potassium sulfate, respectively).
A drip irrigation system was installed as previously
described. Tensiometers (Irrometer Co.) were placed in the
root zone in the center row of plots, to monitor the soil
moisture content. When the tensiometer readings exceeded
the level for a specific treatment, the irrigation system was
turned on until the soil moisture content was increased to
the desired level. Semi-automatic timers were used to turn
the irrigation system on and off. Water meters were used to
measure the amounts of water applied to each treatment.
The experimental design, plot sizes, layout,
establishment and harvesting method were similar to the
nitrogen rate trials. Rainfall during the 1992 and 1993
growing seasons were 830 and 511 mm, respectively (Figure
2). Yams were harvested at seven and six months after
planting in 1992 and 1993, respectively.
The response to the irrigation rates was similar for both
years. Yields were higher in the first year, probably due to
the longer growing season (Tables I and 2). During both
growing seasons the only parameter to be influenced by
the application of irrigation was the dry matter content of
the tubers. There was a quadratic decrease in dry matter
content as the irrigation rate, hence the soil moisture
increased (Tables 1 and 2).
These trials indicated that cultivars 'Binugas,' Gunung,'
'Seal top,' Forastero' and 'PR-PI 15580' exhibited
tolerance to anthracnose and produced high yields of good
quality tubers. The optimum combination of planting
density (.3m x I m) and size of planting material (115-125g)
have been identified. Under local environmental conditions,
supplemental nitrogen applications to soils testing at levels
of 95 to 116 ppm nitrogen will suppress yam yields. Rainfall
amounts of 511 to 830 mm during a six to seven growing
season appears to be adequate for yam production.
This research was supported in part by the U.S.
Department of Agriculture under Hatch Grant No. 201038.
Agricultural Experiment Station Personnel
Darshan S. Padda....Vice President for Research and Land
Grant Affairs & Director
James Rakocy....Associate Director
Raquel Santiago Silver....Assistant to the Vice President
for Research and Land Grant Affairs
Yvonne Horton....Administrative Assistant III
Audrey Valmont Schuster....Administrative Assistant III
Francis Diaz....Trades Leader
Martin Adjei....Research Assistant Professor
Terry Gentry....Research Analyst I
Osvaldo Lopez,....Agricultural Aide III
Antonio Rodriguez....Agricultural Aide II
Robert W. Godfrey....Research Assistant Professor
Mark Gray....Research Specialist II
Joni Rae Collins....Research Analyst III
Victor Callas....Agricultural Aide II
James Rakocy....Research Professor
William Cole....Research Specialist II
Donald Bailey....Research Specialist II
Kurt Shultz.... Research Analyst II
Ezekiel Clarke....Agricultural Aide II
Manuel Palada....Research Assistant Professor
Stafford Crossman....Research Specialist III
Jacqueline Kowalski....Research Analyst I
Paulino Perez....Research Assistant I
Nelson Benitez....Agricultural Aide II
Reinardo Vasquez....Agricultural Aide II
Forestry and Biotechnology
Christopher Ramcharan.... Research Assistant Professor
Thomas Zimmerman....Research Assistant Professor
James O'Donnell....Research Specialist II
Aberra Bulbulla....Research Analyst II
Jeanette Richards....Research Analyst I
Jeremiah Hassan....Agricultural Aide II
Agustin Ruiz....Agricultural Aide II
Current Research Projects
Evaluation of Forage Conservation Systems in the Caribbean,
Evaluation of Integrated Mechanical and Chemical Control
of Casha (Acacia spp.) on Native Pasture.
Improving Forage Feeding Value by Urea Treatment.
Breeding and Biotechnology for Forage Yield, Quality and
Evaluation of Native Pasture and Agro-By-Product-Based
Systems for Market Lamb Production.
Herbage Allowance and Pasture Rotation Systems for Animal
and Forage Production on Tropical Pasture.
Increased Efficiency of Sheep Production.
Reducing Effects of Heat Stress on Reproduction in Dairy Cattle.
Studies on the Production of Tilapia in Marine Cages.
Evaluation of the Culture Potential of Selected Caribbean
Integration of Tilapia and Hydroponic Vegetable Production
in Recirculating Systems.
Economic Analysis of Integrated Recirculating Systems.
Integrating Tilapia Culture in Tanks with Field Production of
Micro-Irrigation of Horticultural Crops in Humid Regions.
Horticultural and Economic Evaluation of Vegetable Varieties
in the U.S. Virgin Islands.
Alley Cropping Systems for Sustainable Vegetable Production
in the U.S. Virgin Islands.
