Living Fences

Live Fence Posts in Costa Rica: A Compilation of the Farmer's Beliefs and Technologies Gerard? Budow,ski Ricardo 0. Russo

ABSTRACT. Live fence posts are widely used in Costa Rica and other Cenhal American countries as a sustainable agricultural practice. The Costa Rican farmers' empirical knowledge was compiled through a questionnaire, field measurements and literature review. Ninety two species used for live fences were recorded and tabulated. The management practices (preparation of stakes,planting, attaching wire, pruning regimes) and the various uses are discussed. The biomass production of a kilometer of one live fence was measured. It is concluded that this is indeed a very promising sustainable practice which deserves more research and dissemination. INTRODUCTION

The construction and maintenance of agricultural fencing has the potential to become sustainable through the use of renewable resources. In Costa Rica, as well as in other tropical American countries, many fences or enclosures are constructed by stringing barbed wire on trees serving as fence posts. According to the diversity of climates and the corresponding life zones found in Costa Rica, a great variety of tree species are used for live fence posts (see Photo 1). They are usually propagated from large - -

-

Gerafdo Budowski is affiiated with the University for Peace, Program fm Natural Resources and Quality of Life, San Jos6, Costa Rica Ricardo 0.Russo is affiliated with Yale University, School of Forestry and Environmental Studies, New Haven, CT. Journal of Sustainable Agriculture, Vol. 3(2) 1993 O 1993 by The Haworth Press, Inc. All rights reserved.

65

Research, Reviews, Practices ond Tech~tology

67

cuttings or stakes (about 2-2.50 m). Each species has its own particular growth characteristics, cultivation requirements, management practices, derived products, and benefits. These trees may provide wood for fuel and charcoal or for construction (poles, posts, pillars, etc.), edible fruits and flowers, flowers for honey, leaf forage for cattle and other domestic mimals (e.g., goats, rabbits, and chickens), handicraft (seeds used for beads, ornamental wood), medicinal products, gums and resins, dyes as well as various other products. These plants also have great value as ornamentals and as a refuge for wild animals, notably birds. Additionally, they continually yield new cuttings for more fence posts. However, their main purpose is to serve as physical support to attach rows (usually 3, but their number may vary from 2 to 5) of barbed wire so as to effectively protect houses, crops, cattle and pastures or differenttypes of gardens. Furthermore,living fences provide conspicuous and fm boundaries to separate properties or areas within properties or pastures. When cuttings are planted very closely and sometimes associated with smaller plants that may be spiny (e.g., Bromelia pinguin), or poisonous (e.g., Euphorbia cotinifolia and Hura crepitans), they can also provide an effective barrier to cattle and people, even without use of barbed wire. Many other benefits and services are also provided. The leaves shed by live fences serve as mulch and release nutrients to adjacent crops. The periodic pollarding (cutting back the crown of the fence posts) results in starving and death of roots, leading to small air channels in the soil which favors water infiltration. Other benefits include nitrogen fixation by some species, specially legumes, erosion control and better infiltration of water, provision of shade, use as wind-breaks, niches for insecteating buds, support for orchids and ornamental plants and many other additional uses. A fair statement is that they have considerableaesthetic value. Whether as part of the landscape or from the beauty of their flowers and branching forms, live fence rows provide a pleasant break to the monotony of many grass-dominated landscapes. Many farmers, ranchers and rural laborers are experienced with the use of live-fence posts; some have well defined criteria concerning planting practices and the advantages and disadvantages of various species. But their knowledge is empirical and with the exceptionof a few species, such as Gliricidia sepium (Baggio, 1982; Beliard, 1984), this has not been quantified. The purpose of the present investigation is to overview present technologies and beliefs concerning live fence posts used in Costa Rica and to support obse~atiofl~ with farmer interviews, some field measurements, a literature review, and other experiences by the authors over the last 10 years.

68

JOURNAL OF SUSTAINABLE AGRICULTURE

TRADITIONAL USES OF LIVE FENCES

Live fence posts have been recorded in the literature by many authors and travelers. While barrier and ornamental hedges had been employed traditionally by native people and by the Spaniards after the conquest of America, references are scanty, and the practice may not have been as common as it is today. It was not until the 19th Century that the use of living trees as fence posts came to be recognized particularly for use as a support to attach barbed wire. Early and continuing mention of some of the various species used is found in some treatises of economic plants, propagation and gardening books, and forestry manuals (Holdridge, 1970; Logsdon, 1978; Mintz, 1962). Most references on live fence posts species are found in some national or regional floras on the notes concerning utilization,or as short descriptivenotes inmanuals and books on ornarnental or commercial trees in the tropical American region (Duke, 1972) as well as in other regions of the world (Bond, 1944, Standley and Steyermark, 1946; Howes, 1946; Martinez, 1959). Several of the most common species such as Gliricidia sepium, Bursera simaruba and Spondiaspurpurea are now widely planted throughout the tropics. Moreover, several species of genera such as Erythrina are native to both the old and the new world where they are sometimes widely planted and the knowledge is often transferable. Gliricidia has been introduced to many countries, notably in South East Asia and lately in Africa where it is commonly planted for various purposes, although less frequently than in tropical America. In tropical Africa, for instance in Nigeria, it has become very popular as a live support for yams (Dioscorea spp.), but it has also other uses (Bond, 1944). Sumberg (1986) cites 143 references for Gliricidia sepium, some of them annotated, while at the Tropical Agricultural Center for Research and Training (CATIE) in Tunialba, Costa Rica, a bibliography concerning the genus Erythrina, widely used for fences has been compiled with about 500 references (Centro Agron6mico Tropical de hvestigaci6n y Enseiianza, 1986). Howes (1946), in one of the first reports specifically concerning live fences in the tropics, described and enumerated a variety of trees, shrubs and cacti used in or as fences for a diversity of purposes in Africa, India, Australia and New Zealand. Crane (1945) reported informal trials to select the most appropriate species for live fence posts. Specifically mentioned are Gliricidia sepium, Erythrina berteroana, E. poeppigiana, Bursera simaruba and others used at that time in Cuba. Burgos (1952) made recommendations as to species and practices in the humid lowland forest areas of eastern Peru. Allen (1977) recorded 7 species used for live fence posts in

