:‘wl z,A.-‘.x..~ :1“ mm 4x, .. ulna-.- .. .. . ~,: ‘ WWII”I“UHH“"lWWI”WWW 3 1293 009011 This is to certify that the thesis entitled ESTABLISHMENT AND NUTRITIVE VALUE OF NATIVE AND EXOTIC FODDER TREE SPECIES IN JAMAICAN PASTURE SYSTEMS presented by James Michael Roshetko has been accepted towards fulfillment of the requirements for Master degree in Forestry Major professor Date __ _.Se;1temb.e_L4_,_19.91 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution r \ LIBRARY Hickman State L Unlverelty PLACE IN RETURN BOX to remove thle checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Afflnnatlve ActiorVEquel Opportunity Institution emails-n1 ESTABLISHMENT AND NUTRITIVE VALUE OF NATIVE AND EXOTIC FODDER TREE SPECIES IN JAMAICAN PASTURE SYSTEMS BY Jamel Michael Roshetko A THESIS submitted to Michigan State University in partial fulfillment of the requirements for the degree MASTER OF SCIENCE Department of Forestry 1991 53?“??75” ABSTRACT ESTABLISHMENT AND NUTRITIVE VALUE OF NATIVE AND EXOTIC FODDER TREE SPECIES IN JAMAICAN PASTURE SYSTEMS 3? James Michael Roshetko An experiment ‘was conducted to investigate fodder tree establishment in Jamaican grass pastures. Results demonstrated that Gliricidia and Leucaen; can.be established by direct seeding, and that adequate weed control improves fodder tree survival and growth. Circle weeding greatly enhanced survival and growth over bush weeding alone. Fertilization and larger planting pits had little positive effect on fodder tree performance. The use of these treatments is not recommended. In a second experiment the approximate nutritive value of native and exotic Jamaican fodder tree species were determined and compared to the approximate nutritive values of the main pasture grass species used in Jamaica. In general, tree species are inferior to grass in digestibility and mineral content, but superior in crude protein content. These characteristics make tree fodder a good livestock feed supplement for the dry season when grass or of poor quality. ACKNOWLEDGEMENTS This thesis is the result of two and a half years of work in both Jamaica and East Lansing. During that period many people gave me support and advice for which I am genuinely grateful. Drs. Douglas Lantagne and Michael Gold were instrumental in helping me design and complete my research project. Even more importantly their advice and tutelage instilled in me an appreciation for the "learning process" of a thesis program. I certainly would not have benefitted so much from this process if it had not been for them. Drs. Oran Hesterman and Robert Deans both served on my graduate committee and deserve my thanks. Their comments and advice were very useful in improving my research design, analysis, and write-up. Also, I would like to thank Drs. George Wilson and Lyndon McLaren, as well as the entire Jamaican.Agricultural Research Program staff for their steady and cheery support during my year residence in Jamaica. Additionally, Mrs. Gayle, Knight, Malcolm, Simmonds, and Dr. Wellington of Alcan Jamaica Co. , Agricultural Division deserve my thanks for the logistical and technical support they gave me during my study. I am also indebted to my parents, Don and Dolores, and my siblings, Tim, Lois, Debby, Tom, Mary, and Ellen, for their material, moral, and emotional support. Their encouragement and love have been with.me since my earliest memories. I feel truly blessed to have them as family, and hope that I am as supportive of them as they have always been of me. Finally I would like to thank.my follow graduate student and other friends in East Lansing. Their comradery, friendship, and spirited engagement of Friday Afternoon Club activities have made my time here truly enjoyable. ii TABLE OF CONTENTS LIST OF TABLES........................................ v LIST OP PIGURES....................................... Vi INTRODUCTION.......................................... 1 Chapter 1. EFFECT OF NEED CONTROL, EERTILIZATION, AND PLANTING PIT SIZE ON SURVIVAL AND GROWTH OF DIRECT SEEDED AND SEEDLING TRANSPLANTS OF CALLIAND , GL RICI , AND LEUCAEEA...... 3 Abstract.................................. 3 Introduction.............................. 3 Literature Review......................... 6 Materials and Methods..................... 11 Results................................... 1 18 Discussion................................ 26 Conclusions....................... ...... .. 32 Acknowledgements.......................... 34 Bibliogratheeeeeeeeeeeeeeeeeeeeeeeeeeeeee 35 Chapter 2. DIRECT SEEDING OF FODDER TREE SPECIES IN JAMAICAN PASTURES......................... 43 Abstract..................... ....... . ..... 43 Introduction.............................. 43 Materials and Methods..................... 46 Results and Discussion.................... 48 Conclusions............................... 52 Acknowledgements.......................... 53 Bibliography.............................. 54 iii Chapter 3. ESTIMATED NUTRITIVE VALUE OF NATIVE AND EXOTIC FODDER TREE SPECIES IN JAMAICAN AND THEIR COMPARISON TO PASTURE GRASS SPECIES. . Abstract................................... Introduction............................... Materials and Methods...................... Results and Discussion..................... Conclusions................................ Acknowledgements........................... Bibliography............................... SUMMARY”............................. ..... ............ DIRECTIONOF FUTURERESWCBOOOOOOOOOOOOOOOOOOOOOOOOOOO iv 59 59 59 63 67 78 79 80 87 89 LIST OF TABLES Table Page Chapter 1. 1 ANOVA tables of survival data for direct seeded Glizigidi§_and L_Q§§gng.......................... 2 Leucaega survival and growth by treatment, ten months after seeding............................. 3 gliricidia survival and growth by treatment, ten months after seeding............................. 4 ANOVA tables of survival data for transplanted seedlings of Leugagng and Calliand;a............. 5 Leugaena survival and growth by treatment, six months after transplanting ....................... 6 Calliandra survival and growth by treatment, six months after transplanting....................... Chapter 2. 1 Germination (at three months), survival, height and diameter growth of study species at ten months... Chapter 3. 1 Composition of tree fodder species. .............. 2 Composition of Jamaican grass species............ 3 Daily nutrient requirements of diary cattle...... 4 Relationship between fiber parameters and alfalfa feed qualitYeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 20 21 21 23 24 24 50 68 70 70 73 LIST OF FIGURES Figure Page Chapter 1 1a Main block layout of weed control and fertilizer.. 15 leveISeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 1b Main block layout of planting pit size levels..... 15 2 Effect of weed control x fertilization interaction on Leucaena basal diameter growth six months after transplanting...GOOOOOOOOOOOOOIOOOOOOOOOOOOOOOOOO. 25 vi INTRODUCTION The main premise of this thesis is that fodder trees are extremely valuable livestock feed resources, which are presently under-utilized in Jamaica. Given appropriate testing and investigation, tree fodder could be incorporated into the Jamaican livestock production system particularly to improve dry season feed availability. On-farm production of tree fodder resources is not capital intensive. Thus, tree fodder production is specially well suited for incorporation into the farming systems of small-scale farmers who have limited financial resources. The text is divided into three chapters. Chapter 1 discusses the impact of soil cultivation, fertilization, and weed control as establishment techniques for fodder trees in grass pasture ecosystems. Chapter 2 discusses the evaluation of fifteen different exotic fodder tree species for potential incorporation into the pasture component of the Jamaican livestock production system. Chapter 3 presents approximate analysis nutritive values for the more promising exotic fodder tree species in chapter 2 and some commonly used native Jamaican fodder tree species. The basis for chapter 1 and 2 was research conducted in Jamaica between March of 1990 and March of 1991. The data in chapter 3 were determined between January and April of 1991. This research was part of a cooperative Silvo-pastoral Research and Development Project between the Department of Forestry, Michigan State University and the Jamaican Agricultural Development Foundation. The project goal is to integrate tree fodder growth and management knowledge with current small-scale farming systems to develop a sustainable small-scale silvo-pastoral management system. The project has already completed a survey of the indigenous knowledge of small-scale cattle farmers in Trelawny Parish (Morrison 1991), and is currently' funding' a study’ on, the socio-economic conditions and farming systems of the same group of farmers. Additional biological research and the initial testing of different tree fodder production systems are scheduled to begin in the same community in the fall of 1991. EPEECT OP WEED CONTROL, FERTILIZATION, AND PLANTING PIT SIZE ON SURVIVAL AND GROWTH OF DIRECT SEEDED AND SEEDLING TRANSPLAN'I'B 01' W: C D I AND mesm- Abstract. In Jamaica, a field experiment was designed to determine the feasibility of establishing fodder trees in grass pastures by direct seeding, and to investigate the affect of planting pit size, fertilization, and weed control on the survival and growth of direct seeded and transplanted trees. Results demonstrate that Leugaeng and Gliricidia can be established by direct seeding. The potential of direct seeding for Qalliandrg was not thoroughly tested. Of the three experimental factors tested only weed control yielded consistent and positive effects on fodder trees. Circle weeding was superior to bushing alone at improving tree survival and growth. Circle weeding also decreased insect predation when compared to bushing alone. Initial investigation showed that grass mulch is ineffective as a weed control. Planting pit size produced contradictory results. In most instances, larger planting pits had neutral or negative affects on tree survival and growth. The preparation of larger pits is not justified. Nitrogen fertilization eight weeks after sowing or transplanting had little positive influence on tree survival or growth. Fertilization is not advocated. INTRODUCTION Trees can be extremely valuable animal fodder resources. Torres (1983) described the widespread reliance on tree fodder as a dry season emergency feed in Australia. Le Houerou (1980) edited a thorough compilation on the importance of tree browse to animal production in Africa. In Nepal, over fifty percent of fodder needs are supplied from trees (Pandey 1982) . Surveys show that up to 93 percent of Nepalese farmers cultivate fodder trees (Malla 1988), and that trees are the single most important fodder source on some farms (Rusten 1989) . In other countries tree fodder is currently receiving 4 attention as a means of increasing livestock production on small farms (IDRC 1990, Logan and Radcliffe 1985). In the past, tree fodders have had little impact on Jamaican livestock production (JLA 1983) . Pasture grasses are the main source of livestock feed (Kaplan et al. 1976). These grass resources are nutritious and able to sustain high levels of animal production. during' the 'wet season (JLA 1983). However, during the annual December to March dry season (CRIBS 1982) grasses lose much of their nutritional value. Incorporating fodder trees into pastureland can help overcome this dearth of quality dry season feed (Robinson 1985). Trees have extensive root systems that tap subsoil water throughout the year (Sanchez et al. 1985). This enables trees to retain foliage in the dry season, providing a timely supply of fresh succulent fodder“when.grass resources are.dry.' Tree fodders are also nutritious throughout the year (Reynolds and Adeoye 1986, Akkasaeng et al. 1989), containing crude protein levels of 10-20% with some species having levels as high as 25-30% (IDRC 1990). QAIIIQDQIQ calgthyrsus, glirigidia se ' , and Leucaega leggocephala are three widely used fodder tree species. The literature clearly documents their value as livestock feed resources, particularly for the dry season. In Australia, cattle grazing pastures containing Leugaena gained 84% more weight annually then those grazing pure grass pastures (Quirk et al. 1990). In West Africa, Gliricidig and Leucagga are 5 used to increase dry season production of small ruminants (Reynolds and Adeoye 1986). In Thailand, Vearasilp (1981) reports the use of cm; and W as dry season rations to increase sheep production. In Sri Lanka, the use of W with rice polish is as efficient as fresh grass and concentrates in producing weight gain in dairy heifers (Chadhokar and Sivarsupiramajam 1983) . Research on W is not as extensive, but indicates that it is a productive feed when used as a mixed ration (NRC 1983, NFTA 1989a). Researchers, development organizations, and large livestock producers in Jamaica have recently become interested in Calliandra, Gli ic'di , and Leucaeng as forage crops for livestock production (JADF 1988, Wellington 1989, Ruegsegger 1990). The successful use of these species as dry season feeds in other parts of the Caribbean (Paterson et al. 1987, 1988), has lead to the expectation that they could also be used in Jamaica. Before these species are widely recommended or used in Jamaica, establishment techniques which maximize survival and growth, and minimize cost and effort should be developed. The objectives of this study were; 1) to determine the feasibility of direct seeding Calliandr , Gliric'd'a, and We in Jamaican grass pastures, and 2) to investigate the effect of soil cultivation, fertilization, and weed control on the survival and growth of direct seeded or transplanted Qalliagdra, W, and Leugaeng trees. Literature Review Soil Cultivation. In tropical climates competition between crop trees and weeds (i.e. grass, herbaceous, or woody plants) is often intense. Tree establishment can be improved by practicing full site cultivation (Evans 1982) . Soil cultivation reduces weeds and enhances soil aeration and water infiltration (Karlan et a1. 