Stylosanthes guianensis CIAT 184 – review of a tropical forage legume* Stylosanthes guianensis CIAT 184 – reseña de una leguminosa forrajera tropical

A comprehensive review, based on about 180 references, synthesizing research and development about Stylosanthes guianensis accession CIAT 184 is presented. This genotype has been widely tested across the tropics and was developed into commercial cultivars in several countries. Agronomic evaluations in a range of disciplines and other research in numerous countries of tropical America, Africa, Southeast Asia and China, and utilization of the legume by farmers are reported. CIAT 184 is particularly successful in southern China, where it gave origin to 5 cultivars. Its outstanding feature is an apparently durable tolerance of the fungal disease, anthracnose ( Colletotrichum gloeosporioides ), throughout the tropics, with exception of the savanna ecozone in tropical America. Further assets include adaptation to acid, infertile soils, drought tolerance and high production of nutritious dry matter that can be used as traditional forage for ruminants and as leaf meal or pellets to feed monogastrics. CIAT 184 has also found application for soil conservation and improvement (such as erosion control, mulch and green manure), alone or in association with crops and trees. Published work dealing with basic research, mainly conducted in China, is presented. The review concludes with a brief discussion on the current adoption of CIAT 184 and with suggestions regarding future perspectives and research needs.


Introduction and background
Stylosanthes Sw. [family Fabaceae (alt.Leguminosae)] is a mainly neotropical genus with currently 40 species accepted by the taxonomy of the Germplasm Resources Information Network (GRIN) of the US National Plant Germplasm System (GRIN 2023).Since many Stylosanthes species are found on infertile soils and tolerate drought, the genus, often referred to as 'stylo', has been attracting the attention of tropical forage legume researchers since the 1930s (Burt and Miller 1975).
Owing to its yield potential, a species of particular interest has been S. guianensis (Aubl.)Sw., better known up to the late 1960s by practitioners under the name of S. gracilis (Tuley 1968;Audru 1971).In addition to var.guianensis, 5 botanical varieties are recognized in this species by GRIN taxonomy (GRIN 2023).When no botanical variety is mentioned in this paper, 'S.guianensis' always refers to S. guianensis var.guianensis.
S. guianensis has a wide natural distribution ranging from latitude 23° N in Mexico to 27° S in Argentina (Williams et al. 1984) and exhibits a particularly wide morphological diversity (Costa 2006).Wild populations are mostly found on well-drained, low to medium fertility soils at low to medium elevations (10-1,700 masl, occasionally higher), mainly in regions with wet-subhumid climate 1 .Research for cultivar development and use of S. guianensis, until the mid 1970s an essentially Australian activity, is documented in the scientific literature.Compilations can be found in 4 key books : Burt et al. (1983); Stace and Edye (1984); de Leeuw et al. (1994); and Chakraborty (2004a).A factsheet summarizing updated information on S. guianensis and its botanical varieties was recently published in the database, 'Tropical Forages -an interactive selection tool' (Cook et al. 2020).
From the 1970s onwards, an important issue has been tolerance of the fungal disease, anthracnose (caused by Colletotrichum gloeosporioides), the biotic factor constraining the use of all better-known Stylosanthes species, including S. guianensis (Chakraborty 2004b).The search for genotypes tolerant of this disease and adapted to infertile soils and drought, led to increased efforts, in the 1970s and 1980s, to broaden the genetic diversity available to the research community by exploring the native vegetation and collecting germplasm of Stylosanthes in (sub)tropical America.Among the various institutions involved in those collecting and conservation efforts was the International Center for Tropical Agriculture (CIAT), whose genebank conserved a total of 4,200 Stylosanthes accessions in May 2020 (Schultze-Kraft et al. 2020).The subject of this review, S. guianensis accession CIAT 184, stems from one of CIAT´s early collecting endeavors in Colombia.
In this paper we review CIAT 184 by compiling, in an essentially region-and country-oriented approach, information from research conducted throughout the tropics and subtropics, with the objective to contribute to continuing interest of the research and development (R&D) community in this genotype.The focus is on information from studies that we consider to be of interest to a reasonably broad readership.Wherever possible, we concentrated on accessible published literature as information sources.A brief analyis of the reasons for the success of CIAT 184 in some regions and lack of adoption in others and some thoughts on future perspectives are presented.

Origin of CIAT 184
The origin dates back to 12 October 1973, when one of the authors (R.S.K.) collected seeds from a S. guianensis population at a location (03°13'28'' N, 76°33'45'' W; 975 masl) close to Jamundí town, south of Cali in the Valle del Cauca department, Colombia.The sampled plants were part of occasionally grazed roadside vegetation (major component: native Paspalum notatum).Topography 1 An important exception is the 'tardío' morphotype of S. guianensis, which is found in dry subhumid savanna climate on sandy and very acid, infertile soils.This morphotype is also known in the literature as 'S.guianensis var.pauciflora M.B.Ferreira & Sousa Costa', a taxon not accepted as a botanical variety by GRIN taxonomy, where it is referred to as S. guianensis var.guianensis.

97
Review of Stylosanthes guianensis CIAT 184 in the area is flat and the soil was recorded as a welldrained clayey Ultisol, acid (pH 4.9 in H 2 O) and with low phosphorus (1.8 ppm P; Bray II).Annual rainfall in the area is about 1,900 mm.
The collected seed sample entered the CIAT tropical forages collection as accession CIAT 184.Within subsequent germplasm exchanges with genebanks it was alternatively catalogued under different accession numbers (e.g.'ILCA/ILRI 164' and 'CPI 133548').