Research, Reviews, Practices and Technology

69

Southwest Costa Rica making a useful contribution toward elucidating the multiple purposes which live fence posts serve. At Tunialba, Costa Rica, many research theses are related to live fence posts. For instance, de Vastey (1962), included several live fence post species in an exploratory study to determine the possibilities of vegetative propagation of small cuttings using hormones. More related to the subject of fences, Lozano (1962) investigated improved methods for successful rooting of large cuttings from Gliricidia sepim, Erythrina costaricensis and E. poeppigiana. This author analyzed which species were best suited for different conditions, suggested better planting practices and emphasized seasonal timing for planting the cuttings. Budowski (1977) included live fence as a promising line of research for an international research program in agroforestry. Later, the author brought forth the various advantages and drawbacks of live fence posts over usual (dead) wooden posts (Budowski, 1982) and emphasized living fences in tropical America as a widespread agroforestry practice (Budowski, 1987). A more specfic paper on live fence posts used in Costa Rica was prepared by Sauer (1979). He listed 43 species including not only those planted by large cuttings,but also those that have been planted by seed or are allowed to grow from natural regeneration to serve as hedges or fences. He delineated their climatic boundaries and included various economics and historical notes and observations.Sauer made a strong plea to maintain diversity as part of the fence-post hedgerow complex. METHODS An extensive literature review on species used for live fences in Costa

Rica was carried out at Orton Memorial Library, CATIE, Tumalba, Costa Rica. Species reported in at least one reference, were recorded and tabulated by scientificname, common name, family, life zone, uses and propagation procedures when possible. This data was compiled over a three year period. An informal questionnaire was used to int&view farmers who were particularly knowledgeable and experiencedwith living fences in different regions of the country. These interviews involved visits to the farms and conversationswith the fanners. Emphasis was placed on the number and frequency of species used in each region; thus some areas received more intensive scrutiny. Farmers were asked about their experiences, practices and observations concerning the live fence species they used As the investigation progressed, trials on rooting stakes were undertaken (Shchez and Russo, 1985). Furthermore, a 100 m fence of Eryrhrina

70

JOURNAL OF SUSTAINABLE AGRICULTURE

berteroana was pollarded and measured at Turrialba (Budowski, Russo and Mora, 1985).

RESULTS AND DISCUSSION

The results of the literaturereview, complemented by the interviews are summarized in Table 1. It can be observed that some fence species are quite site specific while others seem to be adapted to broad range of climatic zones. Caesalpinia eriostachys Benth., for instance, is planted only in lowland1 regions of relatively low rainfall around 1400-1800mm, with a well marked, 5-6 months dry season, while Drimys winteri fences are used at higher elevations (between 1500-2800 m) on the slopes of the volcanoes, or elsewhere above 1500 m. Moreover, the ubiquitous Cliricidia sepium ranges from low elevation in wet habitats to middle elevation with alternating low and high rainfall and to regimes of relatively dry and hot climate where it often forms natural pure stands. Bursera simaruba is employed as a fence tree at low and medium elevation in dry and wet climates on good to extremely poor soils. In dry habitats it appears to be a native species on rocky outcrops reaching about 6 to 15 m high, as part of the climax deciduous forest of the pacific lowlands. But on-the Atlantic coast. with rainfall usuallv over 2500 mm and sometimes above 4000 mm. it is usually a secondary species and may reach 30 m height or more, rarely losing its leaves throughout the year. he growth of fen& posts: Most live fence posts species, if not pruned and allowed to grow, will eventually revert to the natural tree form. This reversion is evident at many abandoned fences throughout the country. However, most live fences when planted and maintained as fence posts, usually display a quite different form than that of the same species occurring naturally in the forest or in open communities. The crown shape is particularly altered by the effect of periodic pruning. The general appearance along a maintained fence is changed enough to cause confusion when compared with the same species found in the forest in open fields or in orchards. Bursera simaruba for instance, is an occasion&atural component of the upper forest story. However, when planted as a fence tree, its height is reduced to a rather squat hunk of about 2 m in height with a confusion of branches originating from the top. 1. For convenience,here the word lowland applies from sea level to about 600m elevation, middle elevation from 600 to 2000 m and highland further above. The limits of 600 and 1500 represent approximately the lower and higher ranges adequate for commercial coffee cultivation.