1990) . Under these improved conditions, root growth is easier and plant energy inputs necessary for establishment are reduced (Lal 1979). Full site cultivation is common in South Africa with short rotation tree crops (Germishuizen and Marais 1980, Schonau et a1. 1980). After establishment, annual soil cultivation also improves survival and growth. White (1985, 1986) reported that five years of annual soil cultivation increased fuelwood production up to 25 times. In south Texas, Felker et al. (1986) found that annual surface cultivation doubled Lgucaena and Ergsopis seedling biomass in the first 21 months. A two year study in Egypt found that monthly cultivation to 15 cm increased.Casua;iga height and diameter growth by 30% and 100% respectively (Badran et al. 1986). Although full site or annual cultivation benefits tree survival and growth, long-term negative effects can occur. Repeated cultivation reduces soil moisture and organic matter levels, and destroys soil structure and pore stability (Lal 1979). Reduction of vegetative cover also increases the risk of soil erosion (Hamilton and Pearce 1987). 7 To avoid these hazards, partial site cultivation during establishment is commonly practiced. In many places, tree seedlings are transplanted into planting pits of various sizes (Evans 1982, Germishuizen and.Marais 1980, Shrestha 1980). A planting pit is an area, normally prepared shortly before planting, where the soil is finely cultivated to certain dimensions. In Nepal (Neville 1987) and semiarid West Africa (Nicou 1986) studies indicate that wider and deeper planting pits improve seedling survival and growth of various species in the first two years after establishment. Studies in South Africa by Schonau et al. (1980) demonstrated that.pit width is more important then depth for increasing Euuulyuuus grandis seedling growth. All studies do not show that increased pit size significantly improves tree survival or growth. Sloan (1988) tested three different pit sizes with P_iu_u_s ponderosg in southern Idaho, and Nair et al. (1986) tested six pit sizes with Eucalyptus grandis in India. In neither study was tree survival or growth positively affected by planting pit size. In summary, if soil characteristics are not limiting plant growth, then soil cultivation beyond that necessary for seed sowing or seedling planting is superfluous. Fertilization. Nitrogen fixing trees increase the amount of available soil nitrogen ‘through. their symbiotic relationship with Buiguuium bacteria (Postgate 1987). This relationship often helps improve tree growth, but potential 8 growth may still be limited by a shortage of nitrogen. Hill (1970) found that the application of 30 kg N/ha improved survival and growth of Lguuugug. .Aziz and Habte (1989) found that nitrogen fertilization not only improved growth, but also increased mycorrhizal activity, phosphorous uptake, and nodulation of W on eroded soils. Saininga et al. (1988) reported that the application of nitrogen fertilizer improved establishment of Lguuugng and hastened the nitrogen fixation process by eight weeks. Other studies have reported that nitrogen fertilization can impede nitrogen fixation (Maasdorp and Gutteridge 1986, Faria et al. 1985, Umali-Garcia et al. 1988). The inhibition is not always complete or irreversible. Initial work with Robiuig pseudoauggig indicated that nitrogen fixation rates returned to previous or higher levels once the chemical fertilizer was no longer available to the plant (Johnsen 1991). Nitrogen fertilization is particularly important in improving the slow germination and early growth of some nitrogen fixing trees (Jones et al. 1983a). Moloney et al. (1986) found that the equivalent of 100 kg N/ha significantly increased the dry yields of 15 week old Adeuanthezg puyuuing, Aluizig figlcatarig, and Suulgiuiuuiu insuluuuu seedlings by 100% when phosphorus was not limiting. Jones et al. (1983a) reported that nitrogen application at sowing hastened Leucaena emergence and improved growth at 10 weeks. Similarly, Jones 9 (1985) found that nitrogen fertilization improved Leucaeug growth at bomb 10 and 30 weeks. In a field establishment study in Zimbabwe, Stewart and Gwaze (1988) reported NPK fertilization increased Auuuiu glbigu survival by 10 to 20% and shoot dry weight by 100% at 30 weeks. Weed Control. Whether trees are being grown as timber, fuelwood, or fodder crops, weed control during establishment is a common practice (Evans 1982, Shrestha et al. 1980). Grass, herbaceous, and woody weeds all compete with crop trees for sunlight, soil moisture and nutrients. Control of weed species redirects the growth potential of the site towards the intended crop. I Weed control can be applied in many forms. Clean weeding is the removal of all vegetation on site. Strip weeding is the use of alternating weeded and unweeded strips across the site. Circle weeding is the removal of weed vegetation in a circle around the crop trees. With clean, strip and circle weeding control of competition includes the removal of weed root masses in the designated area. Bushing is the cutting of all above ground vegetation. Studies have shown that clean weeding improves the survival and growth of trees (Cooksley 1987, Evans 1982, Sabas and Kalaghe 1986), however it also leads to increased soil erosion (Hamilton and Pearce 1987) . Circle weeding and bushing also improves tree survival and growth (Evans 1982, Sabas and Kalaghe 1986), but without as great a risk of soil 10 erosion. In addition, both methods retain grass species between fodder trees, thereby providing both wet and dry season feed from the same land (Paterson et al. 1987, 1988). In the Philippines, Balmocena and Casa (1986) compared circle weeding, strip weeding, and clean weeding with a control. Although the dimensions of strip and circle weeding treatments were not reported, circle weeding was the most effective. After two years, circle weeding increased survival and diameter growth of mm guadrialatus seedlings by 12% and 400% respectively, over the control. Besides directly competing with crop trees for site growth potential, weed vegetation may also mask the positive effect of other tree establishment techniques. Badran et al. (1986) reported that soil cultivation was effective at increasing Qasuauing survival and growth only when combined with the control of competing vegetation. Squire (1977) found that giuus Lugiuuu only benefit from fertilization and soil cultivation when weed control is practiced. Jones et al. (1983b) found that fertilization of Leucaena without weed control reduced growth. Similarly, in a study of five leguminous tree species, fertilization without weed control reduced growth of Calliandug, Leucaena, and Acacia unguguiggiuu (Maasdorp and Gutteridge 1986). Summary: Tropical grass species grow fast and compete intensely, denying trees the chance to benefit from favorable nutritional and physical characteristics of the soil (Squire 11 1977). The slow early growth of nitrogen fixing tree species (Jones et al. 1983a), make them more susceptible to grass competition during establishment then other species (Hill 1970, Cooksley 1987). When establishing nitrogen fixing trees in tropical grass pastures measures must be taken to assure acceptable survival and growth. Soil cultivation before planting, early fertilization and post-planting weed control all improve survival and growth of trees (Hill 1970, Squire 1977, Jones et al. 1983a, Badran et al. 1986, Maasdorp and Gutteridge 1986, Nicou 1986, Neville 1987, Stewart and Gwaze 1988). Their combined use should improve establishment of nitrogen fixing fodder trees in improved grass pastures. Materials and Methods Species: Three fodder tree species, Qulliandra calothyrsus, Gliricidiu se inn, and Leucaena leucoce hala, were tested for survival and. growth, responses to ‘three different establishment techniques. These species produce nutritious fodder and are well suited to the elevation, precipitation and soil type of the study site (Falvey 1982, NRC 1983 and 1984). Additionally, they are fast growing, nitrogen fixing species which establish well by direct seeding (NRC 1983 and 1984, NFTA 1989b). Seed of Calliandr; and Leucaena were provided by the Nitrogen Fixing Tree Associationl. Gliricidiu sepiuu seed 1Nitrogen Fixing Tree Association, P.0. 680, Waimanalo, Hawaii 12 was acquired from Bodles Agriculture Research Station, Ministry of Agriculture, Jamaica. Additional supplies of Qulliuugnu and Lguuugng were collected locally. Site: The study site was located on a flat grass pasture at Alcan Jamaica Company's Rio Hoe Farms near Moneague, St. Ann Parish. The surrounding area is comprised of rolling pastures and steep forested hills. Elevation of the site is 500 meters. Annual temperatures vary from 15-25°C, and average annual precipitation is 200 cm (Baker 1968). Holdridge (1967) classifies this area as a subtropical moist forest life zone. Suggested land use for the area is pasture and livestock production and natural forest cover (Baker 1968). The predominate pasture species on site are guinea grass (Euniuum uuximuu) and para grass (Brachiaria muuigg). These species are considered "improved grasses" in Jamaica and are commonly used in livestock production (JLA 1983). The soil is classified.as a St. Ann clay loam, a medium deep red clay over white limestone which contains rich deposits of bauxite (Baker 1968). Available phosphorus and.potassium are considered low, while nitrogen levels are considered adequate (Hewitt 1964). Soil analysis determined a soil pH of 7.48, total nitrogen of 0.42%, and available phosphorus and potassium of 7 and 95 ppm respectively. 13 Treatments: Tree survival and growth were monitored in a 2x2x4 factorial experiment testing establishment techniques. The techniques investigated.were weed control, fertilization, and soil cultivation. Weed control was tested at two levels: 1) bushing - cutting back of all above ground vegetation within the replication, and 2) circle weeding - the above treatment combined.with the removal of all vegetation, including roots, from a 40 cm diameter circle around each seedling. Weed control was applied on a monthly basis for six months, at which time the trees achieved a dominant position over the grass canopy (Roshetko et al. 1991). Fertilization was also tested at two levels: 1) control (no fertilization), and 2) a single application of two grams of ammonium sulphate per planting location. Assuming 40 square cm of surface area per location, the approximate rate per hectare was 125 kgs. This is equivalent to the annual spring fertilization rate used on surrounding pasture lands (Malcolm 1990). To avoid burning the plants, fertilizer was applied in a circle 10-15 cm from the seedling (Malcolm 1990) . Although fertilization was to occur immediately after germination, drought and logistical problems delayed application until eight weeks after sowing. Soil cultivation was tested at four levels: 1) 0x0x0 cm (control), 2) 10x10x10 cm, 3) 20x20x20 cm, and 4) 40x40x40 cm. These levels will be referred to as planting pits for the l4 remainder of the paper. Planting pits were prepared one week before sowing with hand tools. The soil was excavated, finely tilled and replaced before sowing. Experimental Design: There were a total of sixteen treatments per tree species (2 weed control x 2 fertilization x 4 pit sizes). Each species was replicated three times in a split-split block design. This design facilitated application of the treatments and improved the precision of comparing treatment interactions (Little and.Hills 1978). Weed control levels were allocated to the main blocks, fertilizer levels to the split blocks and planting pit sizes to the split-split blocks (Figures 1a and 1b). 