Integrating tropical legumes with condensed tannins into ruminant grass-based diets for sustainable production

Final Report for LS08-212

Project Type: Research and Education
Funds awarded in 2008: $100,000.00
Projected End Date: 12/31/2012
Region: Southern
State: Puerto Rico
Principal Investigator:
Dr. Elide Valencia
University of Puerto Rico, Mayaguez
Expand All

Project Information


Gastrointestinal nematodes (GIN; Homonchus contortus) account for the greatest the economic loss to small ruminant farmers in Puerto Rico and US Virgin Islands. Legumes can increase animal performance and condensed tannin (CT) containing forages exhibits anthelmintic properties. Use of CT legume calliandra had positive GIN control in Puerto Rico (fresh and dry), but calliandra was not effective in the US Virgin Islands. Monitoring anemic animals and feeding of CT legumes (deworming banks) can reduce anthelmintic use. Perennial peanut, Morus and pigeon pea can improve small ruminants nutrition in the different agro-ecological zones of Puerto Rico.

Project Objectives:

Research objectives were: (i). Determine intake, digestibility and nitrogen balance of lambs fed diets of tropical grass hay (eg. blue stem; Dicanthiumannulatum or guineagrass; Panicummaximun Jacq.) supplemented with tropical legumes with condensed tannin (CT) [perennial peanut (Arachis spp.); desmodium (Desmodium heterocarpum(L.) DC. Subsp. ovalifolium (Prain.) cv. Maquenque, calliandra (Calliandra calothyrus)] and pigeonpea (Cajanus cajan), (ii). Determine average daily gain and parasite load of lamb-crops creep grazing CT tropical legumes (perennial peanut, desmodium, calliandra and pigeon pea), (iii) On-farm demonstrations of CT legumes and their effects on parasite control on small ruminants, and (iv). Prepare economic costs and returns of forage and small ruminant production on conventional dewormers vs. dewormer banks.


In Puerto Rico, urban spawn has forced beef producers to diversify (increased sheep and goats farms) and move pastures to less productive marginal lands. Ruminant production are crucial to the economies of both islands and are reliant on tropical grasses, though abundant in the wet season, are of low quality and cannot meet the nutritional requirements of growing ruminants. Concentrates typically used to bridge this nutrient requirement gap are imported at considerable costs ($20 per 50lb of concentrate feed) are not economically viable. In the emerging small ruminant industry (Mutton and Chevon sells for $4-5lb in Puerto Rico), producers are now faced with parasitism of gastrointestinal nematodes (GIN) resulting in marked production losses pre- and post weaning, in addition to the developed resistance to anthelmintic drugs by GIN. Creep grazed or conserved tropical legumes containing condensed tannins (CT) to provide supplementary nutrients may reduce parasite loads and will likely enhance growth rates and increase profitability, but these hypotheses needs to be verified in on farm research trials.


Click linked name(s) to expand
  • Dr. Gerard E. D'Souza
  • Stuart Weiss


Materials and methods:

Study 1, was conducted at the University of Puerto Rico, Agricultural Experiment Sub-Station of Corozal during the period between April-November, 2009 and repeated in 2010. St. Croix White hair sheep growing lambs [n=5; 15 kg body weight (BW)] and replicated twice were used for this study. There were two treatments; treatment 1 consisted of daily and night rotational grazing of an already established guineagrass (Panicum maximum Jacq.)– perennial peanut (PP; Arachis pintio) stand during the period between 1:00PM to 7:00AM and supplemented with guineagrass hay (3% BW) during the day, while treatment 2 consisted of rotational grazing of pure guineagrass stands during the same period. Mineral salt block was offered ad lib during the morning. Weekly, lambs were offered 200g/animal of the condensed tannin (CT) containing Calliandra calothyrsus shrub and weight taken. FAMACHA to assess anemia was done biweekly and lambs with a score above 3 were treated with a de-wormer. A follow study was undertaken to assess chemical composition, in vitro degradability and intake of the pigeon pea, perennial peanut (Arachis glabrata), and Stylosanthes guianenesis. Diets included blue stem hay (50%) and legumes (50%) and fed at 3.5% of goat body weight.

A second study conducted at the Corozal Experiment Station assessed live weight performance of pre- and post weanling milk Nubian goats creep grazing CT Calliandra calothyrsus and followed by evaluating two contrasting diets using dried and ground Morus alba (MA) produced in situ within the farm compared to goat concentrate feed. Twelve kids (2-wks old) of Nubian milk goats creep grazed perennial peanut and CT Calliandra for 3-mo and repeated over time.