Research. Reviews, Practices and Technology

71

The harvesting of stakes for more posts always takes place from the top of the trunk (originally a cutting itself). The top gradually assumes a swollen bulb or headlike appearance (callus). Owing to the propensity of fence mee species to sprout vigorously with many new shoots after pollarding, the upper part of the stake may be much larger than the diameter of the stake itself. This impression may be increased by the abundant growth of epiphytes that find suitable niches between the branches or the stubs of past pruned branches. Cutting back repeatedly promotes extensive callus and woody tissue that progressively build up from wound healing. A typical fence post of Erythrina costaricensis, for example has a trunk top diameter between one and a half and two times that of the trunk. Preparation ofposts: The majority of the important fence forming trees in Costa Rica are propagated by planting large stakes (cuttings), usually around 4-12 cm diameter at the base, and 1.5-2.5 m in length, depending on the purpose of the fence and the preferences of the farmer. Cuttings are derived from branches arising from live fences that are pollarded after one or two years of growth. The number of shoots produced by each established live fence tree varies according to age, vigor and period since the last pollarding. A living post can normally produce from 2 to about 8 cuttings for planting every two years. Cuttings are easy to prepare and require no special skill in planting, although handling with care between cutting and planting is necessary. From early February through early April, mostly when the weather is relatively dry, the trees are pruned to obtain cuttings for planting as well as to remove excess or undesirable growth from the parent tree. Many of the unused branches and leaves form a mulch around the fence. However, cutting has also taken place at other periods during the year. The time-lag between the planting of a cutting and when it begins to produce shoots for new cuttings of plantable size varies among species but it is generally six months to two years. Sometimes it is even less as in the case of the fast growing Erythrina spp. and Gliricidia sepium. Once this production is initiated, sprouting vigor apparently does not decrease for a relatively long time (over 20 years, at least, has been recorded for some well known species with no apparent decline). When in healthy conditions the trees show considerablelongevity. Older trees of Gliricidia sepium can display deformationwhere the barbed wire has been attached. Some fence trees have been estimated to be 90 years old. However, species with relatively soft wood such as Bursera sirnaruba,and Spondias purpurea are reported by local farmers to have a shorter life expectancy than species with denser wood, such as Gliricidia sepium or Diphysa robinioides. The annually repeated pollarding seems to enhance juvenility. The

om mm m m a m m aa 0 m m m m m m m u m m w m m m a w u u u m w W w U U o m m m u m m m m u m u a m m m m m o m m c m c c m w a m B 0 - r l c 01 m . 3 u m ~ ~ 1 m.r l6- 3 c m o B m w . 6 o O W 0 m m u m.6 m z a u a a i o m o m a n m m u 4 w o u o - u m ~ m ~ d u o m m m m w C 3 m 1 U ~ m m u w ~ 6 m m m m m m U ~ u U O m a u ~ d m m u n E w a o o m m a . 4 m o m m r - u m a u u m ~ o u - m m m m + u a u u a o w c m o o c o ~ ~ m m 3 m a m m o a a a s

~

~

C

~

Z

T

~

P

~

~

Z

U

T

O

U

U

T

E

C w 0

? 0 u

A

w

&3V

0

2

0

.+ 3

o

m a

m

c

u

zsaai~,,

4..

w Y

-As 4 4

w O w w

s o d s a a l o 0 O n i o m Do-3.0 o m u Y > r l * m P C s u

rn rn

: - 2m

>

e z

i'

- d - d a o m a d d o o o w o n m.3 U c n m 0 0 N r l E d 4 iu m m a c u 0 . 4 4 0 4 m m m m u

0 0 n o

0 4

c

E g s z z ~ g J ~ ~ X ! ~ ~ ~ L8~8$$~~~';I~~P,P,&~~z P z P z p ~ t ~

&f::Z:g':

& & & & & : & g $ g $ g gkg;l.Xg&ggSg:g

Research, Reviews. Practices and Technology

75

woody shoot regrowth of live fence trees generally appears to have more leaves in proportion to the amount of woody tissue than shoots of naturally growing trees of similar age. They are also said to show higher rooting success than shoots from non pollarded trees. However replicated tests are needed to c o n f i i these differences. Erect, straight, vertical branches are preferred over crooked or lateral branches. Both are found along fences, sometimes side by side. But it is also common to F i d entire rows of straight vertical branches, a possible indication that the same clonal material had been used. Accordine to some farmers, the progeny of crooked cuttings manifests the curving teidencies, making for less even fences. Lateral branches reputedly have lower rooting effectiveness than more vertical branches. Both these beliefs need to be confiied. From the last week of January, throughout February, March and into mid-April, considerable activity along the fence lines takes place, as shown by abundant evidence: pollarded pruned fence rows, leafy material on the ground, stacks of smaller branches piled for fuelwood, cuttings placed under some shade, with lateral branchlets and leaves neatly removed to produce "clean" poles and of course newly set live fence posts. The coincidence of this seasonal silviculturalpractice with the major dry season h u g h o u t Costa Rica is considered quite important. The major factors in the efficacy of this timing seems to be related to the effect of the approaching rains and morpho-physiologic periodicity of the species. In the dryer areas, many cuttings are leafless or have few leaves. Even in the wetter areas, there seem to be less leaves at that time. The practices which are followed are predicated upon the fact that planting should precede the onset of the heavy rains so that root and bud initiation will already have begun when the season starts. The cuttings are prepared before the appearance of buds that precede the vegetative growth flush associated with the rainy season. Flowering cuttings are reputedly not good propagation material. Even in locations with adequate soil moisture it is said that lower percentage rooting success is obtained when cuttings are derived from flowering branches. It is usually considered unadvisable to plant the cuttings during the rainy season itself, but many exceptions exist. The reasons given for no planting in the rainy season are the interruption of active vegetative growth, lack of time for adequate healing of cuttings and subsequent susceptibility to fungal infections due to excessive water and poor soil drainage which limit the production of roots. The seasonal changes in rooting effectiveness may involve carbohydrate mobility, hormonal factors and bud dormancy, all of which need to be investigated.Many farmers