'This arrangement provided for the strongest statistical comparison of planting pit size, followed by fertilization, and then weed control (Little and Hills 1978). Each replicated block was 25.5 x 25.5 meters. Blocks were split to accomodate the two weed control levels, then split again perpendicular to the first division to accomodate fertilizer levels. The resulting quadrants were further subdivided to accommodate the four planting pit sizes. Each of the split-split blocks contained nine measurement trees at a 1.5 x 1.5 meter spacing; Border trees were used to separate the split-split blocks. Each main block contained 17 rows of 17 trees, for a total of 289 trees per replication. Measurements and Analyses: Stewart and Gwaze ( 1988) reported shoot dry weight as superior to height as a measure 15 ir I in No Fert. Fert. mum: No Part. Fert. Figure 1a. Main block layout of fertilizer and weed control levels. A. 40x40x40 cm 8. 20x20x20 cm C. 10x10x10 cm 0. 0x0x0 cm Figure 1b. Main block layout of planting pit size levels. 16 of growth but it requires destruction of the plant. Long-term monitoring plans for the study required that height and diameter be used to measure growth. Additionally, insect damage on gliriuigiu seedlings was recorded at six months, insect damage data was not collected for Lgugggu; because of extremely low survival with some treatments. Analysis of variance (ANOVA) and. contrast. analysis ‘with linear coefficients were used for statistical comparisons (Little and Hills 1978). Growth data was tested for violations of the assumptions of analysis of variance before analysis, results were non-significant. Survival data were converted from percentages to degrees using arcsine transformation before analysis (Little and Hills 1978). Establishment: The site was mowed and planting pits excavated one week before seed sowing. Seeds were pre-treated and inoculated with species specific guizobiuu according to accepted guidelines (NFTA 1989a, 1989c), and.then sown at a 1- 2 cm depth (NFTA 1989b). .As a hedge against poor germination, three seeds were sown per pit (Halliday and Naiko 1984). After 2 months, seedlings were thinned to one plant per pit. Seeding Failure: Although soil moisture and growing conditions were good before seed sowing (Thompson 1986) , establishment of some species was poor. All Calliangrg plots failed to germinate (3.5%) and two of the Leucgeug plots germinated poorly (21%). Poor germination was assumed to be 17 related to a three week drought that followed seed sowing (Roshetko et al. 1991). Excavation of seeds however, determined that Lguuuggg had been sown at twice the recommended depth in failed blocks. Qu_l_;_i_a_n_c_i_r_a seed, although sown properly, had not received the appropriate pretreatment due to conflicting information (NFTA 1989a, 1989c). To achieve replication in the only surviving w block, the fertilizer treatments were not applied. This provided for two replications of Leucaeug testing weed control and planting pit size. Seedling Establishment: Five blocks with insufficient pit occupancy were re-established with seedling nursery stock four months after the initial sowing operations. Four blocks were planted with Leucaena seedlings, and one with Calliandra. All sixteen treatments previously described were applied to the four Leucaena blocks. Only soil cultivation was tested with Calliaudug seedlings. The original single block was divided into four replications. Both Lguuueug and gulliuudxg seedlings were established in four replications to increase the number of replications and therefore the precision of the statistical analyses (Little and Hills 1978). Seedlings were raised in 5x5x15 cm polyethylene nursery tubes according to Liegel and Venator (1987). The tube soil mass of nursery raised seedling altered the size of the planting pit experimental control from zero to roughly 5x5x10 centimeters. Cultivation to 15 cm depth was not necessary due 18 to soil settlement in nursery tubes. other pit sizes were unaffected. As with direct seeding, three seeds were sown per tube (Halliday and.Naiko 1984). Seedlings were thinned to one per tube after six weeks. Before seedlings were transplanted, the blocks were treated with glyphosate herbicide applied as a 1.0% solution combined with a 1.0% surfactant, frigate (Monsanto 1985) . Herbicide application effectively controlled weed growth for two months. Seedlings were transplanted after eight weeks in the nursery. Calliandra seedlings were 15 to 20 cm tall and Leucaena seedlings were 12 to 15 cm tall. Weed control treatments were applied monthly throughout the study period. Fertilizer was applied eight weeks after transplanting. RESULTS Germinatiou, All germination percentages were based on three seeds per pit. Occupancy was determined by the number of pits with at least one living seedling. All Calliandug replications and two of three Leucaeng replications germinated poorly due to methodological error (See Methods and Materials) . In the surviving Leugaeug replication germination was only 52%, however, pit occupancy was 86%. Cilifluidiu germination was 78% with pit occupancy of 96%. Gemination was recorded ten days after sowing and re-checked ten days later. Data was consistent between the two dates. Germination percentages for Calliandug and Leucaeng grown in the nursery were not recorded. l9 fi!£!1!;1.32§.§£2!§h; Direct Seeding: ANOVA tables for survival analysis of direct seeded species are in Table 1. Results of survival and growth data analyses are given in Tables 2 and 3. Interactions between factors were not found to be significant in any case. Weed control was the only practice tested that affected tree survival and growth in a consistent and positive manner. Survival of direct seeded Leugaeng and fiLiuigiuia were 73% and 14% greater, respectively, with circle weeding compared to bushing alone (Table 2 and 3). Total height and basal diameter growth of Leucaena and Gliricidiu were also improved. Leucgeua total height and basal diameter growth increased by 181 cm (197%) and 19.7 mm (229%) respectively (Table 2). Gliuiuigiu total height and diameter increased by 44 cm (34%) and 8.2 mm (33%) respectively (Table 3) . Additionally, circle weeding significantly (P< .05) decreased severe insect damage by 524% (from 19.4% to 3.7%) over bushing alone. Fertilization and planting pit size had little effect on Leucaeng or Gliricidiu in the first ten months after direct seeding. Fertilization increased Gliuicidig survival by 4% but.had.no significant affect onlgrowth. Survival of Leucaeng was higher in the largest planting pit sizes (Table 2), however contrast analysis determined no significant linear trend between means. Contrasts between treatment means or groups of means were not designated during experimental design, and thus not analyzed (Little and Hills 1978). 2 0 Table 1. ANOVA tables of survival data for Gliricidia and Leucaena trees ten months after direct sowing. QLIRICIDIA Factors DF Mean F Square ratio Main plot Replication (Rep) 2 820.83 Weed Control (WC) 1 3185.02 313.59** Error A (Rep x WC) 2 10.16 Split plot Fertilization (Fer) 1 212.52 13.45+ Error B (Rep x Fer) 2 15.79 Split-split plot Pit size (PS) 3 23.40 .32 Error C (Rep x PS) 6 72.16 Interactions (Ints) WC x Fer 1 146.30 1.31 WC x PS 3 146.39 1.31 Fer x PS 3 82.84 .74 WC x Fer x PS 3 29.25 .26 Rep x Primary Ints 14 ------ ---- Error 6 111.93 LEUCAENA Factors DF Mean F Square ratio Mainplot Replication (Rep) 1 83.72 Weed Control (WC) 1 14969.52 636.39* Error A (Rp xWC) 1 23.52 Split plot Pit size (PS) 3 244.72 5.51 Error B (Rep x PS) 3 43.58 Interaction (Int) WC x PS 3 78.50 Error 3 133.49 .59 * Significant at P< .05. *0 Significant at P< .01. + Significant at P< .10. 21 Table 2. Leucaena survival and growth by treatment, ten months after e d Treatment Survival Growth: Ht. Basal On. '1’ “GE“ -mm:, Bushing 10* (4.1)1 92+ (18.7) 8.6+ (1.6) Circle weeding 83 (4.1) 273 (18.7) 28.3 (1.6) Pit size 0x0x0 50+ (5.8) 198 (26.9) 20.0 (2.3) (cm): 10x10x10 50 (5.8) 171 (26.9) 17.5 (2.3) 20x20x20 47 (5.8) 168 (26.9) 16.9 (2.3) 40x40x40 61 (5.81 194 (26.9) 19.5 (2.3) * - Significantly different by ANOVA at P< .05. f - Significantly different by ANOVA at P< .10. - Values in parenthesis are standard errors of the means. lele 3. Cliricidia survival and growth by treatment, ten months after d1 3 1 Treatment Survival Growth: Ht. Basal Dm. -‘- - - -m- Bushing 81*. (2.2)‘ 129* (1.8) 24.9. (0.3) Circle weeding 95 (2.2) 173 (1.8) 33.1 (0.3) Control 86+ (2.2) 147 (1.8) 28.5 (0.3) Fertilization 90 (2.2) 156 (1.8) 29.4 (0.3) Pit size 0x0x0 74 (3.0) 147 (2.6) 28.4 (0.3) (cm): 10x10x10 72 (3.0) 158 (2.6) 30.5 (0.3) 20x20x20 75 (3.0) 151 (2.6) 29.0 (0.3) 40x40x40 73 (3.0) 149 (2.6) 28,1 (9,3) *' - Significantly different by ANOVA at P< .01. * - Significantly different by ANOVA at P< .05. + - Significantly different by ANOVA at P< .10. 1 - Values in parenthesis are standard errors of the means. 22 Seedling Establishment: ANOVA tables for survival analysis of Leucaeug and.§ulliuug;u seedlings are in Table 4. Survival and growth data are given in Tables 5 and 6. The interaction between weed control and fertilization treatments was significant for Leugueuu basal diameter growth. Fertilization improved diameter growth of Leucaeuu (by 12.9%) only when circle weeding was practiced. Bushing alone did not provide adequate weed control for seedlings to benefit from fertilization (Figure 2) . Leucaeua survival and height growth were unaffected by fertilization. All other interactions were non-significant. As a single factor, weed control significantly affected growth of Leucaeug seedlings but had no affect on seedling survival. Total height and basal diameter growth increased by 7 cm (6.6%) and 3.4 mm (29.6%) respectively (Table 5). Fertilization had no direct affect on Leucaena seedlings. Planting pit size affected seedling transplants of both species. Larger planting pits positively affected Calliandra seedling survival, but negatively affected basal diameter growth of both Calliandra and Leucaeng seedlings (Table 5 and 6). Contrast analysis using linear coefficients confirmed the linear relationships in all three of these data parameters, as well as for total height.growth.of Calliandra seedlings (Table 5 and 6). Although not significant under ANOVA or contrast analysis, Leucaena total height growth also showed a negative trend with increased pit size. 23 Table 4. ANOVA tables of survival data for Leucaena and Calliandra seedlings six months after transplanting. LEEQAENA Factors DF Mean F Square ratio Main plot Replication (Rep) 3 3271.62 Weed Control (WC) 1 1756.66 3.16 Error A (Rep x WC) 3 556.10 Split plot Fertilization (Fer) 1 17.33 .08 Error B (Rep x Fer) 3 321.39 Split-split plot Pit size (95) 3 278.14 .13 Error C (Rep x PS) 9 2091.51 Interactions (Ints) WC x Fer 1 5.23 .01 WC x PS 3 198.70 .15 Fer x PS 3 242.42 .19 WC x Fer x PS 3 385.49 .30 Rep x Primary Ints 21 ------ ---- Error 9 432.57 QALLIANDRA Factors DF Mean F Squaue; ratio Plot Replication (Rep) 3 4.12 Pit size (PS) 3 477.40 3.69+ Error (Rep xPS) 9 129.46 + Significant at P< .10. 24 Table 5. Leucaena survival and growth by treatment, six months after tr Treatment Survival Growth: Ht. Basal Dm. “' ‘QE? :mmr Bushing 68 (3.7)1 106+ (4.0) 11.5* (0.3) Circle weeding 81 (3.7) 113 (4.1) 14.9 (0.3) Control 75 (3.7) 106 (4.1) 12.8 (0.3) Fertilization 79 (3.7) 113 (4.0) 13.6 (0.3) Pit size 5x5x10 69 (5.2) 118 (5.8) 14.3*# (0.4) (cm): 10x10x10 74 (5.2) 108 (5.8) 13.8 (0.4) 20x20x20 75 (5.2) 107 (5.6) 12.5 (0.4) 40x40x40 77 15.21, 102, 15.5) 12.2 (0.4) * - Significantly different by ANOVA at P< .05. + - Significantly different by ANOVA at P< .10. f - Significant linear relation by contrast analysis at P< .05. - Values in parenthesis are standard errors of the means. Table 6. Calliandra survival and growth by treatment, six months after tra s 1an . Treatment Survival Growth: Ht. Basal Dm. '8'- -Qm' -1111- Pit size 5x5x10 81+! (5.7)‘ 244: (4.9) 36.3+# (1.1) (cm): 10x10x10 78 (5.7) 230 (4.9) 36.9 (1.1) 20x20x20 92 (5.7) 232 (4.9) 34.3 (1.1) 40x40x40 97 (5.7) 225 (4.9) 32.6 (1.1) + - Significantly different by ANOVA at P< .10. I - Significant linear relation by contrast analysis at P< .05. 