A third study was conducted at the Agriculture Experiment Station of the University of the Virgin Islands, on St. Croix to evaluate condensed tannin containing legume calliandra for its ability to successfully establish a perennial hedgerow for creep grazing and in cut-and-carry feeding systems on average daily gain and as a natural anthelmintic. In October 2008, calliandra was germinated from seed in the greenhouse and young plants were transplanted into an alley hedgerow at the in April 2009. The hedgerow was allowed to establish for approximately 18 months. The first creep grazing trial with St. Croix White Hair x Dorper (STXD) lambs began on October 13, 2010, but ended 4 weeks into the trial due to extreme internal parasite loads resulting in heavy animal attrition and the lambs de-barked the calliandra trunks which had detrimental effects upon the calliandra forage bank. On June 15, 2011 the second experiment began that consisted of two treatment groups; the control treatment consisted of lambs (n=18) grazed on native guinea grass (Panicum maximum) pasture and the experimental treatment consisted of lambs (n=18) grazed on native guinea grass that were supplemented 3 times per week with calliandra that was cut and carried to the lambs and fed in a quantity greater than lamb intake levels. Data was collected on individual animals over the course of the 6 month trial at 30 day intervals. To quantify internal parasite load, FAMACHA© scores, fecal egg count (FEC), and packed cell volume (PCV) was determined; and animal performance was measured by total weight gain and average daily gain (ADG).

Finally a demonstration study was conducted at one farmer location (Guillermo Martinez) and took place during the 8-week period between 16 April and 11 June 2011 and had as objectives the evaluation of three supplemental forage species in terms of animal consumption, estimated quality and live weight gain by young lambs. The animals grazed daily in guineagrass plots during the hours from 7:00 am to 2:30 pm. Thereafter, they were placed in individual stable cages and fed cut whole-plant forage according to the following treatments: MA (Chopped Morus alba), CC (Chopped calliandra) and guinea grass as control. All the chopped forages were following a 12-wks regrowth after previous harvest. The quantity of forage provided daily under each treatment was determined on a dry matter basis, and was equal to 2.0% of animal body weight. Two young lambs (each weighing 16 to 18 kg) were placed in each cage. The level of forage consumption of each treatment was determined during each of four 2-week cycles of grazing. The animals remained on the same treatment throughout the experimental period. At the end of each cycle the animals were individually weighted during the morning hours. During the first week of each cycle the animals were considered to be adapting to the new daily offerings of supplemental forages. During the second week daily 2011 samples of the offered forage were taken to produce a 7-day composite for each treatment. The composites were air-dried in an oven at 65° C for a period of two weeks. For sample evaluation, the 30 April to 13 May, and 14 May to 28 May cycles were chosen for representative estimations of forage quality criteria.

Research results and discussion:

In study 1, there were differences on average daily gain (ADG). Lambs on the legume association gained 62 g/d compared to 30 g/d on grass alone (a two-fold increase in ADG). Lambs on guineagrass-PP required less deworming (2 xs during the period of the study compared to 5x those on grass alone). The CT legume appeared to have minimum effects when offered on the low quality guineagrass. In the follow up study we noted that digestibility was higher for diet containing the legumes than grass hay alone. The NDF digestibility was lower however for both pigeon pea and stylosanthes because of the higher lignin values (10%) observed for these two legumes and also a lower leaf component compared to the perennial peanut (3.5% lignin). Intake was also higher for perennial peanut and stylosanthes (1200 g/d ) vs. pigeon pea (956 g/d). Browsing with growing kids, however, did not how differences in selectivity for either of the legumes.

In the second study, there were no differences on weaning weight (17.5 kg BW). However, this weaning weight surpasses the traditional weaning on non-creep grazed legumes by 5 kg. At post weaning, kids (N=3) were randomly assigned to a semi-confinement feeding systems of a free choice offer of bluestem grass hay supplemented with ground MA (harvested at 3-mo regrowth and sun-dried for 5d) at 3% BW and a free choice offer of bluestem hay plus a goat concentrate (600 g/kid for the first 3 mo and 800 g/kid for the additional 3 mo.), a management practice used by the farmer for developing young goats. Animals allowed pasturing 5 h daily. Kids were weighed weekly. There were no differences between feeding systems. Average daily gains on supplemented MA were 58 g/d compared to 62 g/d on those fed with concentrates. Breeding stock was marketed at 9-mo age.

In third study, animal growth and performance was higher for lambs supplemented calliandra where lamb total weight gain resulted in 5.4 kg and an ADG of 44 g compared to the control lambs on native pasture without supplementation that gained a total of 3.1 kg and had an ADG of 26 g. Results from the FAMACHA scores, FEC, and PCV indicate that there were no significant differences in gastrointestinal parasite infections or anemia levels in lambs from either treatment at any point during the 6-month trial. However, the FAMACHA© eye score system did prove to be a reliable indicator of anemia, validated through PCV (negative correlation) and of GIN population, validated through FEC (positive correlation) in STXD lambs. Calliandra can increase animal growth and performance, but did not exhibit anthelmintic qualities when administered through a cut and carry supplemental feeding system.