76

JOURNAL OF SUSTAINABLE AGRICULTURE

consider.it important that planting is made at the time of the decreasing moon phase, (waning moon) to ensure successful rooting. Pruning techniques: The pollarding or pruning of the parent tree is a fairly simple task. The woody sprouts are lopped off with a single clean cut using a sharp "machete" or a bush-hook, leaving as small a stub as possible. This practice results in less wound covering tissue, the excess build-up of which is said to negatively influence the vigor and form of the shoots and the health of the parent tree. The shoot growth arises from dormant and adventitious buds, whose c o ~ e c t i o nwith the cambium may be lessened with the build-up of wound tissue. Sprouting may also be inhibited by excessive amounts of healing tissues. The heavy population of epiphytes growing upon the trunk top, particularly in moister zones on regularly pollarded trees of advanced age, often contributes to water retention. Eventually water will penetrate the trunk,creating favorable conditions for the entry of heartrot fungi. This is particularly true for Erythrina berteroana in wet areas. Many landowners practice drastic pollarding by eliminating all shoots while others leave a shoot or two. It is said that the latter practice reduces stress and shock, allows reasonable root growth to continue, promotes rapid healing of wounds and provides better conditions for new shoot growth. Larger cuttings also result, since the period of growth prior to harvest would be two or even more years for the uncut branches. Planting: After pollarding, the cuttings are topped and cleaned of branchlets, taking care to minimize injury from gashing or tearing off the bark. The clean poles are then stacked or set vertically under shade and can be planted the same day or later. Stacking up to one month has been observed. The importance of allowing specific time to elapse between cutting and planting is given high priority by some farmers, but not by others. Variation occurs according to the species and the environmental conditions. For example, when using Spondias purpurea in some drier zones in the Guanacaste province, farmers often stack the cuttings vertically in the shade of a tree for two weeks before planting. There does not seem to be a minimal or maximal time for storage with most species. The longest storage period is observed in dry areas, while excessive storage in humid conditionsis generally avoided because it may result in lateral rooting or fungal growth. The poles are transported by hand, oxen, horses, or vehicle to the planting site, always handling them with great care so as not to cause injuries or impair the rooting ability. Commerce of live fence cuttings is profitable, with prices per goodsized cutting ranging in 1991 from US$O.10 in a pioneer farm and cattle

Research, Reviews, Praclices and Technology

77

region to up to US$0.20 in an intensive coffee growing area. In one area it was observed that one "pruner" charged nothing for tending the fence. As compensation, he got free access to the material pruned and was given the right to sell the usable products of his pruning, that is, the poles, to other farmers wishing to establish new living fences. Usually he had a contract to establish a new fence. Among parents and close friends, it is common practice to provide the posts free. However, it is expected that the person who needs them does all the pruning of the fence where the new cuttings are obtained. Once planted, a reasonable time period from two months up to one year, is allowed to elapse before the wire is attached. The thicker the live post, the sooner it is possible to attach the wire. If a fence must be immediately wired, widely spaced dead wooden posts are often initially used. Dead posts allow for immediate fencing with barbed wires, serving only long enough to allow living posts to become established.They will rot in a short period, usually less than one or more years, depending on natural durability. Planting of cuttings between these dead fence posts is done at a 2-3 m spacing immediately after but without attaching the barbed wire. After a year or less, barbed wire can then be attached to these well rooted living posts now. It is also common to see living fence posts planted without the provisional use of dead posts, but allowance must be made for good rooting before barbed wires can be attached. A very common practice is to "flicken" or increase the density of a fence by systematically planting cuttings from the nearby older live fence posts. These fences with different age posts often look rather heterogeneousin architecture and vigor. The holes for planting are dug to a depth of 30-50 cm with a width of 7-15 cm. The poles usually are shaped with a "clean" diagonal cut on the lower end so to expose more cambial tissue. Sometimes the cutting at the bottom has 2 or 3 diagonal "faces." They are then placed in the holes, held vertically, and made as F i as possible at their base by treading with the heels using full body-weight. All round contact of the soil with the cutting and absence of voids is considered essential for good rooting. Movement or disturbance of the posts during the Fist month or so is carefully avoided. Where cattle are present, it is considered desirable to protect the poles from their reach. Damage from cattle disturbance can be a problem with some species because shoots may appear below the top and thus can be reached for browsing. The weight of cattle can also uproot recently planted stakes. One alternativeis to keep the pastures with new fences free of livestock while the cuttings take firm root, a period that usually lasts 3-6 months. When living fence posts are planted for cattle enclosure, the