1 - Values in parenthesis are standard errors of the means. 25 basal diameter (mm) 18 fi 14— 12— \l 10— x Bushing Circle Weeding - Control Fertilizer Figure 2. Effect of weed control x fertilizer interaction on Leucaena diameter growth at 6 months. 26 DISCUSSION Germination. Despite a three week drought immediately following seeding, Lguuugnu and glixiuigiu were successfully established by direct seeding into pastures where heavy grass competition exists. Appropriate seed scarification, fliizobium inoculation, and seed sowing procedures were necessary for direct seeding to be successful. Depth of sowing is extremely important. Leucaen; seeds were particularly sensitive to being sown too deep. Ruegsegger (1991) found similar results when direct seeding Leucaen; and Gliricidiu. In this study, gulliuugzu_did not successfully establish by direct seeding, probably because it was not properly scarified (NFTA 1989). When properly scarified Calliandr; growth in both the nursery and field exceeded that of Leucaena. Results from other researchers recommend Leucaen , Gliuicidia as well as Calliandr; for establishment by direct seeding (Fields and Yassin 1990). Weed control. This study clearly indicates that circle weeding is better than bushing alone when establishing Lguuugug and Glizicidig trees in improved pastures. Tropical grasses compete intensely for sunlight, soil moisture, and nutrients (Squire 1977). Depending upon climatic conditions, monthly grass growth varied from 70 to 150 cm in height per month. Confronted with such intense weed growth, it is assumed that even bushing alone would significantly improve tree survival and growth over no weed control at all. 27 Circle weeding was very effective at improving survival and growth of both directly sown and transplanted species. Elimination of vegetation around each seedling provided a weed free zone in which seedlings could establish and grow. This weed free zone was very important to Leugueng. Only 10% of direct seeded Leucgenu survived in bushing alone plots. The height increases provided by circle weeding enabled Lguuugug and Qliriciuiu to achieve dominant positions in the pasture. Trees that.received.bushing alone*were either suppressed by or equal to the grass canopy after ten month's growth. Susceptibility of Leuuaeng to weed competition has been reported by others (Cooksley 1987, Hill 1980). Establishment of this species should not be attempted without weed control similar to the circle weeding used in this study. gligicidig is also known to be sensitive to weed competition when established by direct seeding (NFTA 1989b). In this study, Gliuiuigiu benefitted from circle weeding, but was much more competitive then Leugaeng. Gliricidiu can establish with bush weeding alone, but circle weeding is recommended. Although the growth increases of Leuuagnu were biologically great, they were not highly significant (P<.10). This is partially due to the low number of replications, fewer replications decrease statistical precision (Little and.Hills 1978), and the wide variation in the data. The low level of statistical significance should not cause the importance of these growth increases to be disregarded. 28 Besides directly decreasing weed competition, circle weeding also decreased insect predation on young trees. Severe insect damage of Glizigidia was over five times greater with bushing alone compared to circle weeding six months after seed sowing. Severe insect damage was defined as complete defoliation, topping, or mortality caused by insects. It is speculated that the removal of the vegetation in the circle weeded plots denied insects the access to the seedlings. Once trees achieved a dominant position in the pasture, six months after sowing, insect attacks greatly diminished. Survival in Lguuggug plots was so skewed (survival only 10% in bushing alone plots) that quantifying insect damage was not attempted. However, as with Gliricidia, once Leucaena achieved dominance in the pasture, insect predation decreased greatly. Studies have shown that complete weed control would increase tree survival and growth more then partial control (Evans 1982, Sabas and Kalaghe 1986) . However the grass retained by circle weeding is a valuable forage asset. The two level forage production offered by a combination of grasses and fodder trees better utilizes site growth potential and distributes forage availability over the year. Grass production is used during the wet season when it is succulent and nutritious, and.tree fodder is used in the dry season when grass nutritive values are low (Ahmed 1986, Paterson et al. 1987, 1988). Trees retain fresh nutritious foliage through the dry season because of their deep perennial root systems 29 (Sanchez et a1. 1985, Wilson.et al. 1986). The value of this grass/ fodder tree system far exceeds the potential fodder tree growth increase possible with complete weed control. Other'means of weed control. In this study weed control was accomplished by manual means. In Jamaica the availability of agricultural labor is often tenuous (LeFranc 1981). When it became necessary to re-establish some of the study plots, herbicide was used as a pretreatment. Herbicide application was effective at controlling weed growth for two months, and made subsequent manual weeding applications less time consuming. Weed control treatments were applied monthly to the Laucaeua seedling plots. Circle weeding significantly improved total height and basal diameter growth, but not as much as in the direct seeding plots where herbicide was not used. Circle weeding had no effect on Laueaena survival when herbicide was used as a pretreatment. In Jamaica, both manual and chemical weed control methods can be expensive or difficult to acquire. Which method to employ is determined by availability workers and equipment, as well as the respective costs which often fluctuate. No economic comparison between manual weeding and herbicide use has been attempted because of this reason and the seasonal variation in plot establishment. The use of mulch as a weed control was also investigated during the study. Four inches of grass mulch (flujguu maximum and fixauhiauia uuuiaa) was applied between plots where weed 30 competition was similar to that within plots. For two weeks, mulch retarded grass growth almost 100%, but after one month grass growth was equal to that in the unmulched area. How this would affect tree survival and growth was not studied, but this method of mulching did not decrease the necessity of weed control. In the Philippines, Principe (1976) found that grass mulching at planting did not improve nine month survival or growth of Gmeliua arbouea seedlings, or decrease weed growth. The use of mulch as weed control is not suggested unless combined with complete weeding. Planting pit size. Studies by Neville (1987) and Nicou (1986) have indicated that larger planting pit size improves establishment of tree species. Conversely, Sloan (1988) and Nair et al. (1986) found that planting pit size had no affect on tree survival or growth. In this study planting pit size had no straightforward affect on tree survival or growth. Trees established by direct seeding were unaffected by pit size. Survival of Qalliandra transplants was positively affected by larger pits. Growth of Leucaena and Calliandza seedling transplants, however, were negatively affected by larger planting pits. Although these results are conflicting, most evidence indicates that larger pit sizes have a negative or neutral effect on trees. Site soil characteristics do not appear to restrict tree survival or growth. The time and costs of preparing planting pits is not warranted. Soil cultivation 31 beyond.what is necessary for seed sowing or seedling planting is not recommended. I Fertilization. Fertilization had little affect on tree survival or growth in this study, even through soil analysis confirmed nitrogen deficiencies (Malcolm 1990). In one case, tree basal diameter growth was increased by fertilization, but only when combined with circle weeding. Trends towards increased survival and growth were apparent with fertilization, but improvements were inconsistent and biologically small. The costs of fertilizer and application were not justified. When nitrogen fertilizers are applied at sowing, improved tree survival and.growth can.occur (Jones 1985, Moloney et al. 1986, Stewart and Gwaze 1988). In this study fertilizer application was delayed until the eighth week. The opportunity to accelerate germination and early tree growth, which can be slow and detrimental to establishment, may have been lost by this delay. In the future, fertilization should occur at sowing or planting. General growth observations. Three and six month growth data for direct sown species have been reported elsewhere (Roshetko and Lantagne 1990, Roshetko et al. 1991). At both intervals, results were similar to those reported here. Circle weeding significantly improved survival and growth of Lauaaaua and gliriciaia, but fertilization and planting pit size had no significant affect. The effect of circle weeding 32 on Lauaaaua survival and growth, and 9.1mm survival accelerated with time. The effect on film growth remained constant. Similar periodic growth data for these species are not available. However, growth rates of Lauaaaua in this study compare favorably with general information (NRC 1984), and other sites in Jamaica (Thompson 1985). Gliuiuidia growth rates for other Jamaican sites were not available. Growth rates of Callianuua in this study parallel general information about this species (NRC 1983). Calliandra height and diameter growth were superior to those of both Leucaena and gliriuiaia, even though the gliuiaiaia trees were four months older (Tables 3, 5 a 6) . General observations determined that foliage biomass production were greatest for galliaudua followed by Gliriciaia and then Leucaena. This suggests that Calliandua is well adopted to the study site, and potentially an excellent fodder tree species for Jamaica. CONCLUSIONS This study indicates that Laucaeua and gliricigia can be established in improved grass pastures by direct seeding, if proper seed scarification, inoculation, sowing procedures and weed control are followed. Direct seeding of Qalliandra was not thoroughly tested, but other researchers recommend its establishment by direct seeding (Fields and Yassin 1990). 33 Weed control was the only factor in this study that had consistent and positive effects on fodder tree survival and growth in improved grass pastures. Results demonstrate that control of weed competition until fodder trees obtain a dominant position in the pasture ecosystem is important. Circle weeding was superior to bushing alone at improving tree survival and growth. Complete weed control would also improve tree survival and growth (Evans 1982, Sabas and Kalange 1986) . However, the grass retained by circle weeding is an additional fodder resource. The value of this grass/fodder tree system far exceeds the potential fodder tree growth increases possible with complete weeding. Circle weeding is recommended for weed control. Soil cultivation (i.e. planting pit size) contributed inconsistent results in this study. Under direct seeding, planting pit size generally had no affect on tree survival or growth. Larger pit size also had a neutral influence on survival of seedling transplants, but a negative affect on growth of transplants. For this site, soil cultivation beyond what is necessary for seed sowing or seedling planting is superfluous or deleterious, and not encouraged. Nitrogen fertilization eight weeks after sowing or transplanting had little positive influence on tree survival or growth, it is not advocated. Survival and growth of nitrogen fixing trees can be positively affected by fertilization at sowing (Jones 1985, Moloney et al. 1986, 34 Stewart and.Gwaze 1988). If fertilization.is desirable, it is recommended at sowing or planting. Contrary to studies by Squire 1977, Jones et al. 1983b, and Badran et al. 1986, no additive or synergistic effects were realized by combining weed control and soil cultivation, or fertilization and soil cultivation treatments. Fertilization positively affected growth only when combined with adequate weed control. These results agree with those of ‘others working on nitrogen fixing trees and fertilization (Jones et al 1983b, Badran et al. 1986, and Maasdorp and Gutteridge 1986). ACKNOWLEDGEMENTS The author would like to thank the Jamaica Agricultural Development. Foundation. and. the ‘United States .Agency for International Development for financial support, Alcan Jamaica Co., Agricultural Division for logistical and technical support and use of the study site, and the Nitrogen Fixing Tree Association for the seeds. BIBLIOGRAPHY Ahmed, B. 1986w Growing forage legumes. CARDI Factsheet.No. AP-F/21.86. Caribbean Agricultural Research and Development Institute. St. Augustine, Trinidad. West Indies. 