Morphological description
The following description is based on that by Cook et al. (2020) for S. guianensis var.guianensis in general, and complemented with accession-specific observations and measurements of morphological features provided by one of the authors (Y.H.) and with photographic illustrations (Figure 1).

Evaluation and anthracnose tolerance
Soon after incorporation of CIAT 184 into CIAT´s active germplasm collection, the production potential of this accession and its anthracnose tolerance became evident in field nursery observations and greenhouse studies conducted at the CIAT experimental station in Palmira, Colombia; its adaptation to acid, infertile soils was recorded in plots established in the experimental stations of Carimagua andQuilichao (CIAT 1975-1977).
From 1978 onwards, CIAT 184 was included in the legume germplasm set for evaluation in multilocational trials of the International Tropical Pastures Evaluation Network (RIEPT, its Spanish acronym).These trials were conducted in the humid and subhumid tropics of Latin America and the Caribbean, using a standardized methodology for evaluation of adaptation ('Regional Trials A') and agronomic performance ('Regional Trials B') (Toledo and Schultze-Kraft 1982).In 1982, CIAT 184 was also introduced in tropical China along with a set of other legumes and grasses for evaluation in Hainan (Liu Guodao et al. 2004a), within early cooperation between CIAT and the then South China Academy of Tropical Crops (since 1994: Chinese Academy of Tropical Agricultural Sciences, CATAS).
During 1989-1993 CIAT 184 was one of the 21 herbaceous legume accessions included in evaluation trials at 17 sites in 9 countries of Subsaharan Africa, within the West and Central African Animal Feed Research Project network (RABAOC, its French acronym), also using a standardized methodology (Rippstein 1998).In the 1990s it was part of the forage germplasm tested in 2 CIAT-CSIRO (Commonwealth Scientific and Industrial Research Organization) R&D cooperation projects in Southeast Asia: the 'South-east Asian Regional Forage Seeds Project ' (1992-1994) and the 'Forages for Smallholders Project ' (1995-1999), which subsequently continued in the form of 2 CIAT-Asian Development Bank (ADB) projects (2000)(2001)(2002)(2003)(2004)(2005) conducted in partnership with national R&D institutions in up to 8 countries (Stür et al. 2007).
The most relevant outcome of these early adaptation and agronomic performance studies around the globe was that in addition to adaptation to acid, infertile soils and drought tolerance, CIAT 184 showed a strong tolerance of anthracnose in comparison with the wellknown Australian commercial stylo cultivars, such as 'Cook', 'Endeavour', 'Graham' and 'Schofield', and with essentially all other S. guianensis germplasm accessions tested in those studies2,3 .

Tropical America
In the humid tropics, where it was one of 13 accessions from 5 different legume genera (10 species) evaluated in 32 'Regional Trials B' conducted between 1979 and1991 in Bolivia, Brazil, Colombia, Ecuador and Peru within the RIEPT network system, CIAT 184 proved to be outstanding with respect to environmental adaptation and dry matter production (Amézquita et al. 1991;Keller-Grein et al. 1992).In contrast, in RIEPT trials conducted in a subhumid savanna climate characterized by a distinct dry season, the accession was badly affected by anthracnose (Pizarro 1983;1985;1988).The presence of antagonistic bacteria on the tissue of stylo plants and prevailing narrow diurnal temperature fluctuations, both preventing the development and spread of the disease in the humid tropics, were reported to explain this contrast between the 2 ecozones (Lenné 1985).
Results of studies with CIAT 184 in tropical America that we found in the literature, beyond those reported in the aforementioned RIEPT trial compilations (Pizarro 1983;1985;1988) and that of Keller-Grein (1990), are as follows: Argentina.The potential of this accession for subtropical conditions with dry and cold winters became evident based on a 3-yr study on a sandy soil in the northwest of Corrientes province, northeast Argentina, where anthracnose is not a major issue and where CIAT 184 along with S. guianensis cultivar 'Graham' had the highest forage and seed yields among 35 accessions from 6 Stylosanthes species (Ciotti et al. 1999).In the same area on a phosphorus-deficient soil (3.4 ppm P; Bray I), an almost 100 % DM yield increase was observed in response to P fertilizer application as low as 11 kg P/ha for CIAT 184 (Ciotti et al. 2003).
Colombia.Low P requirements of CIAT 184 and strong response to very low P fertilizer doses in soil from the Colombian Llanos were also shown in the early work at CIAT (Schultze-Kraft 1976).However, since S. guianensis in general succumbed to anthracnose in the subhumid savanna ecosystem ('Llanos Orientales'), where CIAT´s pasture research in Colombia was focussed in the 1970s and 1980s, studies on CIAT 184 were not continued at that time and resumed in the 1990s in the humid-tropics part of the country (Caquetá department).There, Velásquez et al. (2004) showed that pre-weaned calves with access to a CIAT 184 pasture drank less milk from their dams, than calves with access to only grass pasture with their dams, allowing 21 % more milk per cow to be sold from the CIAT 184 group, while calves from this group made 30 % higher liveweight gains.
At the basic-research level in Colombia, Sarria et al. (1994), using an Agrobacterium tumefaciens-disarmed strain, showed that an efficient transfer of marker genes to CIAT 184 and regeneration of transgenic plants were possible.Subsequently, Kelemu et al. (2005) introduced a chitinase-encoding gene from rice into CIAT 184, which resulted in improved resistance of transgenic plants to foliar blight disease (Rhizoctonia solani).
Cuba.Based on germplasm evaluations within the RIEPT network (Pizarro 1988) and follow-up studies, CIAT 184 was considered a promising legume for pasturebased livestock production on poor soils.In 5-year seed production studies, 182-318 kg seed/ ha/ yr were harvested (Suárez and Villavicencio 1988).Although Review of Stylosanthes guianensis CIAT 184 the accession was not formally released, 'cv.CIAT-184' was promoted and used in protein banks.It has been dominating S. guianensis research in Cuba during the past 3 decades, with a focus on callus formation/tissue culture for generation of plants and on tolerance of soil salinity (Mesa et al. 1993;Fuentes et al. 2000Fuentes et al. , 2008Fuentes et al. , 2015;;González et al. 2000;Fuentes Alfonso et al. 2008).
Peru.CIAT 184 showed high productivity on the acid, infertile soils that prevail in the Peruvian humid tropics, with anthracnose tolerance (Pizarro 1983(Pizarro , 1985;;Keller-Grein 1990).This resulted in the Instituto Veterinario de Investigaciones Tropicales y de Altura (IVITA) and Instituto Nacional de Investigación y Promoción Agropecuaria (INIPA) releasing CIAT 184 in Peru as cultivar 'Pucallpa' in January 1985 (Reyes et al. 1985).The promoted use was mainly as a pasture legume for association with grass, in pure stands as a protein bank or to rehabilitate degraded pastures.
Subsequently, Reátegui et al. (1995) and Loker et al. (1997) reported from long-term on-farm research in the Pucallpa area (13 farms, 4 years management), where CIAT 184 (cultivar 'Pucallpa') was used as a legume species in grass-legume pastures, that 'Pucallpa' established well, even with no fertilizer input.However, with time its proportion in the mixture declined under farmers´ grazing management.It is noteworthy that, in those cases where the pastures were burnt to control weeds and pests, the persistence of 'Pucallpa' improved.
In more recent research in the Pucallpa area, partial substitution of forage for milk from dam by calves with access to a 'Pucallpa' pasture led to a 23 % increase of milk sold per cow without affecting growth of calves (Vela 2004a).Incorporating stubble of the legume in the soil had the same yield-increasing effect on a subsequent rice crop as 50 kg N/ha fertilizer (Vela 2004b).
Velásquez Ramírez et al. (2021) showed that, when used as cover crop, 'Pucallpa' had a significantly positive effect on soil restoration of alluvial gold mine spoils in the Peruvian Amazon, in terms of several soil chemical, physical and biological parameters.
Puerto Rico.Under commercial dairy farm conditions with 10 acres of stylo in the South Region (Guayama), an average of 10.8 t hay/ha with 25 % moisture content was obtained in 1991/92 in each of three 4-monthly cuts from cultivar 'Pucallpa', with mean CP concentrations of 16-18 %.Fed as a supplement to about 290 kg replacement heifers grazing common guinea grass, a daily ration of 8.7 kg of stylo hay per animal was able to replace 5 kg of commercial concentrate (A.Arias-Pedraza, pers.comm.December 2022).