78

JOURNAL OF SUSTAINABLE AGRICULTURE

planting of species especially sensitive to disturbance is avoided. Here the planting of longer cuttings with greater diameter is favored, since the growth of new sprouts at the top is out of the reach for cattle. Moreover, the formation of sprouts and branches originatingbelow the top section of the planted cutting, is carefully avoided by systematic pruning. Cattle relish the leaves from the sprouts of some species and can cause considerable damage to the cambium by pulling on these shoots and tearing off considerableportions of the bark. Attaching barbed wire: A basic function of living fence posts is for the attachment of barbed wire to contain livestock, protect crops from human and animal intruders and to demarcate property boundaries. The interaction of the living trees and barbed wire merits discussion. The barbed wire is not attached until after the live post has rooted solidly and has produced normal branches, generally 3-6 months after planting. If the staples that hold the wire are nailed too early, cambial damage can result in necrotic zones. Posts of small diameter show a higher mortality from nailing than thicker ones. Development time, often up to one and a half years, is usually allowed before proceeding with nailing. For most species, six to twelve months are considered adequate. As already mentioned, it is common to plant dead fence posts at relatively wide spacings and attach the wire to them. Live posts are then planted in between and when they take root firmly the wires are attached. After attachment, several species show poor wire holding properties owing to the low density or softness of the wood, or heavy sap exudation. Another problem is that of the tree "swallowing" or engulfing the wire, which becomes progressively embedded in the trunk rather than remaining on the surface of the bark. Gliricidia sepium, Spondias purpurea, various Erythrina spp., and particularly Bursera simaruba, occasionally exhibit h i s tendency. The consequences of engulfment are: (1) accelarated corrosion of the wire in some species, (2) loss of marketable value of the wood, such as in the case of Bombacopsis quinata and Tabebuia spp., and (3) minor but sometimes important issues of questionable property lines between small, intensively cropped vegetable farms in the highlands, caused by movement of the wire as the trees grow. Severalstrategies are used to avoid engulfment. One of them consists of not nailing deeply into the tree. This serves two purposes: the fence staple is not embedded in the heartwood and a space is left between the tree and the staple, both of which provide a margin for wire movement with increase in tree girth. Another practice is to place a small strip of metal sheet between the wire and the tree. This helps to keep the wire on the bark surface. A third practice is to nail the staple through a small piece of wood

Research, Resiews, Practices and Technology

79

as a backing, thus enabling the wire to move above the bark and not to penetrate into it. Other management practices: During the first year, leafy side branches from the trunk are carefully pruned off. This practice serves to direct growth to the top shoots and to train the tree by promoting corollary inhibition of lateral shoot initiation from the trunk. At the same time, any undesirable lateral, weak, or excessive top branches are removed. The gaps between the living fence posts are gradually filled with new cuttings to reach the desired density. Densities range from 1to 3 posts per meter up to 6 posts per meter in very dense fences around coffee and vegetable fields and gardens to prevent animal entry as well as to protect against strong winds. Most common is a density of 2-3 trees per meter. Another variation is the planting of widely spaced, well-rooted posts 4-5 m apart, with living posts planted in between. The latter may or may not be attached with barbed wire. In dry areas the spiny Bromeliapinguin (which has edible fruits) is a common candidate to fill in the lower portion of the fence. Cuttings for reinforcement usually are of smaller diameter than those used for the establishment of the original fence. The cuttings are sometimes tied to the barbed wire to train them during the period preceding nailing of the barbed wire. W~ththe planting of new cuttings to fill in an older fence, other welldeveloped fence trees are sometimes harvested for firewood, or for construction posts and occasionally lumber and cabinet wood This practice allows a gradual replacement and harvesting without depleting the area of support for barbed wire. Shchez and Russo (1985) reported preliminary results of a survey by questionnaires carried out in Costa Rica by the Nitrogen Fixing Trees Project (NFTP), CATIE, Tunialba on uses of Erythrina spp. including management practices. Table 2 details the findings in relation to fence posts. Later, Shchez and Payne (1987) communicated results of another survey done by the NFTP on Gliricidia sepium. The authors found that 95% of the interviewedfarmers use the species as live fence posts. Uses: For most species both the main stem and the cuttings produced by branches provide a much needed supply of firewood,although the latter is considered of lesser quality since insufficient time is allowed to produce heartwood. The quality varies by species, those with the highest density are considered better quality for firewood or charcoal. A self-sufficient supply of fuelwood is much more than just an economic benefit. A nearby supply also relieves the pressure and thus provides a margin of protection for the existent, natural primary or secondary forest nearby, as well as other trees found in nearby fields.

a

JOURNAL OF SUSTAINABLE AGRICULTURE

80

T a b l e 2. Hanagement p r a c t i c e s o f E r y t h r i n a s p p . f o r f e n c e p o s t s I n C o s t a R l c a (Sanchez a n d Russo, 1 9 8 5 ) . SFECIES E. b e r t e r o a n a 2. fUSCd Other Erythrinas.

EXQUENCY 50 % 5 % 45 %

FOW1 OF PROPAGATION EL4TERIAL

96 % Large c u t t i n g s Direct planting by seed 4 %

SilAPS OF THE BAS?& CUT

SRAPE OF THE

Diagonal Tvo f a c e s Three faces Conical Flat

Diagonal ConicaI Flat

10 %

10 % 10 % 19 %

82 % 4 4 14 %

51 %

PERIOD OF ?.EST BEFORZ PLANTING 7 days Up 8 t o 15 days Without r e s t

APICAL CUT

18 % 41 % 41 %

CUTTING -SEASON Dry p e r i o d S t a r t of rainy season During t h e rainy season Throughout t h e y e a r

35 % 22 % 22 % 17 %

MOON ?SASE

Rising Wanin?

0 - 1 m 1.1 - 2 . 0 m 2 . 1 - 3.0 rn

3.1 - 4.0 m Variable

SURVIVAL RATE

XPLANTING

< 49 % 50 69 %

1 0 - 79 '1 8 0 -. 8 9 %

6 % 12 % 23 %

90 -100 %

53 %

-

6

NO Yes

%

.

.