4 p. Akkasaeng, R., R.C. Gutteridge, and. M. ‘Wanapat. 1989. Evaluation of trees and shrubs for forage and fuel-wood in Northeast Thailand. International Tree Crops Journal 5:209- 220. A212, T., and M. Habte. 1989. Influence of inorganic nitrogen on mycorrhizal activity, nodulation and growth of Iaucaeua laucoceuhala in an oxisol subjected to simulated erosion. Communications in Soil Science and Plant Analysis 20:239-251. Badran, O.A., A.B. El-Sayed, and M.H. El-Lakany. 1986. Effect of tillage and gramoxone weed spray on the growth of Casuariua species. Alexandria Journal of Agricultural Research 31:385-395. Baker, G.H. 1968. Soil and landuse survey No. 24 Jamaica, Parish of St. Ann. Regional Research Centre, Department of Soil Science, University of the West Indies, Trinidad. Balmocena, R.B., and E.V. Casa. 1986a Growth and.development of established toog (Petersiauthus quaduialatus (Merr.) Merr.) plantations under four different weeding methods. Sylvatrop 11:55-60. Chadhokar, P.A., and S. Sivasupiramaniam. 1983. Gliricidia (gliricidia maculata H.B.& K.) leaves as protein supplement to paddy straw in growing cross-bred heifers. Indian Journal of Animal Science 53:120-125. Cooksley, D.G. 1987. Effect of weed competition on the early growth of Leucaena leucocephala. Tropical Grasslands 21:139- 142. CRIES. 1982. Jamaican resource assessment. Comprehensive resource inventory and evaluation system project (CRIES) . Michigan State University - U.S. Department of Agriculture - The Ohio State University. Department of Resource Development, Michigan State University, East Lansing, Michigan. Evans, J. 1982. Plantation forestry in the tropics. Oxford University Press. New York, New York. 472 p. Falvey, J.L. 1982. Gliricidia uaculaua - A review. International Tree Crops Journal 2:1-14. 35 36 Faria, S.M. de, A.R.M. de Jesus, and A.A. Franco. 1985. Field establishment of nodulated Leugaeua lauauaauhala K72. Leucaena Research Report 6:14-16. Felker, P., D. Smith, and C. Wiesmann. 1986. Influence of mechanical and chemical weed contrrol on growth and survival of tree plantings in semiarid regions. Forest Ecology and Management 16:259-267. Field, S.P., and‘M.Yasin. 1990. Improving the effectiveness of multiprurpose tree species research in Nusa Tengarra Timur, Indonesia. pp. 125-132. In: C. Haugen, L. Medema, and C.B. Lantican (eds.) Multipurpose tree species research for small farms: Strategies and methods. Winrock International Institute for Agricultural Development, and International Development Research Centre. Funny Publishing Limited Partnership, Bangkok, Thailand. Germishuizen, P.J., and J.P. Marais. 1980. Site preparation: Establishment and re-establishment of conifer plantations in the summer rainfall region of southern Africa. pp. 26-46. In: Proceedings.of the forest.seed, nursery, and.establishment research working group, third meeting. Saasveld Forestry Research Station, George, South Africa. South Africa Forestry Research Institute, Wattle Research Institute, University of Natal. Halliday, J., and P. Nakao. 1984. Technical notes on the germination of leguminous tree seeds. pp. 231-234. In: Simposio Sobre Fixacao de Nitrogenio em Arvores Tropicais. Brazilian Journal of Agricultual Research ‘Volume 19 (Special Issue). Hamilton, L.S., and A.J. Pearce. 1987. What are the soil and water benefits of planting trees in developing country watersheds? Chapter 4 pp. 38-58. In: 0.8. Southgate and J.P. Disinger (eds.) Sustainable resource development in the third world. Westview Press. Hewitt .W. 1964. Soil technical sheets. Agric. Chem. Div., Ministry of Agriculture, Jamaica. 144 p. Hill, G.D. 1970. Studies on the growth of Leucaena lauauuaphala: Effect of clean weeding and nitrogen fertilizer on early establishment. Papua New Guinea Agriculture Journal 22:29-30. Holdridge, L.R. 1967. Life zone ecology. Tropical Science Center. San Jose, Costa Rica. 37 IDRC. 1990. Shrubs and tree fodders for farm animals. International Development Research Centre. Ottawa, Ontario, Canada. 349 p. JADF. 1988. Jamaica Agricultural Research Program Annual Report 1988. Jamaican Agriculture Development Foundation. Kingston, Jamaica. JLA. 1983. Livestock manual for the tropics. The Jamaica Livestock Association. Kingston, Jamaica. 406 p. Johnsen, K. 1991. Petawawa National Forestry Institute, Canada. Personal Communication at the International Conference on Black Locust: Biology, culture, and utilization. Michigan State University, East Lansing, Michigan. June 17- 21, 1991. Jones, R.J., G. Vilamizar, and S.J. Cook. 1983a. The effect of seed treatments and nitrogen fertilizer on seedling growth of Lauuaana. Leucaena Research Report 4:4-5. Jones, R.J., G. Vilamizar, and S.J. Cook. 1983b. The effect of nitrogen fertilizer, weed competition, and CCC spray on seedling growth and nodulationof Leucaena. Leucaena Research Report 4:6-7. Jones, R.J. 1985. Effect of nitrogen and inoculum levels on establishment and nodulation of Leucaena. Leucaena Research Report 6:8-10. Kaplan, I., H.I. Blustein, K.T. Johnston, and 0.3. McMorris. 1976. .Area handbook for Jamaica. Foreign Area Studies of the American University. Washington D.C. 332 p. Karlan, D.L., D.C. Erbach, T.C. Kaspar, T.S. Colvin, E.C. Berry, and D.R. Timmons. 1990. Soil tilth: A review of past present and future needs. Soil Science Society of America Journal 54:153-161. Lal, R. 1979. Importance of tillage systems in soil and water management in the tropics” pp. 25-39. In: R. Lal (ed.) Soil tillage and crop production. IITA Proceedings No. 2. International Institute of Tropical Agriculture. Ibadan, Nigeria. LeFranc, E. 1981. The utilization and organization of rural labor in Jamaica. In: H. Blustain and E. LeFranc (eds.) Strategies for organization of small-farm agriculture in Jamaica. Center for international Studies, Cornell University, New York. 38 Le Houerou, N.H. (Editor) 1980. Browse in Africa: The current state of knowledge. ILCA Addis Ababa. 491 p. Liegel, L.H., and C.R. Venator. 1987. A technical guide for forest nursery managemant in the Caribbean and Latin America. General.Technical.Report 80-67. New'Orleans, La: U.S. Dep. of Ag., Forest Service, Southern Forest Exper. Station; 156 p. Little, T.E., and Hills, F.J. 1978. Agriculture experimentation - Design and analysis. John Wiley & Sons, Inc. New York. 350 p. Logan, I.A., and J.E. Radcliffe (eds.) 1985. Fodder trees - A summary of current research in New Zealand. 49 p. Crop Research Division - Department of Scientific and Industrial Research. Christchurch, New Zealand. Maasdorp, B.V., and R.C. Gutteridge. 1986. Effect of fertilizer and weed control on the emergence and early growth of fvie leguminous fodder shrubs. 'Tropical Grassslands 20:127-134. Malcolm, C.A. 1990. Agronomist, Alcan Jamaica Company LTD. Manderville, Jamaica. Personal Communication. Malla, Y.B. 1988. The case for placing more emphasis on private tree planting programmes. Forestry Research Paper Series No. 13. HMG-USAID-GTz-FORD-WINROCK Project. Kathmandu, Nepal. 46 p. Moloney, R.A., R.L. Aitken, and R.C. Gutteridge. 1986. The effect of phosphorus and nitrogen application on early growth of agenanthera avonia, Albizia falcataria, and Schleinitaia insulauuu. Nitrogen Fixing Tree Research Reports 4:3-6. Monsanto. 1985 Roundup herbicide - Complete directions for use. Monsanto Company, St. Louis, Missouri. U.S.A. 104 p. Nair, J.M., A.D. Dec, and T.J.C. Nair. 1986. Reclamation of grasslands at. Kulamavu, Kerala. pp. 200-209. In: .J.K. Shgarma, C.T.S. Nair (eds.) Eucalypta in India. Past present and future. Proceedings of the national seminar held at Kerala Forest Research Institute, Peechi, Kerala, India. Neville, G.A. 1987. Site amelioration for the establishment of fodder trees. Banko Janakari 1:32-33. Nepal-UK Forestry Research Project. Kathhmandu, Nepal. NFTA. 1989a. Qalliandra calthyrsua - An Indonesian favorite goes pan-tropic. NFT Highlights. Nitrogen Fixing Tree Association. Waimanalo, Hawaii. 2 p. 39 NFTA. 1989b. W production and use. Nitrogen Fixing Tree Association. Waimanalo, Hawaii. 44 p. NFTA. 1989c . NFTA cooperative planting program: Establishment guidelines. Nitrogen Fixing Tree Association. Waimanalo, Hawaii. 36 p. Nicou, R. 1986. Influence of ploughing on soil physical properties and growth of annual crops in semi-arid West Africa: Relevance to tree planting. Forest Ecology and Management 16:103-115. NRC (National Research Council). 1983 galliauaua: A versatile small tree for the humid tropics. National Academy Press, Washington, D.C. 52 p. NRC (National Research Council). 1984. Leucaena: Promising forage and tree crop for the tropics. Second Edition. National Academy Press, Washington, D.C. 100 p. Pandey, K. K. 1982. Fodder trees and tree fodder in Nepal. Swiss Development Cooperation. Berne, Switzerland. 107 p. Paterson, R.T., G.A. Proverbs, and J.M. Keoghan. 1987. The management and use of forage banks. Caribbean Agricultural Research and Development Institute. St. Augustine, Trinidad, West Indies. 21 p. Paterson, R.T., J .M. Keoghan, and G.A. Proverbs. 1988. Forage-based grazing systems for the drier parts of the caribbean region. Caribbean Agricultural Research and Development Institute. St. Augustine, Trinidad. West Indies. 23 p. Postgate, J.R. 1987. Nitrogen fixation. Edward Arnold LTD. London, Great Britain. 73 p. Principe, E.B. 1976. Mulching and fertilization of out planted yemane (Gmelina aubuuaa Roxb.) seedlings. Sylvatrop 1:223-236. Quirk, M.F., C.J. Palon, and J.J. Bushell. 1990. Increasing the amount of Leucaena on offer gives faster growth rates of grazing cattle in South East Queensland. Australian Journal of Experimental Agriculture 30:51-54. Reynolds, L., and S.A.O. Adeoye. 1986. Planted leguminous browse and livestock production. Paper presented at: International workshop on Alley Farming for humid and sub- humid.regions of tropical Africa. Ibadan, Nigeria. IMarch 10- 14, 1986. 40 Robinson, P.J. 1985. Trees as fodder crops. pp. 281-300. In: M.G.R. Cannell, and J.E. Jackson (eds.) Attributes of trees as crop plants. 592 p. Institute of Terrestrial Ecology. Natural Environment Research Council. Huntingdon, England. Roshetko, J.M. , and D.O. Lantagne. 1990. Fodder tree establishment in Jamaican pastures: Three month results. Paper presented at the Second Annual Seminar for presentation of progress reports on Jamaican Agricultural Research Projects. The Jamaican Agricultural Development Foundation. Kingston, Jamaica. September 18-20, 1990. Roshetko, J.M., D.0. Lantagne, and M.A. Gold. 1991. Influence of weed control, fertilization, and planting pit size on the survival and growth of Leucaena and Gliricidia establishment by direct seeding in pastures. Nitrogen Fixing Tree Research Reports 9:64-67. Ruegsegger, G. 1990. On-farm. evaluation. of Leucae , glluiuiaia and legume/grass conserved fodder as high protein forages in dairy cattle rations. Paper presented at the Second Annual Seminar for presentation of progress reports on Jamaican Agricultural Research Projects. The Jamaican Agricultural Development Foundation. Kingston, Jamaica. September 18-20, 1990. Ruegsegger, G. 1991. Personal communications. Rusten, E. 1989. .An investigation of an indigenous knowledge system and management practices of tree fodder resources in the Middle Hills of Central Nepal. Ph.D. Dissertation. Department of Forestry, Michigan State University. 282 p. Sabas, E., and A.G. Kalaghe. 1986. The influence of weeding on early growth of Eucalyptus camaldulensia at Igwala Mwanza Tanzania. Tanzania Silviculture Research Note No. 44 7 p. Sanginga, N., K. Mulongoy, and.A..Ayanaba. 1986. Inoculation of Leucaena leucocephala (Lam.) de Witt with Baizobiuu and its nitrogen contribution to a subsequent maize crop. Biological Agriculture and Horticulture 3:347-352. Sanchez, P.A., C.A. Palm, C.B. Davey, L.T. Scott, and C.E. Rusell. 1985. Tree crops as soil improvers in the humid tropics. pp. 327-358. In: M.G.R. Cannell, and J.E. Jackson (eds.) Attributes of trees as crop plants. 592 p. Institute of Terrestrial Ecology. Natural Environment Research Council. Huntingdon, England. 41 Schonau, A.P.G., R.V. von Themaat, and D.I. Boden. 1980. The importance of complete site preparation and fertilizing in the establishment of W. pp. 1-18. In: Proceedings of the forest. seed, nursery, and. establishment research working group, third meeting. Saasveld Forestry Research Station, George, South Africa. South Africa Forestry Research Institute, Wattle Research Institute, University of Natal. Shrestha, R.B., L.B.S. Tulaadhar, and P.K. Tyysljarvi. 