Tropical Africa
R&D with CIAT 184 in tropical Africa concentrated on West and Central Africa and Madagascar.The accession was, however, not available for fodder bank studies conducted in West Africa by International Livestock Research Institute (ILRI) and partners in the early 1980s (Mohamed-Saleem and Suleiman 1986).In final reports of the RABAOC project on the agronomic evaluation of a set of grass and legume species in humid and subhumid environments in Benin, Cameroon, Central African Republic, Côte d´Ivoire, Ghana, Guinea, Nigeria, Senegal and Togo (NARS-CIRAD/EMVT-CIAT-ILCA 1995; Rippstein 1998), CIAT 184 consistently ranked among the top herbaceous legumes, out of 21 accessions representing 17 species.Further regional studies involving CIAT 184 (de Leeuw et al. 1994) are country-specific and referred to below.In some of the publications on R&D in tropical Africa, the accession is mentioned under its synonym numbers 'ILCA/ILRI 164' or 'FAO 46004'.
Benin.Saito et al. (2010) concluded from a tillage management experiment in southern Benin that manual tillage combined with a CIAT 184 stylo fallow is a recommendable practice for smallholder farmers to improve upland rice productivity.
Cameroon.Several studies with FAO 46004 (CIAT 184) were conducted in the Adamawa Plateau.In an agronomic evaluation of 17 Stylosanthes accessions, CIAT 184 was outstanding in terms of dry matter production, anthracnose tolerance and seed yield (400 kg/ha) (Yonkeu et al. 1994).While Pamo and Yonkeu (1994) showed increased grass production of a subsequent Urochloa ruziziensis crop after harvesting stylo for hay during 2 years, Enoh et al. (1999) reported from dry-season sampling of 4-5 yrs old on-farm and on-station stylo-grass (U. ruziziensis) and stylo-native pasture (mainly Hyparrhenia and Panicum) paddocks (which they called "fodder banks") increased dry matter yields and nutritive value in comparison with grass-only pastures.Poor management and yearly bush fires were mentioned as major limitations to the promoted stylo-grass technology.
Côte d'Ivoire.Based on a study at 5 benchmark sites representative of upland rice systems with contrasting edaphoclimatic conditions, Becker and Johnson (1998) identified S. guianensis (accession CIAT 184 according to M. Becker, pers. comm. February 2022) as a promising soil-improving and weed-suppressing fallow legume species, particularly for acid infertile forest and savanna soils.In feeding studies with rabbits, sun-dried hay of CIAT 184 was successfully included as fiber source up to 30 % in feed pellets (Kouadio et al. 2021;2022).DR Congo.Bulakali et al. (2013) tested the seedproduction potential of 3 S. guianensis accessions at 2 sites on the Batéké plateau and found that, by the soilsieving method, 600 kg seed/ha could be harvested from CIAT 184, at a total production cost of USD 1.31/kg.Guinea.CIAT 184 was introduced as part of RABAOC trials and was reported as useful for improvement of natural pastures in Guinea (Elbasha et al. 1999).
Madagascar.In Madagascar, the use of CIAT 184 (and other legumes and grasses) as soil cover in notill agriculture ('conservation agriculture') is being promoted in a manual including a species factsheet on S. guianensis (Husson et al. 2008) and a chapter containing description and discussion of S. guianensisbased conservation agriculture systems (Husson et al. 2013).The manual´s explicit recommendation of CIAT 184 is because of anthracnose tolerance.
Examining the relationship between biomass removal for forage and the mulch cover remaining for a subsequent no-till rice crop, a study on 91 farmers´ fields using CIAT 184 and 3 other legumes, showed the interdependence between cover species used, soil cover rate and mulch quantity and quality (Naudin et al. 2011).In a no-till experiment with upland rice, soil cover by CIAT 184 mulch as high as 95 % was necessary to effectively control emerging weeds (Naudin et al. 2012).In another conservation agriculture study, Randrianjafizanaka et al. (2018) found that of 4 cover legumes, CIAT 184 provided the most effective control of the parasitic weed, Striga asiatica.Using the natural abundance method, Zemek et al. (2018) measured N fixation in CIAT 184 shoots over 17 months ranging from 96 to 122 kg N/ha.
In a feeding trial, Razafinarivo et al. (2014) showed the potential of CIAT 184 in a mixture with Urochloa brizantha for feeding dairy cows.From another feeding trial with dairy cows, Rakotomanana (2021) concluded that, although not increasing milk yields, stylo CIAT 184 silage (in the publication erroneously referred to as 'CIAT 194') contributed to improved physical condition of dairy cows in the dry season.
In 2021, a facility to produce CIAT 184 feed pellets was established at Antananarivo as part of Madagascar-China cooperation.
Nigeria.As Nigeria hosted the Subhumid Research Site and Humid Research Site of ILRI, a considerable portion of the research on Stylosanthes in West Africa during the 1980s and 1990s was done in this country.Several evaluation studies in the dry-subhumid Northern Guinea Savanna ecozone were done by Tarawali et al. (1994), who reported that ILRI 164 (CIAT 184) was outstanding in terms of anthracnose tolerance, forage production and persistence.This was complemented by the highest grain yield of a subsequent maize crop, probably due to the accession´s high dry matter production (Tarawali 1994).However, evaluating it in a fodder bank situation, Peters et al. (1994) recorded that ILRI 164 competed poorly with invading native vegetation.In a relative-palatability test under grazing, Peters et al. (2000) reported high palatability of ILRI 164 throughout the year, possibly associated with the greenness and leaf retention of the accession well into the dry season.Seed yields recorded in that area across 3 years ranged between 57 and 106 kg/ ha/year (Kachelriess and Tarawali 1994).
In a small-plot grazing experiment conducted in the more humid Derived Savanna ecozone, Tarawali et al. (1999) showed that ILRI 164, in association with native Megathyrsus maximus, was particularly valuable in the dry season when it led to higher liveweight gains of grazing calves than grass alone plus rice bran supplement.In the same environment and working with mixtures of legume species under grazing, Peters et al. (2001) reported that ILRI 164 had potential to contribute to stable and productive mixtures.In the same environment, assessing species mixtures primarily used as cover crops and improved fallow, Tarawali (2000) showed that ILRI 164 was a suitable component in several 3-species mixtures.
In that same Derived Savanna zone of subhumid southwest Nigeria, Muhr et al. (1999a;1999b) conducted several experiments to test and analyze the short-term ley potential of 11 legume species for crop-livestock systems.They reported that ILRI 164 had the highest DM yields and that, in spite of dry-season utilization of the forage, grain yields of a subsequent corn crop exceeded those after natural fallow by up to 147 %, equivalent to 96 kg fertilizer N/ha.In subsequent farmer-controlled trials, ILRI 164 ranked highest under the prevailing low-input management (Muhr et al. 2001).