Diphysa robinioides, a handsome tree with numerous yellow flowers during the dry season, is highly valued for its dense and very resistant wood It is in great demand for piling foundations,dead comer fence posts and railroad ties. This tree, similar in appearance to the black locust (Robinia pseudoacacia) of North America, will grow from a cutting of 6-7 cm diameter to a tree with a diameter of 20 cm in 5-6 years. It thrives in dry and wet soils. Even though the trunk is usually not straight, fences have been observed with a sizeable percentage of trees with good form characters. Ample possibilities for genetic improvement exist for this species. GIiricidia sepium, probably the most extensively used live fence tree in

Research, Reviews, Practices and Technology

81

the American tropics, yields a wood of similar qualities but its diameter growth is slower, although it produces cuttings at a faster rate. It has durable wood which is used for cabinets, for artisanal work and for poles in contact with the ground. A workshop on management and improvement of this species was recently held in the Tropical Agricultural Center for Research and Training (CATIE), Turrialba,Costa Rica (Withington et al., 1987). few other fence trees are often allowed to grow into adult trees because of the value of their wood. Among those Bombacopsis quinata, Tabebuia rosea and Tectona grandis stand out in the lowlands. In the highlands, Cupressus Iusitanica and Alnus acuminata, both propagated by seed, are most common. The leaves, flowers and fruits of several live fence trees are an important source of fodder for domestic animals and food for humans. Little research on this subject has been done, especially in the lowland humid tropics. The leaves of Gliricidia sepium are commonly fed to cattle, but are said to be semi-toxic to horses. The flowers, however, are edible when cooked. Spondias purpurea, a cosmopolitan tree, bears red or yellow plum-like edible fruits appreciated by livestock as well as humans. The leaves are also eaten when young. Some varieties, particularly with large fruits are often seen in markets. The tree shows surprising variation in appearance and growth form. There is great variation in the quantity and quality of fruit, but the highest yields and the best quality come from rather dry areas. The names of some particularly productive varieties have been reported from Guatemala (Standley and Steyermark, 1946). Eugenia jumbos bears a crisp fruit, the rose apple and is grown from polyembryonic seeds which produce multiple fast growing shrub-like stems. The tree is often planted for windbreaks to protect coffee plantations and other crops at middle elevations. Erythrina berteroana responds to regular pollardiig with a great volume of large leafed shoots in only 3 weeks to 1 month. These shoots are often used for rabbit feed and are also palatable to livestock. Goats are particularly eager to eat these shoots. Production: At CATIE (elevation 610 m, mean annual temperature 22.3"C and annual rainfall 2700 mm), a 100 m fence of E. berteroana after pruning and carefully measurements showed the following production (Table 3). The posts had an average spacing of 0.6 m and the lopped branches were 4,6, and 8 months old after being pollarded. A member of the Liliaceae, Yucca elephantipes Regel, with long stiff dagger shaped leaves is notable for its large paniculate, whitecreamish, bell-shaped flowers.These are sought out as a delicious vegetable which is

JOURNAL OF SUSTAINABLEAGRICULTURE

82 .

-

-

-

.

..

-

.

Table 3: Biomass production (dry weight) of one kilometer of fence of Erythrina berteroana pruned every 4, 6 and 8 months. Average distance between posts: 0.6 m. Every 6 months

Every 4 months

Every 8 months

(kg d. w. /prunning)

700 500 1200

Leaves Branches Total ..

. -

-

.

780 1050 1830 -

-.

- -.

.

750 2440 3190 - .. .

cooked with eggs. The stigma is usually removed to avoid a bitter taste. Flowers are sold at US$0.50-1.00 her cluster in Costa Rican markets. Each stem produces one cluster per year. It is rapidly propagated from cuttings in most types of soil under dry and wet climatic conditions from sea level to about 1500 m altitude. The plant is also used to stabilize terrain along roads, particularly when there are landslides. Anacardium occidentale, the cashew, will grow on very degraded soils in relatively dry climates. Psidium guajava, the well known guava fruit, is used for jellies, canned for juice or eaten raw, although usually the latter is infested with larvae when ripe. This tree also does well on poor and compacted soils. The fruits like those of various Ficus spp. are relished by cattle ,and other domestic animals and in fact the seeds are spread through excrement. The wood is frequently used for fuel. Bursera simaruba exudes on oleo-resin called elemis. This aromatic essential oil played a role in commerce in the past and provided a sticky base for varnishes, printers, inks and in some ointments where stimulant and antiseptic qualities were required. Home use is made of the gum as a cement for mending crockery and as a crude varnish. The thick colored sap exuded from wounds in Spondias purpurea is used in medicine.

CONCLUSIONS The maintenance and other management practices of living fences vary with the species, the ecological conditions, the local markets and the cultural backgrounds. Living fence posts are of relatively low cost, in comparison with normal wooden "dead" posts, particularly when live fences from which cuttings can be obtained are found nearby. They are also more durable, and rows of live fence posts well over 50 years have been recorded.

Research, Reviews, Practices and Technology

83

But there are also drawbacks when using live fence posts in comparison with non-living wooden, metallic or cement fence posts. They require much more care for their establishment and maintenance, such as careful handling of the large cuttingsprior to planting, and their possible replacement if they do not take root. Periodic pruning is required and often constitutes a main management problem, as can be witnessed by many abandoned live fences. Some of the advantages and drawbacks concerning live fences were previously presented (Budowski, 1982). Live fences may become weeds and grow tall when they are not maintained. They can, and sometimes do, harbor pests. They may also be considered as obstacles to the use of machinery and even for small airplanes that spray pesticides. Under certain conditions, they compete for water, nutrients, and light with adjacent crops, although over limited surfaces. Some of the fruits, and even the fofiage of some species, can be toxic if ingested by livestock. And once planted live fences are difficult to move or to eradicate because of their rooting systems. They are not easily relocated, such as with electric fences or even dead fence posts. Despite the important role of live fence posts in the rural landscape and the vast amount of empirical knowledge accumulated, information is scarce. More needs to be known about the cultural, edaphic and climatic requirements of the many species used and the anatomical and physiological basis of propagation from large cuttings and the various factors affecting their regenerationand rooting. Certain superior varieties among fence post species seem to exist, in terms of growth habits, vigor, desirable architecture, sprouting behavior, absence of spines ( i what would be normally spiny varieties), quality of edible fruits and flowers produced, protein content of foliage and resistance against drought or swampy conditions. Local people speak of varieties and races, but these are tittle known since the information has not yet been validated, much less quantified. The same may be said about the ability of living fence posts to contribute to soil stabilization, to serve as windbreaks or to supply sources of animal feed or forage. For the latter, additional data about their protein content, digestibility, drying requirements or their pelleting capability is needed. There is no doubt that a great potential exists for the development of improved strains through the application of known selection and varietal improvement techniques. Much potential exists for commercialization of some of the multitude of derived products obtained from live fences and the subsequent generation of additional income, particularly for the small farmers. Finally, the living fence is indeed a very promising practice to be