1980. Manual on nursery and plantation management practices for Community Forestry Development Project. HMG/ UNDP/FAO Community Forestry Development Project. Kathmandu, Nepal. 102 p. Sloan, J. 1988. Auger hole shape, size, and tree placement affect survival and root form of planted ponderosa pine in south central Idaho. Proceedings combined meeting of the Western forest nursery associations. Fort Collins, Colorado. Rocky Mountain Forest and Range Experiment Station, Forest Service, U.S. Dept. of Agriculture. General Technical Report RM:167. Squire, R.O. 1977. Interacting effects of grass competition, fertiliing, and cultivation on the early growth of Pinua {aaiaua D. Don. Australian Forest Research 7:247-252. Stewart, H.T.L., and D.P. Gwaze. 1988. Effect of fertilizer and sowing date on eswtablishment and early growth of direct- seeded M albida in Zimbabwe. Nitrogen Fixing Tree Research Reports 6:40-42. Thompson, D.A. 1985. Leucaena leucocephala trials in Jamaica. Leucaena Research Reports 6:60-62. Thompson, D.A. 1986. Leucaena leucocephala and other fast growing trees. Organization of American States, Caribbean Region LEUCAENA project, and Department of Forestry and Soil Conservation, Kingston, Jamaica. 36 p. Torres, F. 1983. Role of woody perennials in animal agroforestry. Agroforestry Systems 1:131-163. Umali-Garcia, M., J.S. Libuit, and R.L. Baggayan. 1988. Effect of Rhizobium inoculatio on growth and inoculation of Albizia falcataria (L.) Fosh. and Acacia Manguiu wild. in the nursey. Plant and Soil 108:71-78. Vearasilp, T. 1981. Digestibility of rice straw rations supplemented with Leucaena W19 ansd 9111121311: maculaua. Thailand Journal of Agricultural Science 14: 259- 264. 42 Wellington K.E. 1989. Agricultual Director, Alcan Jamaica Company LTD. Manderville, Jamaica. Personal Communication. White, K.J. 1985. Intercropping strategies and practices in the Bhabar Terai of Central Nepal. Manual No. 1 Sagarnath Forest Development Project. Ministry of Forests. Kathmandu, Nepal. 57 p. White, K.J. 1986. Tree farming in the Bhabar Terai of Central Nepal. Manual No. 2 Sagarnath Forest Development Project. Ministry of Forests. Kathmandu, Nepal. 199 p. Wilson, G.F., B.T. Kang, and K. Mulongoy. 1986. Alley cropping: Trees as sources of green-manure and mulch in the tropics. Biological Agriculture and Horticulture 3:251-267. DIRECT SEEDING OP FODDER TREE SPECIES IN JAMAICAN PASTURES Abstract. .A ‘trial of fifteen fodder' tree species ‘was established by direct seeding in an improved pasture in St. Ann Parish, Jamaica. The objective of the experiment was to determine if any of the species have potential to be incorporated into the forage component of Jamaican livestock production systems. After ten months of growth pauhiuia yariegaua, Lauuaaua_hybrid.Kx3, Sesbania gpauaiflpra NFTA 835 and 872, and Sesbauia aaapan NFTA 812 and 873 showed satisfactory survival, height and diameter growth. Cajanua uajan performed satisfactorily for six months, but by study's end had been infected with a root rot and died. All other species performed poorly under direct seeding on this site. INTRODUCTION Trees can be extremely valuable animal fodder resources. Torres (1983) described the widespread reliance on tree fodder as a dry season emergency feed in Australia. Le Houerou ( 1980) edited a thorough compilation on the importance of tree browse to animal production in Africa. In Nepal, over fifty percent of fodder needs are supplied from trees (Pandy 1982). Surveys show that up to 93 percent of Nepalese farmers cultivate fodder trees (Malla 1988), and that trees are the single most important fodder source on some farms (Rusten 1989). In other countries tree fodder is currently receiving attention as a means of increasing livestock production on small farms (IDRC 1990, Logan and Radcliffe 1985). In the past, tree fodders have had little impact on Jamaican livestock production (JLA 1983) . Pasture grasses are the main source of livestock feed (Kaplan et al. 1976). These grass resources are nutritious and able to sustain high levels of animal production during the wet season (JLA 1983). 43 44 However, during the annual December to March dry season (CRIES 1982) grasses lose much of their nutritional value. Protein is one of the most important nutrients for sustained animal maintenance and production (FAO 1988) . Crude protein levels in tropical grasses, as high as 14% during the wet season, decline to 5 to 6% during the dry season. This is below the 7% maintenance requirement reported for mature cattle in the Caribbean (Ahmed 1986, Paterson et al. 1987). Low protein, dry forage diets are also low in digestibility and.remain longer in the stomach.(Minson 1980, Paterson et a1. 1988) . This reduces feed intake and causes corresponding deficiencies of minerals, vitamins, and energy. Restricted to a dry forage diet, livestock decline in health and may die (Vohnout and Jimenez 1975, FAO 1988). For a variety of reasons, incorporating fodder trees into pastureland can help overcome this dearth of quality dry season feed (Robinson 1985) . Trees have extensive root systems that.tap subsoil water throughout the year (Sanchez et al. 1985). This enables trees to retain foliage in the dry season, providing a timely supply of fresh succulent fodder when grass resources are dry. Tree fodders are nutritous throughout the year (Reynolds and Adeoye 1986, Akkasaeng et al. 1989), and contain crude protein levels of 10-20% with some species having levels as high as 25-30% (IDRC 1990). Although too many trees can decrease grass production by competing for light and moisture, at optimal spacing trees 45 ameliorate environmental conditions which indirectly increase pasture and animal production (Torres 1983) . Tree roots enhance soil aeration and water permeability, and tree shade helps decrease soil temperature and desiccation. These factors improve the microsite of the pasture plants (Nair 1983) . Deep tree roots also recycle soil nutrients which have leached beyond the grass rhizophere and through litterfall redeposit them on the soil surface (Sanchez et al. 1985). If nitrogen fixing trees are used, the soil enriching ability of the tree litter may lead to increases in grass production (Lowry 1989, Lowry et al. 1988). Trees also provide animals with wind and sun protection. In the tropics, such stress reduction increases animal efficiency and productivity (Daly 1984). There is increased interest by researchers, development organizations, and large livestock producers to incorporate fodder trees into Jamaican livestock production (Ruegsegger 1990, Wellington 1989, JADF 1988). To date most efforts have concentrated on glipicidia aapiuu and Leucaeua leucocephala. Considering the vast number of valuable fodder trees already identified (Pandy 1982, Singh 1982, Le Houerou 1980), it is reasonable to believe that many other species, both native and exotic, hold potential for Jamaica. The objective of this study was to identify other fodder tree species that may be useful in Jamaica and determine if they can be established by direct sowing. Fifteen fodder tree species were established 46 by direct sowing in an improved grass pasture and evaluated by survival and growth measures ten months after sowing. Materials and Methods Species: Based on a search of the literature (Pandey 1982, Singh 1982, Le Houerou 1980) , fifteen fodder tree species were included in the trial (Table 1). Only one of the species, Leupaena, is "native" to Jamaica (Adams 1972) . Of the remainder all but two have been naturalized in Jamaica or the Caribbean for a long period (Adams 1972, Little and Wadesworth 1964, Little et al. 1974). The two newly introduced exotic species are Chamaecytisus palmensis and Flemingia macrophylla. Seed for thirteen of the species were provided by the Nitrogen Fixing Tree Association (NFTA)1. Seed of Bauhinia variagata and one of the Albizia aauau sources were collected locally. All of the species except Bauhinia vauiegaua are nitrogen fixing species (Allen and Allen 1981, Koul et al. 1989). Site: The study site was located on a flat grass pasture at Alcan Jamaica Company's Rio Hoe Farms near Moneague, St. Ann Parish. The surrounding area is comprised of rolling pastures and steep forested hills. Elevation of the site is 500 meters. Annual temperatures vary from 15—25°C, and average annual precipitation is 200 cm (Baker 1968) . Holdridge (1967) classifies this area as a subtropical moist forest life 1Nitrogen Fixing Tree Association, P.O. 680, Waimanalo, Hawaii 47 zone. Suggested land use for the area is pasture and livestock production and natural forest cover (Baker 1968) . The predominate pasture species on site are guinea grass (Realm miner) and Para grass (music Mica) - These species are considered "improved grasses" in Jamaica and are commonly used in livestock production (JLA 1983) . The soil is classified as a St. Ann clay loam, a medium deep red clay over white limestone which contains rich deposits of bauxite (Baker 1968) . Available phosphorus and potassium are considered low, nitrogen levels are considered adequate (Hewitt 1964). Soil analysis determined a soil pH of 7.48, total nitrogen of 0.42%, and availalbe phosphorus and potassium of 7 and 95 ppm, respectively. Establishment: Site preparation was conducted two weeks before sowing operations. First, the pasture was mowed with a tractor. Then, sowing pits of 10x10x10 cm were excavated with hand tools, the soil was finely tilled and replaced. Seeds were sown at a 1-2 cm depth. Before sowing, seeds were pre- treated and inoculated with species specific Baizobium according to accepted guidelines (NFTA 1987-90). The only exceptions were: 1) Eleuinqia uacrophylla, which was not pretreated with hot water due to conflicting information, and 2) the peamodium species, which did not receive fihizppiuu inoculum because none was available. As a hedge against mortality multiple seeds were sown per pit (Halliday and Naiko 48 1984). Depending' on. expected. germination. and. available supplies the number of seeds sown varied from 4 to 10. Competition from pasture grasses for moisture, nutrients, and sunlight was intense. Depending on climatic conditions, grass species grew from 70 to 150 cm in height per month. Manual weed control of these grasses occurred monthly and took two forms: 1) the cutting back of all above ground vegetation within the trial, and 2) the removal of all vegetation, including roots, from a 40 cm diameter circle around each seedling. This second operation maintained a weed free zone in which the seedling could thrive. Monthly weed control was conducted for six months. Experimental Design: Each species in the trial was established in a single unreplicated block of twenty-five sowing positions at a spacing of 1.5 x 1.5 meters. The nine interior positions of each block were used to measure survival and growth. Eight weeks after sowing the interior pits were thinned to contain one seedling. Germination was measured three weeks after sowing. Nodulation was observed at eight weeks. Survival and growth were measured at ten months. RESULTS AND DISCUSSION Germination. Despite a three week drought immediately following seed sowing, most of the species exhibited satisfactory germination. Only Alpiaia,lappapk, Ay_pupua;a, Remedies 91:21:19.5. and Elemineia magnesium had germination 49 rates below 50 percent. Additionally, Albizia aauau (local) and gnauaecytisus palmeusis did not germinate (Table 1), and will not be discussed further. Some species displayed delayed germination. This is discernible in Table 1 where survival is greater than germination. It is speculated that dry conditions during the drought temporarily suspended seedling emergence. Once the rains returned more germination was observed. Only with pa gyugides was delayed germination great, increasing from 5 to at least 20 percent. Hot and dry conditions during the drought had a negative effect on pesmodium uicaraguense, although germination was 60% at three weeks most of the germinants died before becoming established. This species appears very sensitive to heat stress and weed competition. Nodulation. With the exception of §_._ varie at , field examination showed that all species in the study, including those not inoculated, formed root nodules. In all cases nodules were fresh, pink and appeared to be "effective" (Postgate 1987). Survival and Growth. Subsequent survival after germination was satisfactory for only six of the species after ten months (listed in Table 1). All six, Bauhinia yapiagaua, W, and the four Saapauia species, had survival rates that when combined with sowing multiple seeds per pit ensured high site occupancy. These species also demonstrated fast growth despite heavy vegetative competition. After three months these 50 .oo>a>usn mmmmmmmummdd amdmmflmma no coauoomm eco xaco .nnucoe an :msoucu >uouueunaues ooSUOuuom e>sn menu newcomm .uuouos m.