Ecological, Sustainable and Economic Impact of Legume-based Pasture Systems for Limited-Resource Small Ruminant Farmers in the Virgin Islands

Final Report for LS99-107

Project Type: Research and Education
Funds awarded in 1999: $110,410.00
Projected End Date: 12/31/2003
Region: Southern
State: U.S. Virgin Islands
Principal Investigator:
Elide Valencia
University of the Virgin Islands
Expand All

Project Information


Inter-planting improved guineagrass (Panicum maximum Jacq.) cv. Mombasa-lablab (Lablab purpureus; a semi-annual legume) provided high quality browse forage as early as 60-d on rehabilitated pasture and reduced infestation of native hurricanegrass by 90%. Weanling goats on Mombasa grass reached marketable weight (>30 kg) 4-mo earlier than those grazed on native or naturalized guineagrass. During the dry season, lablab provided two-fold increases in herbage yield compared to cowpea and forage soybean, and growing lambs had three-fold higher animal weight gains on lablab-local guineagrass grazing than those supplemented with lablab in a cut-and carry systems. On-farm research provided valuable data for publication of tech-pacs on pasture establishment, management, and also provided an alternative legume to enhance weight gains in critical stages of small ruminant growth. The seeding of lablab when establishing new pastures has now been adopted by farmers and expanded to neighboring islands.

Project Objectives:

1. The overall goal of the project was a multi-disciplinary effort to develop profitable, environmentally safe, and sustainable small ruminant production in the Caribbean. Initial studies assessed means of introducing legumes in degraded pastures and or rehabilitating with improved grasses (e.g., Mombasa).

2. Compare pasture and small ruminant performance on N-fertilized guineagrass, no-till seeded legume hedgerows, and pure guineagrass swards.

3. Compare pasture productivity and economic costs and returns on forage and goat production on no-till planting of legumes (25% of planted area) in native/or rehabilitating with improved pastures.

4. Promote and facilitate adoption of technology of these legume-based systems by forage and livestock producers.


The small ruminant industry is an integral part of the US Virgin Islands’ (USVI; St. Croix, St. Thomas, and St. Johns) agriculture, economic and social activity. Beef cattle and dairy farming on St. Croix are limited to a few large land owners and most small ruminants farms are family-owned and managed by limited-resource farmers. A survey by the US Dep. of Commerce (1995) reported 100 small ruminant farms (ranging from 5 to 20 hectares) with an average flock of 50-100 animals. St. Croix has the majority of small farms that graze or browse on native grasses and leucaena where available, while goat or sheep farms on St. Thomas and St. Johns are maintained on semi-confinement dues to scarce land resources and are fed guineagrass or imported alfalfa hay. The demand for mutton and chevon in the islands has outgrown production as a result of population and income growth. In addition, expanded markets are available for the hotels and restaurants in the tourist islands of St. Thomas and they demand high quality carcasses.

Small ruminant nutrition on the USVI relies on native or naturalized grass pastures as the main sources of feed. Scarcity and low feed value grasses (e.g., hurricanegrass; Bothriocloa pertusa) during drought periods limit expanded production of both sheep and goats on St. Croix. Naturalized local guineagrass (Panicum maximum Jacq.) used in grazing or browsing by goats on St. Croix exhibit seasonal availability, thus causing periods of nutritional stress and reduced animal performance (e.g., slow weight gains and increased susceptibility to parasites).

On St. Croix, adequate dry season feed represents one of the biggest challenges at the small farm level. Sustainable, year-round availability of improved grass cultivars and identification of adequate and manageable browse or graze-able legumes are needed. There is also need to assess the potential feeding value of legumes and compare them to N fertilized guineagrass to determine if they can benefit small ruminant farm profitability.


Materials and methods:

On-farm research were conducted on St. Croix US Virgin Islands (17 43′ N; 64 48′ W). Mean monthly temperature on St. Croix varied form 22 to 28 C and the 20-yr norm annual rainfall was 1030 mm and mean annual rainfall in the 3-yr of the studies average 1000 mm. Soil types used were mainly of the mildly alkaline Fredensborg clay (fine carbonatic, isohyperthermic, Typic Rendolls, Mollisol). Farmer cooperators (both sheep and goat farmers) participated in the development and design of these studies and provided pastures for rehabilitation and animals for conducting the experiments.