Southeast Asia
In tropical Asia, network research in the 1990s on forages, including S. guianensis CIAT 184 ('Stylo 184'), was conducted mainly within several regional CIAT-CSIRO-ADB R&D projects, in partnership with Review of Stylosanthes guianensis CIAT 184 national institutions in up to 8 countries (Cambodia, China, Indonesia, Lao PDR, Malaysia, Philippines, Thailand and Vietnam).There are several reports and papers informing about the advances and research results of these projects with contributions from individual countries (Stür et al. 1995(Stür et al. , 2000a(Stür et al. , 2002;;Stür 1997Stür , 1998)).All documents indicate that, based on multilocational and on-farm trials, of all herbaceous legume germplasm tested, Stylo 184 was consistently among the top legumes regarding adaptation and productivity in a wide range of ecophysical environments, as well as acceptance by farmers.It was ultimately one of the 2 legumes (the other was Centrosema molle CIAT 15160 named 'Barinas'), that participants in the Forage Seeds Project recommended in 1994 for release throughout SE Asian countries (Stür et al. 1995).In a final, developmentoriented overview, Stür et al. (2007) reported that, of all legume species and accessions tested during the 13 combined project years, the only legume ultimately used by farmers was Stylo 184.The primary use is as cut-andcarry livestock feed.
The most noteworthy country-specific results concerning CIAT 184 include: Cambodia.Based on a 4-yr R&D project with largeruminant-keeping smallholder farmer participation in 6 villages in southern Cambodia, including comprehensive forage and livestock production measurements, Bush et al. (2014) found that growing introduced forage species (4 grasses and Stylo 184) was a credible entry point for converting livestock keepers into livestock producers.From a review of forage options to sustainably intensify smallholder farming systems on sandy soils in Cambodia (and southern Laos), Philp et al. (2019) concluded that Stylo 184 was the only forage legume to be recommended for drought-prone areas with acid, sandy soils.At the livestock nutrition level, Pen et al. (2013) reported that supplementation of a basal rice straw-grass diet with 30 % of CIAT 184 significantly increased forage intake and N balance of zebu cattle.
Indonesia.In a 3-yr multilocational trial (6 sites in East, Central and South Kalimantan), CIAT 184 ('Pucallpa'), along with C. molle CIAT 15160, stood out as the only well-adapted and productive accessions among a total of 35 legume species and accessions tested (Cameron et al. 1995).Liu Guodao et al. (1997) reported that, in East Kalimantan, Stylo 184 is used to rehabilitate degraded Imperata cylindrica uplands.In eastern Indonesia, after 6 years of participatory research supported by the Australian Centre for International Research (ACIAR), the resulting manual recommends CIAT 184 (plus the Australian composite S. guianensis hybrid cultivars, 'Nina' and 'Temprano') for use in East Nusa Tenggara province, mainly as a component of fodder banks (Nulik et al. 2013).
Lao PDR.Following a first report by Phengsavanh (1999), Phengsavanh and Phimphachanhvongsod (2007) summarized that, of 118 legume species and accessions tested during several years at 5 sites in upland areas across the country, Stylo 184 showed the broadest adaptation, including dry season tolerance, to different environments; it was ultimately the only legume selected by farmers for use to feed not only ruminants but also pigs.The benefits of the latter use were reduced time for collecting and cooking natural feed and increased growth rates of pigs.
The potential and adoption of Stylo 184 for smallholder pig feeding in Laos were subsequently summarized by Phengsavanh et al. (2008) and Stür et al. (2010).Individual studies were concerned with the role of Stylo 184 for: replacement of rice bran (Keoboualapheth and Mikled 2003); combination with cassava leaves as protein supplements to a basal diet of broken rice (Norachack et al. 2004); mixture with cassava foliage leading to increased DM intake and N retention in comparison with a basal diet (Khoutsavang et al. 2006); use by smallholder farmers in the uplands as locally grown protein supplement (Kopinski et al. 2008); ensiling in mixture with taro leaves (Kaensombath and Lindberg 2013); and replacement of soybean protein (Kaensombath et al. 2013).They were complemented with combined agronomic/nutritive value studies (Kaensombath and Frankow-Lindberg 2012;Phengsavanh and Frankow-Lindberg 2013).
Regarding the potential of Stylo 184 as forage for goats, Keopaseuht et al. (2004) found that forage intake and digestibility were unexpectedly reduced when leaf only was offered rather than branches with intact leaves, while Phengsavanh and Ledin (2003) reported that the inclusion of Stylo 184 in a grass diet (Andropogon gayanus) improved the quality of the diet, resulting in higher intake and better growth rates of animals.Stylo 184 as basal diet for growing rabbits was not successful (Phimmmasan et al. 2004), an observation that was at variance with reports from Côte d´Ivoire (Kouadio et al. 2021;2022).
In northern Laos, Saito et al. (2006) showed the potential of Stylo 184 as improved fallow/relay crop for increased subsequent upland rice production of 0.6 t/ha with 60 % reduced weed population in comparison with natural fallow.
Malaysia.Assessing the production potential of 47 legume accessions, reportedly being shade-tolerant, as weed-controlling soil cover and feed for sheep grazing under young rubber trees, Ng et al. (1997) found that CIAT 184 along with S. scabra cultivar 'Seca' presented consistently highest DM yields in 2 experiments with photosynthetically active radiation means of 40 and 70 %, respectively.Chen et al. (1995) reported seed yields, obtained in Northwest Peninsular Malaysia, in the range of 230-390 kg/ha.
Myanmar.Stylo 184 was included in evaluation of 5 herbaceous legumes in a drought-prone environment without fertilizer application and irrigation (Gyue et al. 2021).The authors considered Stylo 184 to be a "good" forage because of its DM yield, high crude protein concentration and gas production and low fiber concentration.
Philippines.Stylo 184 showed excellent adaptation to all environments where it was evaluated, particularly on farms with acid soils (Lanting et al. 1995).Horne et al. (1997) mentioned that in Mindanao it is used to control soil erosion and suppress growth of Imperata cylindrica, when establishing forestry plantations.In northern Luzon at about latitude 18° N, Pardinez et al. (2000) reported seed yields from 3 sites and 2 years averaging 201 kg/ ha/ yr.Based on a feeding trial with sheep, Lanting et al. (2003) showed the potential of Stylo 184 as a protein supplement in low-quality (rice straw) basal diets.
Thailand.In 1996, the anthracnose-susceptible S. guianensis cultivar 'Graham' was replaced by disease-resistant Stylo 184 (Phaikaew and Hare 2005), which was released by the Department of Livestock Development in 1997 as 'Ta Phra' stylo.Research on CIAT 184 concentrated on the northeastern region of the country characterized by infertile, sandy soils.For southern Thailand, Satjipanon et al. (1995) mentioned an adaptation study conducted in Narathiwat, where CIAT 184 showed promise on an acid, infertile soil.In a subsequent cutting management trial at Narathiwat, 50-day cutting frequency and 20-cm cutting height gave the best result for CIAT 184 yield and nutritive value combined (Sukkasem et al. 2003).
In the northeast, Stylo 184 was shown to have potential for intercropping with cassava (Kiyothong and Wanapat 2004a).In a dairy cow feeding trial, supplementing cassava leaf hay with Stylo 184 hay reduced concentrate use and resulted in improved milk yield and quality, and ultimately higher economic returns (Kiyothong and Wanapat 2004b).Stylo 184 can also successfully be used as silage, alone or in mixture with guinea grass (Bureenok et al. 2016).Fermentation quality and nutritive value of the legume could be significantly improved by inoculation with thermotolerant lactic acid bacteria (LAB) isolated from fermented juice of tropical forage crops (Pitiwittayakul et al. 2021).From experiments on inter-row planting of legumes to improve crude protein concentration in pastures of Paspalum atratum cultivar 'Ubon', Hare et al. (2004) concluded that Stylosanthes species, including 'Ta Phra' stylo, are suitable legumes on infertile upland soils of northeast Thailand.Homma et al. (2008) found in farmers´ field trials that stylo as relay crop, used in the dry season for livestock feeding, did not affect subsequent rainfed-rice yields.
Optimization of 'Ta Phra' stylo seed production and quality was the subject of several studies (Kiyothong et al. 2005a;2005b).While seed yields of about 1,200 kg/ ha were obtained in one study when the nylon-gauze-bag method was used (Kiyothong et al. 2002), Phaikaew et al. ( 2004) reported a range of 300-1,400 kg seed/ ha, depending on cutting management, and Hare et al. (2007) reported 365 kg/ha with a 1,000-seed weight of 2.59 g (equivalent to almost 390,000 seeds/kg).
Na Chiangmai et al. (2013) concluded from a field study that 'Ta Phra' stylo was not shade-tolerant.This finding is at variance with the report from Malaysia (Ng et al. 1997).
At the laboratory level, a study on regenerative capacity of Stylo 184 showed that in-vitro culture of this accession is a suitable tool for shoot generation, even after 3 years maintenance of the culture (Veraplakorn et al. 2012).Both the activity of antioxidative enzymes in in-vitro shoots and ion accumulation in seedlings can be used to distinguish between Stylo 184 clones in regard to their salt tolerance (Veraplakorn et al. 2013a;2013b).
Vietnam.Khanh and Ha (2007) reported that, as a result of regional evaluations of 70 accessions for adaptation to the environmental conditions of the Dak Lak Central Highlands, Stylo 184 was the only herbaceous legume ultimately selected by upland farmers for cattle feeding.Thang et al. (2010) found that Stylo 184 foliage, alone or in mixture with cassava leaves, improved DM intake, digestibility and liveweight gain of cattle fed a basal diet of urea-treated rice straw.As a result of on-station and onfarm research in South Central Coastal Vietnam, where several grasses and Stylo 184 were evaluated on sandy soils, farmers (n=45) preferred high-yielding grasses (onstation DM yields up to 50 t/ha/yr) to the only legume tested (CIAT 184; up to 17 t DM/ha/yr) (Ba et al. 2014).