Improving Crop Management Systems for the Production of
Culinary Herbs in the U.S. Virgin Islands.
Evaluation of Horticultural Practices for Enhancing Root Crop
Production in the Virgin Islands.
Evaluation ofCultural Practices for Sweet Potato Weevil Control.
Evaluation of Integrated Production Methods for Tropical
Evaluation of Minor Tropical and Subtropical Fruits and Nuts
for Production in the U.S. Virgin Islands.
Evaluation of Trees for Agroforestry in the U.S. Virgin Islands.
Potential for Ornamental Pot Crops in the Virgin Islands Using
Bioengineering Plants with the Role gene to Improve Water
Use Efficiency and Drought Tolerance.
Bioengineering Papaya Ringspot Virus Resistance in Carica
papaya for the Caribbean.
Transformation and Regeneration of Hibiscus and
Effects of Bioherbicides on Competitive Ability ofNutsedge.
Biochemical Basis of Resistance of Nutsedge Biotypes and
Species to Nutsedge Rust.
Adjei, M. B. 1994. Component forage yield and quality of
grass-legume cropping systems in the Caribbean. Tropical
Grasslands (in press).
Adjei, M.B. 1993. Desmanthus, a legume for forage
production in the Caribbean. UVI Research, Agricultural
Experiment Station, University ofthe Virgin Islands 5:6-8.
*Adjei, M.B. 1992. The control of casha (Acacia spp.) on
native pasture. Proceedings ofthe Caribbean Food Crops Society,
Santo Domingo, Dominican Republic 28:374-383.
Adjei, M.B. and TJ. Gentry. 1994. Deferment date and
harvest management effects on yield and quality ofgrass/legume
forage bank systems in semi-arid tropics. Agronomy Abstracts:72.
Adjei, M.B., W.F. Brown and TJ Gentry. 1994. The effects
of moisture, urea level and method ofapplication on the chemical
composition and digestibility of native grass hay in the Caribbean.
Proceedings of the Caribbean Food Crops Society, St Thomas,
U.S.V.I. 30:(in press).
Adjei, M.B and W.F. Brown. 1994. Improving feeding value
of guineagrass (Panicum maximum) hay by urea treatment.
Proceedings of the National Conference on Forage Quality,
Evaluation, and Utilization, Lincoln, Nebraska. (abstract).
Adjei, M.B., T.J. Gentry, S.C. Schank and A. Sotomayor-
Rios, 1994. Forage yield, quality and persistence of interspecific
Pennisetum hybrids in the Caribbean. Proceedings of the
Caribbean Food Crops Society, St Thomas, U.S.VJ. 30:(in press).
*Adjei, M.B., K. Albrecht and C. Wildeus. 1993.
Performance ofDesmanths wrgaft accessions in the Caribbean.
Proceedings of the International Grasslands Congress, New
Zealand and Queensland, Australia 17:2129-2130.
*Adjei, M.B. and Pitman, W.D. 1993. Response of
Desmanlhus to clipping on phosphatic clay mine-spoil. Tropical
Adjei, M.B. and W.D. Pitman. 1993. Response of the
temperate perennial grass Reed Canarygrass to defoliation on a
Peninsular Florida Spodosol. Soil and Crop Science Society of
Florida Proceedings 52:1-3.
Bailey, DS., J.E. Rakocy, W.M. Coleand KA. Shultz. 1994.
Economic analysis of two commercial-scale integrated systems
for the production of tilapia and hydroponic lettuce. World
Aquaculture Society, Book of Abstracts: 166.
Brown, W.F., M. Gray, MB. Adjei and R. Godfrey. 1994.
Growth response ofhair sheep fed urea-ammoniated guineagrass
(Panicum maximum) hay. Proceedings of the Caribbean Food
Crops Society, St Thomas, U.S.VJ. 30:(in press).
Cole, W.MN, M.C. Palada, J.E. Rakocy, S.M.A. Crossman,
D.S. Bailey, K.A. Schuttz and J.A. Kowalski. 1994. Integration
of tilapia culture in tanks with field production of bell peppers.
World Aquaculture Society, Book ofAbstracts: 167.
Cole, W.M., J,.E. Rakocy, K.A. Shultz, D.S. Bailey and J.A.
Hargreaves. 1994. The effects offour diets on the survival, growth
and feed conversion ofjuvenile schoolmaster snapper (Lutjanws
apodus). Proceedings of the Gulf and Caribbean Fisheries
Institute-46, Corpus Christi, Texas. (in press).
*Collingwood, C.D, S.M.A. Crossman and M.C. Palada.