For pasture rehabilitation, 10-yr-old local guineagrass stands on medium stage of degradation (<50%) were mob-grazed with lambs and disked lightly. Prepared fields were drilled with lablab and allowed un-interrupted growth for 6-mo. Pastures that were more than 50% degraded were mob-grazed; allowed a 2-wk re-growth and sprayed with glyphosate (1.5%). Fields were then lightly disked and seeded with cv. Mombasa and then seed drilled with lablab. Both guineagrass and lablab growth and spread were measured in permanently marked 2 x 2-m quadrats. Six months after planting, weanling goats of an average of 15 kg BW were assigned to local guineagrass and improved Mombasa pastures and weight gains determined. In the animal feeding studies, forage-based treatments included guineagrass-leucaena, guineagrass-N (56 kg/ha) and unfertilized guineagrass pastures. Treatments were replicated in a randomized complete block design. At the initiation of the rainy season (September) 12 one-acre paddocks of guineagrass were grazed by lambs to a 10-cm stubble height. Four paddocks were no-till direct seeded with leucaena cv. Cunningham so that legume hedgerows cover 25% of the pasture. Glyphosate herbicide was banded on the planted row at a rate of 0.75 L /ha to suppress guineagrass growth at the time of planting. Leucaena was planted in a skip-row pattern of hedgerows (2 rows 1-m apart every 6 m). The remaining eight pastures will be maintained as pure stands of guineagrass. Prior to the initiation of the study, paddocks with the established leucaena hedgerows and pure guineagrass stands were mob-stocked with sheep to a 10-cm stubble height. Guineagrass-N experiment was fertilized with 56 kg/ha N, P and K based on soil recommendations for guineagrass. Legume hedgerows were fertilized with 25 kg/ha P. The experimental units were subdivided in five paddocks (0.20 ha) using a solar powered electric fence. Thirty five days after fertilizer application, subdivided paddocks were rotationally stocked with weaned lambs (600 kg live-weight ha/d) 7-d grazing and 28-d rest period.

Research results and discussion:

Slight disking and seed drill of lablab on degraded pastures had minimal effect in increasing local guineagrass plant population. Lablab, however, provided a fast short term feed for goats within 60-d of planting and this help cover the cost of land preparation and seed costs. Inter-planting improved guineagrass cv. Mombasa-lablab resulted in full stands 6-mo after planting. Lablab provided high quality browse forage as early as 60-d after planting and reduced regeneration of native hurricanegrass by over 90%. Re-growth of lablab was affected by competition of the growing grass and contributed little forage thereafter. The average daily gain of goats grazing Mombasa was 73 g/d compared to 54 g/d on the local guineagrass. Weanling goats on Mombasa grass reached marketable weight (>30 kg) 3 to 4-mo earlier than those grazed on native or naturalized guineagrass.