~. arose: unseen an assesses as use .nnucoa CA as some Ca ocqcaesou condenses nuconoumou a As>w>usm .csoe moose uo wanes: Heuou co ocean .exooz sewn» us caduecASuoo HNMV+ .mdamm mummamm new: nsoexcocme a4 mmamm aflmwmdd «ouoz an on woo me up armless: commmmlmwmmmmmw so we saw ms mm was res: cannon assesses an no can mm mm «so cssz canon access assesses we mm son as ms mam ass: ceuoduaesmuo assesses mm mm sen mm as nxz than»: anemones I- as cow 4 a has res: eaaasmosuus assessmam in +o~ +msa o.o oo Ann can: oecosoeusuwc Ssaooenoo In as can on m one res: nooaouso ssaeosuoa II II II II II one can: mascoaasm essauaumnEcco It in i: I: do one can: cease uscmfieo an me men as ea sauna caucusesa> cesarean I: It ii I: ll Aequ cases emumnac In an mos an no man res: sass: oneness II on moH an an new fiahz euoooum edwwnad II ea can v ea - «on oak: muennea saneoac j IEUI IOI lvl nonessdo anode: He>q>usm nauseouauoo OOH poem newcomm I ”A.“ O one snows: .Hs>«>wse . axes: so u as. sodasnumueo .« canes 51 species, with the exception of Laucaaua, were in codominate or dominate positions compared to the pasture vegetation. Although.weed control was continued for six months, it was not necessary after four months. After six months, fieapa_n_1_a_ species established a closed canopy which suppressed grass growth. Qajanus uajan performed very well for six months, height growth was 292 cm - greater then all species except fiy grandiflora NFTA 835 and £1 sesbau NFTA 812, and survival was 67 percent (Roshetko et al. 1991). By ten months all the gajanus trees were infected with a root rot and most were dead. Field identification of the pathogen indicated Macrophomina phaseolina (Sinclair et al. 1987), a common disease of Qajanus (Daniel and Ong 1990). The other six species have performed poorly under direct seeding in pastures. The Albizia species, L W, and the Desuoaiuu species are all recommended for direct seeding, if weed control is practiced through establishment (NRC 1979, 1980, NFTA 1987, 1988a, 1989a, Jackson 1987, Paterson 1933). In. this study' monthly' weed control for six :months was inadequate to furnish these species with a competitive advantage over the pasture grasses. After ten months, pasture grasses continue to dominant and suppress all of these tree species except L uaarpphylla, which is in a codominant position. All six of these species also displayed low germination and survival rates. 52 In Indonesia, A1 aauau survived poorly when planted in a grass ecosystem. In the same study, 5... mm survived satisfactorily, but grew slowly for three years. In the fourth year growth improved greatly, and fodder production equalled that of Laucaeua and.glipipiaia (Blair et al. 1988). For the species that performed poorly in this study, more intensive weed control for a longer period of time may lead to acceptable establishment. This represents an increase expense of time and money. In view of the superior growth of L yaziagata, Laucaena, and. the §esbania species 'under ‘the original weed control conditions, this increased expense would not be justified. CONCLUSION From this study it appears that pauhinia yauiagaua, Laupaena hybrid KX3, §esbania grandiflora 835 and 872, and §asbania sesban 812 and 873 can be established in improved pastures by direct seeding. All of these species are valued fodder resources (Pandy 1982, NRC 1984, NFTA 1990), and have potential to supplement Jamaican livestock production. Follow up studies to determine nutritive value of these species are reported in the following chapter. Of the species that performed poorly in this study one deserves special mention. gajauus pajau is a common pea crop in Jamaica, but it is unknown as a fodder crop. Under proper management this species can produce both pods and fodder. 53 Although it had died out by the end of the study, up aajau did establish successfully by direct seeding. It may hold great potential as a fodder crop in Jamaica. The other species that performed poorly in the study may have a useful role as fodder resources in Jamaica, but because of their low survival and relatively slow growth are not recommended for direct seeding into improved grass pastures for this site. Improved.weed control, fertilization, and use of seedling or stump sprouts (Jackson 1987) may be useful in establishing these species in grass pasture ecosystems. Species that demonstrated low germination should be retested to determine if failure was due to poor seed quality. ACKNOWLEDGEMENTS The author would like to thank the Jamaica Agricultural Development Foundation and the United States Agency for International Development for financial support, Alcan Jamaica Co., .Agricultural. Division, for logistical and 'technical support and the use of the research site, and the Nitrogen Fixing Tree Association for the seeds. BIBLIOGRAPHY Adams, C.B. 1972. Flowering plants of Jamaica. MacLehose & Co., The University Press Glasgow, Scotland. 848 p. .Ahmed, B. 1986. Growing forage legumes. CARDI Factsheet.No. AP-F/21.86. Caribbean Agricultural Research and Development Institute. St. Augustine, Trinidad. West Indies. 4 p. Akkaseng, R., R.C. Gutteridge, and. M. Wanapat. 1989. Evaluation of trees and shrubs for forage and fuel-wood in Northeast Thailand. The International Tree Crops Journal 5:209-220. Allen, O.N. and E.K. Allen. 1981. The Leguminosae: A source book of characteristics, uses and nodulation. Wisconsin Press. Madison, Wisconsin. 812 p. Baker, G.H. 1968. Soil and landuse survey No. 24 Jamaica, Parish of St. Ann. Regional Research Centre, Department of Soil Science, University of the West Indies, Trinidad. Blair, G.J., M. Panjaitan, D. Ivory, B. Palmer, and M. Sudjadi. 1988. An evaluation of tree legumes on an acid ultisol in South Sumatra, Indonesia. Journal of Agricultural Science (Cambridge) 111:435-441. CRIES. (1982). Jamaican resource assessment. Comprehensive resource inventory and evaluation system project (CRIBS) . Michigan State University - U.S. Department of Agriculture - The Ohio State University. Department of Resource Development, Michigan State University, East Lansing; Michigan. Daly, J.J. 1984. Importance of trees for cattle. Tropical Grasslands 18:154-156. Daniel, J.N., and C.K. Ong. 1990. Perennial pigeonpea: a multi-purpose species for agroforestry systems. .Agroforestry Systems 10:113-129. FAO. (1988) . Some guidelines for designing integrated farming systems. FAO Regional Office for Latin America and the Caribbean. Santiago, Chile. 50 p. Halliday J., and P. Nakao. 1984. Technical notes on the germination of leguminous tree seeds. pp. 231-234. In: Simposio Sobre Fixacao de Nitrogenio em Arvores Tropicais. Brazilian Journal of Agricultual Research ‘Volume 19 (Special Issue). Hewitt, C.W. 1964. Soil technical sheets. Agric. Chem. Div., Ministry of Agriculture, Jamaica. 144 p. 54 55 Holdridge, L.R. 1967. Life zone ecology. Tropical Science Center. San Jose, Costa Rica. IDRC. 1990. Shrubs and tree fodders for farm anaimals. International Development Research Centre. Ottawa, Ontario, Canada. 349 p. Jackson, J.K. 1987. Manual of afforestation in Nepal). Nepal-United Kingdom Forestry Project. Department of Forestry. Kathmandu, Nepal. 402 p. JADF. 1988. Jamaica Agriculture Research Program Annual Report 1988. Jamaica Agriculture Development Foundation. Kingston, Jamaica. JLA. 1983. Livestock manual for the tropics. The Jamaica Livestock Association. Kingston, Jamaica. 406 p. Kaplan, I., H.I. Blustein, K.T. Johnston, and U.S. McMorris. 1976. .Area handbook for Jamaica. Foreign Area Studies of the American University. Washington D.C. 332 p. Koul, V.K., S.D. Bhardwaj, and A.N. Kaushal. 1989. Bauuiuia yapiegata does not nodulate. Nitrogen Fixing Tree Research Report 7:75. Le Houeron , N.H. (Editor) 1980. Browse in Africa: The current state of knowledge. ILCA Addis Ababa. 491 p. Little, E.L., and F.H. Wadesworth. 1964. Common trees of Puerto Rico and the Virgin Islands. Agriculture Handbook No. 249. USDA Forest Service. Washington D.C. 556 p. Little, E.L., R.O. Woodbury, and F.H. Wadesworth. 1974. Commons trees of Puerto Rico and the Virgin Islands, Second Volume. Agriculture Handbook No. 449. USDA Forest Service. Washington D.C. 1024 p. Logan, I.A., and J.E. Radcliffe (Eds.). 1985. Fodder trees - A summary of current research in New Zealand. Crop Research Division - Department of Scientific and Industrial Research. Christchurch, New Zealand. 49 p. Lowery J.B. 1989. Agronomy and forage quality of Alpiaia lappack in the semi-arid tropics. Tropical Grasslands 23:84- 91. Lowery J.B., J.B.C. Lowery, and R. Jones. 1988. Ehhanced grass growth below the canopy of Albizia lebpeck. Nitrogen Fixing Tree Research Report 6:45-46. 56 Malla, Y.B. 1988. The case for placing more emphasis on privatet tree planting programmes. Forestry Research Paper Series No. 13. HMG—USAID-GTZ-FORD-WINROCK Project. Kathmandu, Nepal. 46 p. Minson, D.J. 1980. Nutritional differences between tropical and temperate pastures. pp. 143-157. In: F.H.W. Morley (ed.) Grazing animals. Elsevier Publishing Company, Amsterdam. Nair, P.K.R. 1983. Tree integration on farmlands for sustained productivity of smallholdings. pp. 333-350. In: W. Lockeretz (ed) Environmentally sound agriculture. Praeger, New York. NFTA. 1987. Nitrogen Fixing Tree Highlights: The multipurpose rain tree, Samana samau. Nitrogen Fixing Tree Association, Waimanalo, Hawaii. 2 p. NFTA. 1988a. Nitrogen Fixing Tree Highlights: Alpiaia M - A promising fodder tree for semi-arid regions. Nitrogen Fixing Tree Association, Waimanalo, Hawaii. 2 p. NFTA. 1988b. .Nitrogen Fixing Tree Highlights: Cajauua pajau - It's more than just a pulse crop. Nitrogen Fixing Tree Association, Waimanalo, Hawaii. 2 p. NFTA. 1989a. Nitrogen Fixing Tree Highlights: Flemingia maazpphylla - A valuable species in soil conservation. Nitrogen Fixing Tree Association, Waimanalo, Hawaii. 2 p. NFTA. 1989b. NFTA cooperative planting program - Establishment guidelines. Nitrogen Fixing Tree Association, Waimanalo, Hawaii. 36 p. NFTA. 1990. Perennial Sesbania production and use. Nitrogen Fixing Tree Association. Waimanalo, Hawaii. 41 p. NRC (NATIONAL RESEARCH COUNCIL). 1979. Tropical legumes: Resources for the future. National Academy Press, Washington, D.C. 332 p. NRC (NATIONAL RESEARCH COUNCIL). 1980. Firewood crops: Shrub and tree species for energy production. National Academy Press, Washington, D.C. 236 p. NRC (NATIONAL RESEARCH COUNCIL). 1984. Leucaena: Promising forage and tree crop for the tropics. Second Edition. National Academy Press, Washington, D.C. 100 p. Pandey, K. K. 1982. Fodder trees and tree fodder in Nepal. Swiss Development Cooperation. Berne, Switzerland. 107 p. 57 Paterson , R. T . 1988 . The introduction of legumes into degraded tropical pastures. Ph. D. Thesis. University of Reading. Reading, England. Paterson, R.T., G.A. Proverbs, and J.M. Keoghan. 1987. The management and use of forage banks. Caribbean Agricultural Research and Development Institute. St. Augustine, Trinidad, West Indies. 21 p. Paterson, R.T., J.M. Keoghan, and G.A. Proverbs. 1988. Forage—based grazing systems for the drier parts of the caribbean region. Caribbean Agricultural Research and Development Institute. St. Augustine, Trinidad. West Indies. 23 p. Postgate, J. 1987. Nitrogen fixation. Edward Arnold LTD. London, Great Britain. 73 p. Reynolds, L., and S.A.O. Adeoye. 1986. Planted leguminous browse and livestock production. Paper presented at: International workshop on alley farming for humid and sub- humid.regions of tropical.Africa. Ibadan, Nigeria. ZMarch 10- 14, 1986. Robinson, P.J. 1985. Trees as fodder crops. pp. 281-300. In: M.G.R. Cannell, and J.E. Jackson (eds.) Attributes of trees as crop plants. 592 p. Institute of Terrestrial Ecology. Natural Environment Research Council. Huntingdon, .England. Roshetko, J.M. , D.O. Lantagne, M.A. Gold. 1991. Direct seeding of fodder tree species in Jamaican pastures. Nitrogen Fixing Tree Research Reports 9:68-70. Ruegsegger, G. 1990. On-farm evaluation of Leuaaaua, Gliricidia and legume/grass conserved fodder as high protein forages in dairy cattle rations. Paper presented at the Second Annual Seminar for presentation of progress reports on Jamaican Agricultural Research Projects. The Jamaican Agricultural Development Foundation. Kingston, Jamaica. September 18-20, 1990. Rusten, E. 1989. .An investigation.of an indigenous knowledge system and management practices of tree fodder resources in the middle hills of central Nepal. Ph.D. Dissertation. Department of Forestry, Michigan State University. 