China
The importance attributed to S. guianensis in (sub) tropical China is reflected by a slogan coined in 2011 by the former Minister of Agriculture of the People´s Republic of China, He Kang: Alfalfa in the North and stylo in the South.It is also reflected by a large number of studies, mainly on CIAT 184 and cultivars developed from this accession, published in Chinese language.While those studies deserve being categorized and compiled in a separate review or annotated bibliography, in the present document we concentrate on Chinese research published in English and accessible in international journals and databases.
CIAT 184 was introduced into tropical China in 1982 and within a few years had become an important forage legume (He Chaozu and Schultze-Kraft 1988).Thereafter, its continuing good performance and anthracnose tolerance are dealt with in a number of reports on forage R&D (Liu Guodao and Kerridge 1997; Liu Guodao et al. 1997Guodao et al. , 2004a;;Liu Guodao and Chakraborty 2005;Yi Kexian et al. 2007).It is particularly noteworthy that 40 years after the introduction of CIAT 184, its field resistance to anthracnose seems to have remained unchanged.This suggests an inherent tolerance to the prevailing anthracnose strains, in contrast with most of the Australian S. guianensis cultivars ('Endeavour', 'Schofield' and 'Cook'), which succumbed to the disease a few years after their introduction in China.
Released as cultivar 'Reyan No. 2' in 1991, CIAT 184 developed into being the most widespread Stylosanthes cultivar in southern China.'Reyan No. 2' is used for forage for livestock (cut-and-carry and grazing, hay and silage), leaf meal production (for ruminants and monogastrics), soil cover and green manure in perennial crops (rubber, coconut and fruit tree plantations), and soil stabilization and reclamation.