84

JOURNAL OF SUSTAINABLE AGRICULTURE

considered in sustainable agriculture. For that reason it deserves more research and dissemination of knowledge. A first step in this direction would be the retrieval of all the empirical information accumulated, not only in Costa Rica but also in the rest of Central America, the Caribbean, and northern South America, where the practice is most widespread. Undoubtedly, much of the value remains to be found. BIBLIOGRAPHY Alavez Lopez, S. 1983. Estudio p r e l i i a r de 10s c ~ c o vivos s en la ganaderia de Teapa, Tabasco, (manejo de Arboles en 10s potreros, una tdcnica silvopastoril). Tesis 1ng.A~.Chapingo, Mexico, Universidad Aut6noma de Chapingo. 77 p. Allen, P.H. 1977. The rain forest of Golfo Dulce. Stanford, California, Stanford University Press. 417 p. Baggio, A.J. 1982. Establecimiento, manejo y utilizacidn del sistema agroforestal cercos vivos de Gliricidia sepiurn (Jacq.) Steud., en Costa Rica. Tesis Mag.Sc. Turrialba, Costa Rica, UCRCARIE. 91 p. Beer, J. 1987. Experiences with fence lime fodder trees in Costa Rica and Nicarag u a In Advances in Agroforeshy Research Proceedings, Edited by J. Beer, H.W. Fassbender and J. Heuveldop. CATIE. Tunialba, Costa Rica (Inf. Tec. no. 117) Beliard, C.A. 1983. Resultados preliminares de la producci6n de biomasa en cercos vivos de Gliricidia sepium bajo dos frecuencias de poda en la regi6n de La Palmera, San Carlos, Costa Rica Trabajo presentado en el curso Corto sobre Metodologias de Investigaci6n Agoforestal en el Tr6pico Hiunedo UNU/CATIE-IICA TROPICOS-CONIF, Cali, Colombia, 26 de noviembre a 7 de diciembre de 1983. Tunialba, Costa Rica, CATIE. 12 p. Beliard, C.A. 1984. Producci6n de biomasa de Gliricidia Sepium (Jacq) Stend, en cercas v i v a bajo tres frecuencias de podas (3,6 y 9 meses). M.Sc. Thesis. CATIE-UCR, Tunialba, Costa Rica, 97 p. Bond W.E.T. 1944. Hedge plants in Northern Nigeria Tropical Agriculture (Trinidad) 21(12):228-230. Budowski, G. 1977. Ago-forestry in the humid tropics, a program of work. Tunialba, Costa Rica, CATIE, Departamento de Ciencias Forestales. 24 p. (report mimeo). Budowski, G. 1982. The socio-economic effects of forest management on lives of people Living in the area: the case of Central American and some Caribbean countries. In Socio-economic effects and constraints in tropical forest management Edited by E.G. Hallsworth. New York, John Wiley, 1982. pp. 87-102. Budowski, G. 1987. Living fences in tropical America: a widespread agroforesby practice. In Agoforeshy: realities, possibilities and potentials. Edited by H.L. Gholz. Martinus NijhoR Publishers, Dordrecht, The Netherlands. pp.169-178. Budowski G., D.C.L. Kass and R.O. Russo. 1984. Leguminous trees for shade. Pesquisa Agropecuaria Brasileira 19(s/n):205-222.

Research, Reviews, Practices and Technology

85

Budowski, G., R.O. Russo y E. Mora. 1985. Productividadde una cerca viva de Erythrina berteroana Urban en Turrialba, Costa Rica. Turrialba 35(1):83-86. Burgos, J.A. 1952. Postes vivos para cercos. T i g o Marla, P d . Estaci6n Experimental Agrkola 6 p. (Circular no. 39). Centro Agronornico Tropical de Investigaci6ny Enseiianza 1986. Bibliography on the genus Erythrina. CATIE, Renewable Natural Resources Department, Turrialba, Costa Rica. 182 p. Combe, J. y G. Budowski. 1979. Clasificacidnde las tknicas agroforestales: una revisidn de litexatura. In G. de las Salas e d Taller de Sistemas Agoforestales en Am6rica Latina. W a l b a , Costa Rica, CATIE, pp. 1748. Cook, 0.E 1901. Shade in coffee culture. Washington, D.C., U.S. Department of Agriculture, Division of Botany. 79 p. Crane, J.C. 1945. Living fence posts in Cuba Agriculture in the Americas 5(2):34-38.