282 p. 58 Sanchez, P.A., C.A. Palm, C.B. Davey, L.T. Scott, and C.E. Rusell. 1985. Tree crops as soil improvers in the humid tropics. pp. 327-358. In: M.G.R. Cannell, and J.E. Jackson (eds.) Attributes of trees as crop plants. 592 p. Institute of Terrestrial Ecology. Natural Environment Research Council. Huntingdon, England. Sinclair, W.A., H.H. Lyon, and W.T. Johnson. 1987. Disease of trees and shrubs. Comstock Publishing Associates, Cornell University Press. Ithaca, New York. 574 p. Singh, R.V. 1982. Fodder trees of India. Oxford and IBH Publishing Co. New Delhi, India. 663 p. Torres, F. 1983. Role of woody perennials in animal agroforestry. Agroforestry Systems 1:131-163. Vohnout, K., and C. Jimenez. 1975. Supplemental by-product feeds in pasture-livestock feeding systems in the tropics. pp. 71-82. In: E.C. Doll, and G.O. Mott (eds.) Tropical Forage in Livestock Production Systems. American Society of Agronomy - Special Publication No. 24. Madison, Wisconsin. 104 p. Wellington, K.E. 1989. Agricultural Director, Alcan Jamaica Company LTD. Manderville, Jamaica. Personal communication. ESTIMATED NUTRITIVE VALUE OP NATIVE AND EXOTIC FODDER TREE SPECIES IN JAMAICA AND THEIR COMPARISON TO COMMON PASTURE GRASS SPECIES Abstract. The estimated nutritive value of eight native and six exotic Jamaican tree fodder species are presented. Exotic species generally have superior crude protein and phosphorus levels, and lower fiber contents. Native species have greater dry matter content. Ash, calcium, and magnesium percentages are about equal between the two groups. The exotic species appear to have greater potential as livestock feed resources. Samples were collected and analyzed in January and April of 1991, respectively. Compared to common pasture grasses in Jamaica, tree fodder species are inferior in digestibility and mineral content, but superior in crude protein. This characteristic makes tree fodders excellent feeds for the dry season when grass crude protein levels fall below animal maintenance requirements. During seasonal dry periods crude protein levels in tree species do not decrease because of their deep perennial root systems. In Jamaica, tree fodder serves a supplementary role in livestock production, it is not meant to replace grass as the major dietary input. INTRODUCTION Trees can be extremely valuable animal fodder resources. Torres (1983) has described the widespread reliance on tree fodder as a dry season emergency feed in various parts of Australia. Le Houerou (1980) has edited a thorough compilation on the importance of tree browse to animal production in Africa. In other countries tree fodder is currently receiving attention as a means of increasing livestock production on small farms (Logan and Radcliffe 1985, IDRC 1990). In Nepal, over fifty percent of fodder needs are supplied from trees (Pandy 1982). Surveys have shown that up to 93 percent of Nepalese farmers cultivate fodder trees (Malla 1988) , and that trees and woody vines are the single most 59 60 important fodder source on some farms (Rusten 1989) . In many parts of the world, fodder trees are deliberately incorporated on to small farms to expand fodder resources, particularly for dry season use (Von Carlowitz 1989). These trees are planted on contour lines to stabilize soil (Fonzen and Oberholzer 1984) , on boundaries to serve as living fences (Budowski 1987) , with forage grasses or shrubs to form fodder banks (Nitis et al. 1987), or scattered around the farm where site characteristics are too poor to support agronomic crops (Ivory 1990). In Jamaica, the use of tree fodder as a livestock feed has been documented for over 200 years (Browne 1789, Lunar 1814, Hooper 1886, Swaby 1941). More recently, in a study of small cattle farmers in northwest Jamaica, Morrison (1991) found that 95% of farmers utilized tree species as dry season feed. Although the use of tree fodder is known in Jamaica, it is not a widespread practice and is mainly limited to small farmers. Tree fodder is under utilized, and given only passing mention in the Jamaican Livestock Association's management . manual (JLA 1983) . Generally, large livestock producers ignore tree fodders, and concentrate on the use of grass and pastures. Most Jamaican grass species contain ample amounts of protein, energy, and minerals when grown under proper conditions (JLA 1983) . Animals raised on these grasses maintain high levels of productivity during the wet season. 61 This situation changes drastically in the dry season, when livestock production begins to suffer from a paucity of nutritious feed. Typically, Jamaica experiences two dry seasons per year; a short period from July to August and a long period from December to March (CRIES 1982) . During these periods precipitation is insufficient to sustain vigorous grass production, and livestock must survive on dry pasture grasses (Kaplan et al. 1976). The nutritional quality of this forage is low. Tropical grass nutritive values decline during the dry season. Crude protein contents, which may be as high as 14% under normal conditions, diminish to 5 or 6%. This level is below the 7% maintenance requirement reported for mature cattle in the Caribbean (Ahmed 1986, Paterson et al. 1987). Furthermore, low protein and dry forage diets are low in digestibility and remain longer in the stomach (Minson 1980, Paterson et al. 1988). This results in reduced feed intake and may cause corresponding deficiencies of minerals, vitamins, and energy. Restricted to a dry forage diet, livestock often suffer a decline in health and in the extreme may even die (Vohnout and Jimenez 1975, FAO 1988). For a variety of reasons, incorporation of fodder trees on pastureland can be helpful in overcoming this dearth of quality dry season feed.(Robinson 1985). Trees have extensive root systems that tap subsoil water throughout the year (Sanchez et al. 1985). This enables trees to retain foliage 62 in the dry season, providing' a timely supply of fresh succulent fodder when grass resources are dry. The deep roots also capture and recycle soil nutrients which have leached beyond the grass rhizophere (Sanchez et al. 1985). Besides remaining fresh, the foliage of many fodder trees are also nutritious. Most tree fodders contain crude protein levels of 10-20%, some species having levels as high as 25-30% (IDRC 1990). If spaced too closely, trees can decrease grass production by competing for light and moisture. Conversely, at appropriate spacing trees also ameliorate environmental conditions which indirectly increase pasture and animal production (Torres 1983). Tree roots enhance soil aeration and water permeability, and tree shade helps decrease soil temperature and desiccation. These factors improve the microsite of the pasture plants (Nair 1983). If nitrogen fixing trees are used, the soil enriching ability of the tree litter may lead to increases in grass production (Lowry et al 1988, Lowry 1989). Trees also provide animals with wind and sun protection. In the tropics, such stress reduction increases animal efficiency and productivity (Daly 1984). Considering the important role tree fodder plays in other countries, there is reason to believe that its potential in Jamaica is not being fully realized. The purpose of this paper is twofold; 1) to present the estimated nutritive values of fourteen Jamaican fodder tree species recently collected 63 and analyzed, and 2) to compare the nutritive values of these tree species with those of the common grass species used in Jamaican livestock production. MATERIALS AND METHODS Species: The fodder tree species analyzed in this study can be assigned to one of two groups: Native Exotic flgosimum alicastruu Bauhinia variegata Bursera simaruba Leupaena hybria gecropia peltata Sasbania grandiflora (Ni-Ea §_3_§)_ githarexylum fruticosum Sesbania grandiflora (NFTA 872) Guazuma ulmifolia gasbania aesban (NFTA 812) Haemaloxylum campechianuu fiesbania sesbau (NFTA §7§) Ipomoea tiliacea Eiapidia piscipula Native Fodder Species. The native species are either indigenous to or naturalized in Jamaica and occur frequently in forests, pastures, or bushlands (Little and Wadesworth 1964, Adams 1972, Little et al. 1974). Seven of the plants are trees, Ipomoea tiliacaa is a herbaceous vine. The term "tree fodder" will be used to include 1; piliacea in future discussion. Each of the species is used.to axdifferent extent by Jamaican farmers. Bppsimum alicastruu, puusera simazupa, guaauua ulmifplia, and la piliaaaa are considered excellent feeds that are readily accepted by livestock. gepxppia peltata. WW. and Moi—die 21__p__a' sci ul are regarded as good feeds, but inferior to the others in quality and acceptability. Qitaareuyluu fuuuicosuu is a dry season 64 emergency feed that is only used when other resources are lacking. Seven of the native tree fodders are used as goat and cattle feed, I; piliacea is fed to goats and pigs. The use of fly campechiauum is limited because of the presence of sharp thorns on its branches. Leaf samples of the native species were collected from forests and pastures in Green Park, Trelawny Parish. Elevation in the area is 150 meters. Annual precipitation averages 100 cm, annual temperatures vary between 20-30°C (Baker 1970). Ecologically, this area is classified. a subtropical dry forest life zone by Holdridge (1967). The land in the area varies from flat pasture and cropland to steep wooded hillsides. Suggested land use is pasture/livestock production and natural forest cover (Baker 1970). The soil in the area is Bonny Gate stony loam, a thin brown or reddish soil over hard limestone (Baker 1970). It is generally considered to be infertile and slightly alkaline. Available nitrogen, phosphorus, and potassium are all low (Hewitt 1964). Exotic Fodder Species. The exotic species are all fast growing, highly valued fodder trees of the Leguminosae family (NRC 1984, Jackson 1987, NFTA 1990). In Asia and Africa they are often established on private land specifically as livestock feed resources (NRC 1984, Jackson 1987, NFTA 1990). Their crude protein values vary from 19 to 29 percent (IDRC 65 1990). All of these species except Bauhinia vauiagata are nitrogen fixing trees (Allen and Allen 1981). These species are currently found in Jamaica, and although not presently utilized as fodder, have potential as protein rich feed supplements for livestock production. Lancaena lauppaapuala is native to Jamaica (Adams 1972). The Laupaeua.hybrid in the study is a cross of La lauapaapnala K636 and Lu giyezsifolia K156 (Moore 1991). All of these exotic species were established in a trial planting near Moneague, St. Ann Parish in May of 1990 (Roshetko et al., in press). Elevation of the site is 500 meters. Annual temperatures vary from 15-25°C, and average annual precipitation is 200 cm (Baker 1968) . Holdridge (1967) classifies this area as a subtropical moist forest life zone. The study site is located on a flat grass pasture. The surrounding area is comprised of rolling pastures and steep forested hillsides. Suggested land use for the area is pasture/livestock production and natural forest cover (Baker 1968). The soil is classified as a St. Ann clay loam, a medium deep red clay over white limestone which contains rich deposits of bauxite (Baker 1968). The soil is generally considered to be infertile and slightly alkaline (Hewitt 1964, Weir 1980). 66 Collection Procedure: Foliage samples for nutritive analysis were collected from each tree species. With the exception of gygpaluaua, samples were collected from.three to five specimens. For g_,_ peltata only two specimens were sampled. Only leaves (i.e. leaves, leaflets and.petioles) on the outside of the tree crown were included in the samples. Foliage samples of exotic species were grown in full sun, but foliage collected from native species were not necessarily grown under full sun, due to shading from competing vegetation. Native trees sampled were mature and of varying age. Exotic trees were seven months old when sampled. All specimens were healthy. Collection was conducted in mid- January, which corresponds to the beginning of the winter dry season, a period of considerable tree fodder use. Processing Procedure: Fresh weight of samples were recorded at collection. Samples were air dried for 24 hours and then oven dried at 61°C for 48 hours. After cooling, samples were ground using a Wiley Mill with a 1x1 mm screen. Approximate analysis values for crude protein (CP) , acid detergent fiber (ADF), neutral detergent fiber (NDF), ash, phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg) content were determined for each sample. These values are only’ estimates of each species' nutritive ‘worth. to livestock, but. are