On-farm demonstrations of CT legumes and their effects on parasite control on small ruminants in Puerto Rico and US Virgin Islands. Farmer Cooperators: Guillermo Martinez, Vicente Ortiz and Tito Palmer, Vega Baja, Barceloneta and Guajanilla, Puerto Rico, respectively participated in the on-farm research. Between July-December 2009, 2010, and 2012 protein banks of Morus alba (MA) and a deworming bank. Four hundred seedlings of MA and CT Calliandra were planted in a 0.20-area. At 4-mo post-planting, 17adult lambs grazed both banks, and shrub plants were cut at 15-cm height to induce tillering. In addition, a 0.20-ha area of guineagrass and PP association was also established. Sheep flock was bred to lamb in the fall of 2010. Following the lambing period, a creep grazing study using MB, CT Calliandra and the association was conducted post-weaning. Lambs on both legumes had higher weaning weights (7.5 kg) compared to these on grass alone at all three farmer cooperator farms.

On farm demonstration showed significantly higher proportional consumption of MA than of CC and guineagrass was observed for the mean of all four experimental cycles. During the individual cycles MA was consistently preferred over the other two forages; however, in the first and forth cycles these differences were not different statistically. Significantly higher CP content was observed in MA in relation to CC and guineagrass by respective margins of 3.4 and 5.0 percentage units. Although no significant differences were observed in the ADF content of the three forages, both MA and CC had considerably lower valves than guineagrass. The NDF contents of MA and CC were very similar and significantly lower than that of guineagrass. Although no significant differences were verified between the treatments, given the small number of observations in this experiment and the high variability in the young lambs’ growth rate, MA presented the highest rate of gain (60.8 g/day) over the eight weeks of experimentation. This result is logical given the better forage quality, higher consumption and higher net energy potential of MA. Supplementation with CC, which was shown to be a forage of lower quality level than MA, resulted in a slower growth rate of 23 g/day, while supplementation with guineagrass the poorest quality forage, led to a loss of live weight, which was not a surprising result. Given these promising results, more research is warranted on the use of the MA and CC shrub under more extensive commercial small ruminant production systems. MA showed promise as supplemental forage for young lambs grazing relatively poor pasture, having the highest level of forage quality in terms of chemical composition, estimated not energy level, and highest forage consumption. Although the differences were not significant under the conditions of this experiment, MA can be expected to outperform CC and guineagrass in promoting live weigh gain when used as supplements.

Participation Summary

Educational & Outreach Activities

Participation Summary

Education/outreach description:

During the course of the grant period four students were on assistantships to conduct thesis research within a graduation rate of 2.5 years. In addition, several papers were presented at regional and national meetings.


J.P. Muir, T.H. Terrill, E. Valencia, S. Weiss, K. Littlefield, P.D. Jones, J. Mosjidis, and R.M. Wolfe. 2008. A wide range in forage condensed tannins in the southeastern USA shows promise in ruminant protein and parasite management. In: Proc. of the American Forage Grassland Council (AFGC). AFGC, Grand Rapids, Michigan, June 2009. CD-Rom.

Valencia, E., B. R. Min, J. Muir, K. Quesenberry y A. Rodríguez Carías. 2009. Concentraciones de taninos condensados en leguminosas tropicales y el efecto de extractos de estas en producciones de gas metano en rumiantes. In: Resumen de la Asociación Latinoamericana de Producción Animal. Isla, Verde, Puerto Rico, Julio 2009.

Valencia, E., R. Ramos-Santana and A. Rodriguez. 2009. FAMACHA and shrub legumes to avoid anthelmintic resistance in small ruminant farms in the Caribbean basin. In: Proc.of the 45th Annual Meetings of the Caribbean Food Crops Society, St. Kitts. July 2009

Santos, A., and E.Valencia. 2010. Season and date of harvest effects on biomass production of sunn hemp (Crotalaria juncea L.) in Puerto Rico. Abst. American. Soc. Agron. Cd-ROM.

Ramos Santana, R., E. Valencia, and R. Machiavelli. 2010. Consumption and chemical composition of shrub and herbaceous species with potential to feed small ruminants under grazing conditions. J. Agric. Univ. PR. 94 (3-4):275-278.

Ramos Santana, R., E. Valencia, and R. Machiavelli. 2010. Morus alba and Hibiscus sinensis fed as whole-plant supplements to growing lambs on guineagrass. J. Agric. Univ. PR. 94 (3-4):279-284.