1992. Tomato germplasm evaluation for growth and productivity
in the Virgin Islands. Proceedings of the Caribbean Food Crops
Society, SantoDomingo, Dominican Republic 28:232-238.
Crossman, S.MA. 1994. The status oftropical yams. Virgin
Islands Agriculture and Food Fair Bulletin No. 8:29-32.
Crossman, S.M.A., C.D. Collingwood, M.C. Palada and
J.A. Kowalski. 1994. The effect of varying rates of nitrogen
and irrigation on yam (Dioscorea ata L.) production.
Proceedings ofthe Caribbean Food Crops Society, St. Thomas,
U.S.V.I. 30:(in press).
*Crossman, S.MA., M.C. Palada and C.D. Collingwood.
1992. Yield evaluation of sweet potato cultivars in the U.S. Virgin
Islands. Proceedings of the Caribbean Food Crops Society, Santo
Domingo, Dominican Republic 28:533-545.
Godfrey, R.W, M.L. Gray and J.R. Collins. 1994. Theeffect
ofram exposure on uterine involution (UI) and postpartum luteal
function ofhair sheep in the tropics. Journal of Animal Science
Godfrey, R.W. and PJ, Hansen. 1994. Effects of coat color
on production and reproduction of dairy cattle on St. Croix.
Proceedings ofthe Caribbean Food Crops Society, St Thomas,
U.S.VI. 30:(in press).
Godfrey, R.W. and PJ. Hansen. 1994. Seasonal influences
on reproductive patterns of dairy cows in the tropics. Journal of
Animal Science 72(2):13.
Kalmbacher, R.S., W.D. Pitman and M.B. Adjei. 1993.
Growth relationships between Vigna adenantha or V. parkeri
and five perennial bunch grasses. Soil and Crop Science Society
of Florida Proceedings 52:10-14.
Kowalski, JA. and M.C. Palada. 1994 Responseofselected
vegetable crops to saline water in the U.S. Virgin Islands.
Proceedings of the Caribbean Food Crops Society, St Thomas,
U.S.VJ. 30:(in press).
Mislevy, P., M.B. Adjei, F.G. Martin and J.D. Miller. 1993.
Response of US 72-1153 energycane to harvest management
Soil and Crop Science Society of Florida Proceedings 52:27-32.
Mislevy, P., M.B. Adjei, F.G. Martin and G.M. Prine. 1993.
Influence of maturity on quality and agronomic characteristics
of energycane. Proceedings of the International Grasslands
Congress, New Zealand and Queensland, Australia 17:581-583.
O'Donnell, JJ. 1994. Nitrogen fixation- a source of nitrogen
fertilizer. Virgin Islands Agriculture and Food Fair Bulletin No.
O'Donnell, J.J., M.C. Palada, S.M.A. Crossman, J.A.
Kowalski and A. Bulbulla. 1994. Growth and biomass production
from four hedgerow species. Agronomy Abstracts: 72.
O'Donnell, J.J., C. Ramcharan and A. Bulbulla. 1994. The
minor tropical and sub-tropical fruit project on St. Croix.
Proceedings of the Caribbean Food Crops Society, St. Thomas,
U.S.V.I. 30:(in press).
Palada, M.C. 1994. Vegetable production using fish waste
water in the Virgin Islands. Virgin Islands Agriculture and Food
Fair Bulletin No. 8:40-4.
Palada, M.C., S.M.A. Crossman and C.D. Collingwood.
1995. Improving vegetable production using microirrigation in
the Virgin Islands. Proceedings of the Fifth International
Microirrigation Congress, Orlando, Florida. American Society
ofAgricultural Engineers, St. Joseph, Michigan. (in press).
Palada, M.C., J.J. O'Donnell, S.M.A. Crossman and J.A.
Kowalski. 1994. Influence offour hedgerow species on yields of
sweet corn and eggplant in an alley cropping system. Agronomy
Palada, M.C. and S.M.A. Crossman. 1994. Improving crop
management systems for the production ofculinary herbs in the
Virgin Islands. HortScience 29:1058. (abstract).
Palada, M.C., W. Cole, S.M.A. Crossman, J.E. Rakocy and
J.A. Kowalski. 1994. Potential offish waste water as an irrigation
and nutrient source for bell peppers in the Virgin Islands.
HortScience 29:508. (abstract).
Palada, M.C, S.MA. Crossman and J.A. Kowalski. 1994.
Growth and yield response of thyme (Thymus vudgaris L.) to
sources ofnitrogen fertilizer. Proceedings of the Caribbean Food
Crops Society, St Thomas, U.S.V.. 30:(in press).