Genetic diversity and new cultivars
Cold tolerance, higher seed production and durable anthracnose resistance are breeding and selection objectives for new germplasm, alongside improved yield and forage quality.Bai Changjun et al. (2004a) reported ongoing evaluations of new material in this regard, but at that time, CIAT 184 (cultivars 'Reyan No. 2' and 'Reyan No. 5') was still the first option.Eighteen years later the situation continues the same, in spite of a report of anthracnose on 'Reyan No. 2' in Hainan in 2014(Jia Yanxing et al. 2017) and insufficient winter survival, in comparison with the grasses tested, reported for 'Reyan No. 5' (and a multiline S. guianensis hybrid cultivar, 'Ubon stylo') in Guangxi (He Chengxin et al. 2017).
Seed production and plant propagation.Details of establishment of a Stylo 184 crop for seed production, its management, seed harvesting and subsequent processing were presented by Liu Guodao et al. (1997); yields reported were in the range of 75-520 kg seed/ ha.Yuan Xuejun et al. (2011) successfully developed a micropropagation method for 'Reyan No. 2'.
Forage.Stylo leaf meal has become an important commercial commodity in southern China (Liu Guodao et al. 2004b) and there are detailed recommendations for the proportion of the legume in rations for different livestock species (Bai Changjun et al. 2004b).Yi Kexian et al. (2007) presented amino acid concentrations in Stylo 184, which for most amino acids were higher than those in alfalfa.Li Mao et al. (2014) showed that 'Reyan No. 2', along with another CATAS germplasm (TPRC-90028), had the highest relative feed value among 10 S. guianensis accessions.
A number of studies dealt with issues related to making silage from S. guianensis, in all cases using cultivar 'Reyan No. 2'.Liu Qinhua et al. (2011) assessed the effects of wilting and storage temperature on silage quality and aerobic stability and Liu Qinhua et al. (2012) tested the effectiveness of 2 LAB strains and concluded that stylo was hard to ensile due to low water-soluble carbohydrates (WSC) concentration and high buffering capacity.Li Dongxia et al. (2019) identified Lactobacillus and Enterobacter as the genera most probably responsible for the enhanced acetate fermentation in low-DM stylo.Li Mao et al. (2017;2021) showed that silage fermentation and ruminal degradation could be enhanced by combining LAB inoculation and treatment with cellulase enzyme and by adding melatonine, respectively.Zhang Yage et al. (2018) assessed the best legume proportion in mixtures of stylo with king grass (Cenchrus hybrid); He Liwen et al. (2019;2020) reported on the potential of gallic acid and Moringa oleifera leaves, respectively, to improve fermentation of stylo silage and Zi Xuejuan et al. (2022) showed the effects of WSC on the bacterial community.
Non-forage uses.Yi Kexian et al. (2007), citing publications in Chinese, reported on improved soil characteristics after 2 years of stylo cropping and enhanced orange tree growth after 2 years, when stylo was applied as green manure.Long Huying et al. (2017), working on degraded soils in the dry-hot valley area of Yuanmou country (almost 26° N), Yunnan province, showed that 'Reyan No. 2' as soil cover and green manure significantly improved soil fertility and micobiological activity.
Wang Dongmei et al. (2008) found that transgenic 'Reyan No. 2' has potential as a plant-based vaccine against foot-and-mouth disease.