The potential for improved weight gains by growing hair sheep and dry season yield of guineagrass (GG; Panicum maximum Jacq.) Leucaena (L; Leucaena leucocephala Lam de. Wit) was investigated. St Croix White hair lambs (4 mo. of age) were utilized to compare two different grazing methods, with three replicates within each treatment. In continuous grazing (CG) each replicate of lambs (n = 5) grazed GG L (10 15% over story) pastures (.14 ha) at a stocking rate of 650 kg BW/ ha/d for a period of 98 d. In rotational grazing (RG) each replicate of lambs (n = 5) grazed guineagrass L (10 15% over story) pastures (.14 ha) at 650 kg BW/ ha/d for 126 d. In the RG, each .14 ha pasture was subdivided into 3 equal paddocks and lambs were moved every 14 d which allowed each paddock a 28 d rest period. Lambs were weighed weekly throughout the trial. Forage dry matter yield was estimated in five .25m2 areas in each paddock. Pastures were sampled every 28 d in CG and every 14 d in RG. There was a trend (P = .09) for higher average daily gain for sheep on CG (68 ± 3.7 g/d) compared to those on RG (46 ± 3.7 g/d). However, seasonal forage yield was favored (P< .05) by RG (3.5 ± .4 Mg/ha) compared to CG (1.8 ± .4 Mg/ha). At season end, sheep on CG had to be supplemented with leucaena on a cut carry system (1.6 kg DM/d) as regrowth of leucaena was affected. Grazing selectivity increased under CG, but also favored ingress of the low quality hurricane grass (Bothriocloa pertusa). These results suggest that weight gains are slightly better under CG, but will require supplementation during the dry season. Forage yield during the dry season is better under RG and may favor the grass legume over story in the long term. This study assessed legume supplementation with leucaena (LS) and N fertilization of guineagrass on the average daily gain (ADG) of St Croix white hair sheep, herbage mass (HM, Mg/ha) and nutritive value. Treatments consisted of N rates (0 or 56 kg/ha; N0 and N56) broadcast-applied on shredded guineagrass stands (15-cm stubble height) at the initiation of the rainy reason (September). St. Croix white hair lambs with an average weaning weight of 13.6 kg (n = 4) were assigned to replicated paddocks (.14 ha). Guineagrass pastures were stocked at 760 kg body weight ha/d for 21 d followed by a 21 d rest period. Legume source was fed three times a week at 0.68 kg head/d (dry matter basis). Herbage mass was estimated in three 0.25 m2 areas in each paddock every 21 d. Lambs were weighed weekly throughout the trial. All data were analyzed using analysis of variance procedures. Both N56 and LS did not affect HM (3.2 Mg ha1), but HM on N0 fell below 1.5 Mg/ha at 14-d grazing limiting grass on offer and requiring cut-grass supplementation for 7 d. Crude protein concentration (8.5%) and in vitro organic matter disappearance (54%) were not affected by treatment. Three was a trend (P=.09) for higher ADG on LS (70 g/d) compared to N56 (60 g/d) and differed (P<.05) from N0 (57 g/d). These results suggest that sheep weight gains can be improved with LS supplementation and this maybe more economical than N fertilization on intensively managed guineagrass pastures. In a follow-up study, the feeding value effect of lablab on average daily gains (ADG) of weaned lambs (15 kg BW) during the dry season was investigated. Treatments were unrestricted grazing of guineagrass supplemented with lablab (SL; 0.5% animal live-weight on a dry matter basis), unrestricted grazing of both guineagrass and lablab (GL), and unrestricted grazing of guineagrass (control) in a randomized complete block with two replicates. Lambs (n = 4) grazed pastures (.14 ha) and were supplemented for a 96-d period, after a preliminary adjustment feeding period of 14-d. Lambs were weighed weekly throughout the trial. Herbage mass (HM) was estimated in five .25m2 areas in each paddock every 21-d. Data were analyzed using GLM procedures of SAS and mean separation when significant was conducted with LSMEANS. There were significant differences (P<.05) among treatments for ADG. There was a four-fold increase in ADG for GL (72.8 ± 6.2 g/d) compared to the control (18.5 ± 6.9 g/d). Average daily gain of lambs on GL was also much higher than SL (36.4 ± 6.4 g/d). There were difference among treatments for HM (P<0.05). At season end, HM on offer for GL (3.64 ± .3 Mg/ha) was two-fold higher than SL (1.9 ± .3 Mg/ha) and control (1.3 Mg/ha). These results indicate that weight gains of St. Croix white hair can be increased with summer grazing or supplementation of lablab. The weight gains observed with unrestricted grazing of lablab justify its establishment as a special purpose pastures for use with weaned lambs during the dry season. In our continued search for forage legumes to increase seasonal forage distribution and quality, desmanthus and glycine (perennial soybean) were inter-planted in degraded pasture stands, but their growth and persistence in this association was limited to les than 1-yr. Introduction in strips with the planting of Mombasa was also not successful because of the aggressive growth of the grass and thus legume forage contribution was minimal. Both legumes may have potential in solid stands and when managed as protein banks under rotational stocking.

Participation Summary

Educational & Outreach Activities

Participation Summary

Education/outreach description:

Valencia, E., and R.W. Godfrey. 2001. Grazing method effects on growth rate of St. Croix White hair sheep lambs on a tropical grass-shrub legume over-story. J. Anim Sci. 79:59.