*Palada, M.C., S.M.A. Crossman, C.D. Collingwood and
JA. Kowalski. 1993. Water use and yield of bell peppers in
hedgerow intercropping with drip irrigation. Agronomy Abstracts:
Palada, M.C, S.MA. Crossman and C.D. Collingwood.
1993. Improving cul inary herb production with drip irrigation in
the U.S. Virgin Islands. UVI Research, Agricultural Experiment
Station, University ofthe Virgin Islands 5:9-12.
*Palada, M.C., S.MA. Crossman and C.D. Collingwood.
1992. Effect of pigeonpea hedgerows on soil water and yield of
intercropped pepper. Proceedings of the Caribbean Food Crops
Society, Santo Domingo, Dominican Republic 28:232-23 8.
Pitman, W.D. and M.B. Adjei. 1994. Response of Shaw
creeping vignaand a perennial alyce clover accession to burning.
Tropical Grasslands 28:53-55.
Rakocy, J.E. 1994. Aquaponics: the integration offish and
vegetable culture in recirculating systems. Proceedings of the
Caribbean Food Crops Society, St. Thomas, U.S.V. 30:(in press).
Rakocy, J.E. 1994. Integrating vegetable hydroponics with
fish culture: promising technology for Caribbean islands.
Proceedings of the Conference on Advances in Tropical
Agriculture in the 20th Century and Prospects for the 21st: TA
2000, Port-of-Spain, Trinidad. (in press).
*Rakocy, J.E. 1994. Waste management in integrated
recirculating systems. Proceedings of the Twenty-First United
States-Japan Meeting on Aquaculture. Bulletin ofthe National
Research Institute of Aquaculture, Supplement 1:75-80.
Rakocy, J.E, D.S. Bailey, W.M. Cole and KA. Shultz. 1994.
Evaluation of two commercial-scale integrated systems for the
production oftilapia and hydroponic lettuce. World Aquaculture
Society, Book of Abstracts: 166.
Rakocy, J.E., J.A. Hargreaves and D.S. Bailey. 1993.
Comparison of tilapiaspecies forage culture in the Virgin Islands.
UVI Research, Agricultural Experiment Station, University of
the Virgin Islands 5:13-17.
Ramcharan, C. 1994. Growth responses of cilantro
(Erynginfoetidum L.) to gibberellic acid sprays. Hortscience
Ramcharan, C. 1994. Physiological responses of 'grande
nine' banana and 'maui' ixora to supraoptimal root-zone
temperature and varying irrigation volumes. Plant Stress in the
Tropical Environment: Proceedings, Kailue-Kona, Hawaii, 27-33.
Ramcharan, C. 1994. Potential for pot crop production using
growth retardants in the USVI. Proceedings of the Caribbean
Food Crops Society, St Thomas, U.S.V.I. 30:(in press).
Ramcharan, C. 1993. Sustainable strategies for reducing
papaya disease problems in the U.S. Virgin Islands. UVI Research,
Agricultural Experiment Station, University of the Virgin
Thomas, T. and M.C. Palada. 1994. The marketing of
medicinal plants in the Virgin Islands; past, present and future
prospects. HortScience 29:558. (abstract).
Wildeus, S. and J.R. Collins. 1993. Hair sheep performance
in an extensive, native pasture system: a five year study. UVI
Research, Agricultural Experiment Station, University of the
Virgin Islands 5:1-3.
*Wildeus, C., M.B. Adjei and S. Wildeus. 1992. Response
ofhair sheep fed silage produced fiom various cropping systems.
Proceedings of the Caribbean Food Crops Society, Santo
Domingo, Dominican Republic 28384-389
Zimmerman, T.W. 1994. Biotechnology leading the way at
UVI. Research Highlights, Research and Land-Grant Affairs,
University of the Virgin Islands, p. 8.
Zimmerman, T.W. 1994. Effect ofexposure to plant growth
regulators on ceDulardifferentiation. Proceedings ofthe Caribbean
Food Crops Society, St. Thomas, U.S.VJ. 30:(in press).
Zinmenuan, T.W. 1994. Papaya ringspot vimrs:a scourge to
papaya production. Proceedings of the Forum on Cultivation,
Production and Marketing of Papaya, Aguadilla, Puerto Rico.
Zimmnunrman, T.W. 1994. Priming papaya seeds reduces seed
germination time. Proceedings of the Caribbean Food Crops
Society, St. Thomas, U.S.V.L 30:(in press).
*An asterisk in front of an entry indicates that it has been
previously listed, but was in press before. Now the entry is