Characteristics of CIAT 184: Positive attributes and shortcomings
There are a considerable number of positive attributes (strengths) that characterize CIAT 184, and several weaknesses:

Strengths
• Tolerance of anthracnose • Broad environmental adaptation, including to subtropical conditions • High productivity (up to 22.5 t DM/ha/yr) (Figure 2) • Adaptation to acid, infertile soils with high levels of Al • Low P demand, but responsive to P fertilizer application • N-fixing capability and rhizobium promiscuity • Ease of establishment • Tolerance of drought • High crude protein concentration • Absence of antinutritive factors • Suitability for grazing and cut-and-carry (Figure 3) • Suitability for ruminants and some monogastrics (Figure 4) • Suitability for hay making and subsequent leaf meal/ pellet production, and ensiling (Figure 5) • Suitability for intercropping (Figure 6) • Multi-use potential: feed, soil cover (weed and erosion control), soil improvement (green manure) (Figures 7 and 8) While most of these attributes are also valid for many other S. guianensis genotypes, the outstanding feature of CIAT 184 is its high tolerance of anthracnose in tropical Africa, SE Asia, southern China, some subtropical regions of Latin America (Argentina, Cuba) and the humid tropics of Latin America.As indicated by Chakraborty (2004b), this appears to be related to the physiological races of the fungus prevailing outside the center of diversity of S. guianensis, which is the New World tropics.We are unaware of any report that the field resistance of CIAT 184, which in China has lasted since its introduction in 1982, has broken down to a major extent at any location.
In Table 1, an account of the releases of cultivars derived from CIAT 184 is presented.No formal release (except for Thailand).

Weaknesses
Insufficient persistence beyond 2-3 years under grazing or cutting is the major shortcoming of CIAT 184 and of S. guianensis in general (Cook et al. 2020).This is the most frequently mentioned technical factor responsible for constraining more widespread use of the accession.In this context it has to be pointed out that 'persistence' is the result of a number of inherent plant characteristics, ranging from the location of growing points on the plant to the concentration of palatabilityaffecting compounds in the forage, as well as utilization management.Low seed production potential and lack of ease of seed production are also mentioned as shortcomings, in comparison with other Stylosanthes germplasm, particularly S. hamata ('Verano').

Adoption of and perspectives for CIAT 184
We do not intend to engage in a discussion about the (lack of) adoption of new tropical forage technologies in general nor of forage legume technologies in particular.This has been most ably handled in a number of publications by others including Stür et al. (2000b) for tropical forages in general, and Shelton et al. (2005) for forage legumes.Rather, as a conclusion of this review, we offer a few thoughts of technical rather than socioeconomic nature around actual adoption of CIAT 184 and possible future developments.There is no doubt that some of them are also applicable to other tropical legume species and genotypes that have shown promise in R&D projects.

Lack of data
There is a lack of data on the actual use of CIAT 184 to assess the extent of its adoption and impact.Information is required about both the area planted to the legume and the number of adopters.In this context it is regrettable

Review of Stylosanthes guianensis CIAT 184
that most R&D projects make no provision for resources and institutional commitments regarding follow-up assessments of technology impacts, in the years after project termination.We consider that ILRI´s ex-post economic impact assessment of the fodder bank project in West Africa (Elbasha et al. 1999) is an example to be followed.Moreover, such assessments should be complemented with the identification and analysis of eventual constraints to adoption and, at the socioeconomic level, impacts on the livelihood of adopters' households.