Zavala, D., E. Valencia, P. Randel and R. Ramos-Santana. 2011. Rendimiento y características fermentativas de ensilajes de maíz (Zea mays)-leguminosas anuales lablab (Lablab purpureus L.) y Sunn Hemp (Crotalaria juncea L.). J. Agric. Univ. PR (accepted).

Zavala, D., E. Valencia, P. Randel and R. Ramos Santana. 2011. Producción de ensilaje de maíz blanco (Zea mays) de alto valor proteico (QPM) con y sin mazorca asociado con dos leguminosas anuales Lablab (Lablab purpureus L.) y Sunn hemp (Crotalaria juncea L.). J. Agric. Univ. PR (accepted).

Santos, E. Valencia, E. Román y R. Ramos. 2011. Producción de biomasa y contribución de nitrógeno de Sunn hemp (Crotalaria juncea L.) en Lajas, Puerto Rico. J. Agric. Univ. of Puerto Rico (accepted).

Project Outcomes

Project outcomes:

In addition to three farmer cooperators participating on the on-farm research, 50 other small ruminant farmers have benefited from the use of MA and CT calliandra. The local Department of Agriculture, Puerto Rico provided additional funds to our program to prepare seedlings to be distributed to other small ruminant famers throughout the island. The results of the economic analysis should provide farmers and extension specialists with an idea of expected costs and returns during the establishment and subsequent years if they were considering switching from a conventional to an innovative system. It is quite conceivable that, given a sufficiently long time horizon, and adoption of one or more of the innovative cost management and/or pricing strategies listed above, the innovative production system could out-perform the conventional system in terms of both profits and risk.

Economic Analysis

An economic costs and returns of forage and small ruminant production on conventional dewormers vs. dewormer banks were prepared. We compared the following two meat goat systems: (1) Conventional production, using concentrate feeds and deworming; and (2) Innovative production, integrating a Morus Alba and calliandra dewormer bank. Using a combination of on-site experimental data provided by the agronomist and extension specialist on the project, informal farmer interviews, and price quotes provided by agricultural supply stores, we developed a series of four enterprise (or cost and return budgets), one for each system and, within each system, a separate budget for the establishment year and for subsequent years.

As the attached budgets show, net returns to both systems are negative during the establishment year (to be expected) as well as during subsequent years (contrary to expectations). Although the establishment costs are similar in both systems, the subsequent years’ returns are greater (or less negative) in the innovative system, indicating that, under the right conditions, this system could be profitable. These conditions include: (a) looking at a larger scale than that investigated in the study, due to the potential for advantage of economies of size; (b) direct marketing meat to restaurants and the public, which could result in a higher product (meat) price (for example, selling shrink-wrapped packages at a central location such as Plaza de las Americas, could generate a price of $12.50 – 13/kg, much higher than the $3.85/kg assumed in the analysis); (c) organizing a cooperative among small-scale goat farmers to take advantage of purchasing discounts on feed and the goats themselves (which together account for the largest proportion of operating costs for the innovative system); and de-wormer and concentrate purchases (together accounting for the largest proportion of operating costs in the conventional system). We also found that break-even prices were lower for the innovative system compared to the conventional system ($6.65/kg versus $12.34/kg).

Given the lack of data for the economic analysis, we used fairly conservative estimates in our calculations, thus these results can be viewed as representing a most likely to worst case scenario. If additional years of data as well as additional replications in other locations were available, they would have been useful in providing insights into riskiness of cash flows across time and space. In addition, given the negative cash flows during the short period investigated, it is not possible to calculate measures of investment feasibility such as Net Present Value or Internal Rate of Return. Finally, given the high degree of inbreeding in the present herd, it would be useful to determine by how much genetic improvement boosts economic performance (in terms of both profit and risk measures).

Farmer Adoption

Through field days conducted at the Corozal Agriculture Stations, Puerto Rico in May of 2009 and 2010 and at the Isabela Agricultural Experiment Station in May 2011, farmers had training on CT legume use and for creep feeding. As a result of requests for planting material, the local department of agriculture approved a three-grant to the Corozal Agriculture Experiment Station to prepare seedlings of calliandra and Morus for establishing deworming banks and protein banks and enhanced its use by small ruminant farmers. To date over 50 small farmers expanded in the use of CT legumes and benefit form the research grant.


Areas needing additional study

The use of these CT legumes and others need to be assessed in larger systems production and also determine carcass quality of animals produced through a SARE producer grant. Record keeping and continued survey with forms developed through our research and education grant are invaluable and can be used for education and extension grants.

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.