Valencia, E. 2002. Legume and N fertilization effects on growth of St. Croix White hair sheep on guineagrass (Panicum maximum Jacq.) pastures. p.267. In: Proc. of the American Forage Grassland Council. Bloomington, Minnesota. July 14-17, 2002.

Valencia, E., R.W. Godfrey, and S. Weiss. 2003. Grazing and supplementation effects of Lablab (Lablab purpureus) on weight gains of St. Croix White hair sheep lambs during the dry season. J. Anim Sci. 86(1):353

Valencia, E. and G. D’Souza. 2002. Factsheet: Pasture establishment in the U.S. Virgin Islands. University of the Virgin Islands, Agriculture Experiment Station Publication. 4 p.

Valencia, E., L.E. Sollenberger, and G. D’Souza. 2002. Factsheet: Managing pasture resources. University of the Virgin Islands, Agriculture Experiment Station Publication. 4 p.

Two forage workshops were conducted each year in St. Thomas and St. John to facilitate and provide information on nutritional needs of sheep and goats. On St. Croix, forage field days, workshops, and pasture walks were held to disseminate information to small ruminant producers, extension agents, and local media. More than 100 producers and extension personnel participated in these activities.

Project Outcomes

Project outcomes:

1. This research demonstrated a viable method of grass establishment with the inter-planting of a semi-annual legume. Earlier studies in pasture establishment showed that when planting grasses (small seeded type) in minimum till or prepared seed–beds the competitive effective of existing seed reserves of the native grasses. In this research, Mombasa and lablab established well with little competition from hurricanegrass and provided an early feed source for animals.

2. The legumes leucaena and lablab can be used efficiently and economically in different feeding system, and that animal weight gains were better with legumes than N fertilized guineagrass or guineagrass alone. With the provision of low-input and high quality legume forages, resource-limited small ruminant farmers have been able to up-grade their sheep and goat-herd.

3. Sound management practices to favor grass persistence and legumes in the farming systems used in the Virgin Islands were developed. Our research showed that in limited land area and with either sheep or goat, rotational grazing should be used to maintain a productive pasture. Farmer co-operator farms, today, serve as a model for managing productive grass-legume pastures.

Goat farmers on St. Croix have upgraded their herds with pure-bred Boer goats to cross with the Spanish-native goats. Sheep farmer introduced Dorper Rams to cross with the St. Croix white hair sheep for improved hybrid vigor. An upgrade in the small ruminant nutrition with use of Mombasa and lablab should favor these crosses and production of better quality carcasses.

Economic Analysis

To compare cost and revenue differences, the following information were collected (cost of fencing, land preparation, seed costs and fencing). Sales from lambs and goats produced were recorded. Analysis shows that the returns from the second year of sales from animals covered initial investment of pasture establishment. The projected life of the pastures with the sound management recommended to farmers was 10-yrs.

Farmer Adoption

Farmer cooperators were an integral part of the project from writing to implementation of the research. Three goat farms participated in this research and invested in rehabilitation of pastures to new and improved grasses such as Mombasa, Tanzania, and Brachiaria brizantha cv. Marandu. They market goats for meat and also provide breeding stock to the other islands. Also, these farms serve as model farms for use by the Cooperative Extension Service. Fact-sheets produced under this project serve as guidelines for establishment and management of improved pastures. Farmers, Senior Research and Industry as well as Department of Agriculture personnel visited or participated in the
Numerous field days conducted at these research sites. Fact-sheets were published and made available to livestock farmers at the yearly Agricultural and Food Fair.


Areas needing additional study

Although this research was able to provide an alternative and adapted grass for the climatic conditions of St. Croix for grazing, research on conservation methods (e.g., hay) are needed to meet the needs of the other islands of St. Thomas and St. Johns. Follow-up outreach and training by the Cooperative Extension Service should continue and addressed under the PDP SARE grant. Lablab is semi-annual legume and can be used as a high protein feed source. Seed production to meet the needs of farmers rehabilitating their pastures should be addressed. The potential of browse tree legumes such as Moringa Oleifera and Pigeon peas to provide feed for the dry season should be investigated.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.