Regional adoption
Tropical America.Traditionally, neither small-nor largescale farmers have been accustomed to sowing forages and assuring their proper management in general, less so regarding legumes.In many cases livestock producers are still more animal-oriented ranchers than agriculturalists dealing with pasture plant production.Taking into account that appropriate grazing management for legume persistence is a major issue for pastures in the humid tropics, where CIAT 184 has shown to have potential in tropical America, it could be questioned whether CIAT 184 has any role in current production systems.Miles and Lascano (1997) found that there had been no major adoption in the humid tropics of Latin America, but this might change depending on evolving land use policies and production systems as well as economic pressures, such as the need to replace commercial concentrates.
Tropical Africa.As can be deduced from the abovementioned impact study of fodder bank technology (Elbasha et al. 1999), there is potential for adoption of CIAT 184 in West Africa.Unfortunately, the accession was not available when the fodder bank research by ILRI and its partners started in the early 1980s.It is regrettable that the RABAOC project was discontinued without considering follow-up activities based on the multisite identification of promising germplasm, CIAT 184 being consistently one of the top-ranking herbaceous legumes.
In tropical Africa, there are 2 different production environments for future use of CIAT 184: in West Africa, where there is a continuing trend for transhumant livestock keepers to become sedentary crop-livestock producers; and in East Africa, where Napier grass dominates planted forages for livestock nutrition.
Southeast Asia.It appears that CIAT 184 has found its niche in this region.However, long-term adoption has been limited.We suggest the initial successes were due to the prevailing smallholder production systems operating under the pressure of very limited land resources, in combination with the long-term involvement of what Shelton et al. (2005) called 'champions', and intensive interaction between researchers and farmers (farmer participatory research).Stylo 184 has been successful where R&D projects arranged seed supply but, once projects were completed, the areas of stylo started to decline over time.A recent ACIAR study evaluating critical factors that promote or impede demand-driven uptake of forages in Cambodia, Lao PDR and Vietnam identified poor access to forage seeds as a major factor that impedes uptake of forages in the region (Yadav et al. 2022).This is particularly important for short-lived or weakly perennial forage legumes like Stylo 184 that require some regular replanting in cut-and-carry systems.
China.No doubt the sustained use of CIAT 184 in southern China is due to the outstanding long-term performance of this germplasm and to strong institutional interest and promotion.Adoption has been on both smallholder farms (where the innovative use as leaf meal originated) and in larger holdings, at private as well as state level.However, continuing use of the accession appears to be increasingly influenced by limited availability of seed produced at reasonably low cost, particularly in terms of labor.In this context, lack of machinery suitable for harvesting both seed and forage in hilly areas is seen as a specific limitation to its promotion.

Access to seed
The experience from R&D projects across the tropics has shown that availability of and access to seed constrain the use of CIAT 184, in spite of the technical strengths of this legume.Most likely this impediment to wider adoption also holds for other legumes.It is an issue that was recognized decades ago (Ferguson and Sauma 1993) and we suggest that it requires particular attention by tropical forage R&D policy makers.No doubt, because of eventual seed market implications, regional differences regarding production systems (e.g.smallholders vs. large estates) need to be taken into account.Model examples in tropical America and Southeast Asia are of seed being produced by smallholders in Bolivia and Thailand, respectively, and subsequently commercialized by seed trading companies (Sauma and Pizarro 2007;Hare 2014).

Future perspectives
We see increased opportunities for CIAT 184 (and other Stylosanthes species and ecotypes), when: • R&D resumes a major interest in developing low-input pasture and forage technologies for infertile soils; • Farmers perceive that soil loss and declining soil fertility are becoming increasing constraints to agricultural production; • The economic environment favors the replacement, or complementation, of synthetic nitrogen fertilizer with the usage of nitrogen-fixing legumes; • There is a major recognition and promotion of the multipurpose potential of forage legumes (Schultze-Kraft and Peters 1997).In this context, CIAT 184 in China is an interesting example: A legume, originally selected for extensive South American pastures grazed by cattle, resulted in being used mainly as a soil cover crop and for the production of concentrate to feed pigs and poultry.This points, again, to regional and production system differences regarding the potential role of S. guianensis.

Research suggestions
In terms of future research on CIAT 184, we suggest the following: Adoption: Region-specific identification of constraints to major adoption in future R&D activities.
Anthracnose: In spite of the continuing field resistance of CIAT 184, breeding towards a broader and durable genetic resistance, owing to the potential danger of new physiological races of the fungus appearing.This includes the strategy of developing new cultivars derived from CIAT 184, which will have to be compared with more recent composite hybrid cultivars such as 'Nina', 'Ubon' and 'BRS Bela' (Grof et al. 2001;Cook et al. 2020).
Forage quality: To optimize use of CIAT 184 for leaf meal/pellet production, breeding for higher nutritive value (improved leaf:stem ratio, lower fiber concentration).
Seed production: Breeding for increased seed yields, while maintaining herbage production.
Rhizobiology: Taking into account (1) that symbiotic nitrogen fixation is among the economically most important attributes of tropical forage legumes, and (2) that seed inoculation with improved Rhizobium strains may well have a place in intensified production systems, screening and selecting more effective and competitive Rhizobium strains for CIAT 184.
Environmental adaptation: Breeding for better adaptation to low temperatures and flooding to extend the environmental conditions where CIAT 184 can be grown successfully.

Figure 3 .
Figure 3. Stylo CIAT 184.A: grazing of a protein bank by dairy cattle at Pinar del Río, Cuba, Sep 2005 (Photograph by Rolando Núñez, Galeca); B: intensively cut plots at Luang Prabang, Laos, Sep 2001; C: in a mixture with grass for cattle feeding at Kampong Cham, Cambodia, Feb 2009.(Photographs B and C by Werner Stür, CIAT).

Table 1 .
Releases of cultivars derived from S. guianensis CIAT 184.