Duke, J.A. 1972. Isthmian ethnobotanical dictionary. Fulton, Maryland, Duke, 96 p. Espinoza, J.E. 1984. Caracterizaci6n nubitiva de la fracci6n nitrogaadit del fomaje de madero negro Gliricidia xepium y por6 Erythrina poeppigiana. M.Sc. Thesis CATLE-UCR W a l b a , Costa Rica. 90 p. Flinta, CM. 1960. Rdcticas de plantaci6n forestal en Adrica Latina. Roma, FAO. 449 p. Fonseca, M.T. 1968. El por6. Revista de Agriculma (Costa Rica) 40(6-7): 102-112.

Garcia W J . 1974. Emaizamiento de estacas de seis especies forestales con tres niveles de 6cido indol-butirico.Tesis Mag.Sc., Twilaba, Costa Rica IICA 41 p. Halliday, J. and P.L. Nakao. 1982. The symbiotic affities of woody species under consideration as nitrogen-fixing trees. NiffAL Project, University of Hawaii. 85 p. Hardwood, R.R. 1979. Small fann development. Understanding and improving farming systems in the humid tropics, Boulder, Colorado, Westview Press. 160 p. Hartmann, H.T. and D.E. Kester. 1983. Plant Propagation: principles and practices. 4ed Englewood Cliffs, New Jersey, Prentice-Hall.727 p. .Holdridge,L.R. 1966. The life zone system. Adansonia 6(2): 199-203. Holdridge, L.R. 1970. Manual dentrol6gico para 1000 species arb6reas en la Repdblica de Panam4 Rograma de las Naciones Unidas para el Desarrollo y FAO. Panarni 325 p. Holdridge, L.R. y L. Poveda. 1975. kboles de Costa Rica v.1. San Jose, Costa Rica, Centro Cientffico Tropical. 546 p. Howes, F.N. 1946. Fence and barrier plants in waxm climates. Kew Bull of Miscellaneous Inf. 251-87. Krukoff, B.A. 1939. The American species of Erythn'na. Brittonia 3:205-337. Little, E.L., Jr. Common fuelwood crops: A handbook for their identification. Morgantown, West Viginia, w.d, 354 p.

86

JOURNAL OF SUSTAINABLE AGRlCULTURE

Logsdon, G. 1978. Living hedge fences, shrub and trees varieties, planting and pruning. Org. Gard Farming 25(5):78-84. Lozano J., O.R. 1962. Postes vivos para cercos. Tesis Mag.&. Turrialba, Costa Rica, IICA. 77 p. Martin, F.W. y R.M. Ruberte. 1980. Techniques and plants for the tropical subsistence fann. Mayaguez, Puerto Rico, U.S. Department of Agriculture. 56 p. Martinez, M. 1959. Plantas dtiles de la flora mexicana. Mexico, D.F., Botas 621 p. Mintz, S.W. 1962. Living fences in the Fond des-Neges region, Haiti. Econoniic Botany 16:lOl-105. Morton, J.F. 1981. Atlas of medicinal plants of Middle America. Illinois, Thomas. 1420 p. Nobhan, G.P. and T.E. Sheridan. 1977. Living fence rows of the Rio San Miguel, Sonma, M6xico: traditional technology for floodplain management. Human Ecology S(2): 97-111. National Academy of Sciences. 1979. Tropical legumes: resource for the future. Washington, D.C., N.A.S., 326 p. National Academy of Sciences. 1980, 1983. Firewood crops. Shrub and tree species for energy production. V.l, V.2. Washington, D.C., N.A.S. 237 p. and 89 p. Nuiiez MelBndez, E. 1978. Plantas medicinales de Costa Rica y su folclore. San Jose, Costa Rica, Editorial Universidad de Costa Rica. 318 p. Otarola T., A. y Martinez H., H.A. 1985. Manejo y producci6n de cercas v i v a de Gliricidia sepium en el Nordeste de Honduras. Paper presented in Second Meeting of I.U.F.R.0 Group S1.07.07-Agroforestry, CATIE, Tumalba, Costa Rica, June 24-28,1985. Picado, W, y R. Salazar. 1984. Produccidn de biomasa y leiia en cercas vivas de Gliricidia sepium (Jacg.) Steud de dos aiios de edad en Costa Rica. Silvoenergia (Costa Rica) 1:l-4. Pittier, H. 1957. Ensayo sobre plantas usuales de Costa Rica. Zed rev.; Editorial Universitaria. 186 p. (Serie Ciencias Naturales no. 2). Russo, R.O. 1984. Studies on Erythrina poeppigiana (Walpers) O.F. Cook, a multipqose tree in Costa Rican farms. Ph.D. Dissertation Southeastern UNversity, New Orleans, Louisiana, U.S.A., 1984. 148 p. Sanchez, G.A. and R.O. Russo. 1985. Biomass production of Etythrina spp. (Papilionaceae) in the humid tropics. Paper presented to the Conference "Research in Foreshy for Energy." Copenhague, Denmark, October 28-November 1,1985. Sanchez, G.A. and L. Payne. 1987. Survey of the cultural practices and uses of Gliricidia sepium by farmers in Costa Rica In Gliricidia sepium (Jacq.) Walp.: Management and Improvement. Edited by D. Withington, N. Glover and J.L. Brewbaker. Nitrogen Fixing Tree Association, Waimanalo, HI. pp. 8-13. Sauer, J.D. 1979. Living fences in Costa Rican agriculture. Turrialba (Costa Rica) 29(4):255-26 1. Schroeder, C.A. 1976. Living fence posts in Central America. Los Angeles,

Reseatch, Reviews, Practices and Technology

87

California, University of California, Department of Biology. 11 p. (manuscript). Standley, P.C. and J.A. Steyermark. 1946. Flora of Guatemala. Fieldha Botany 24(5): 1-502.

Sumberg, J.E. 1986. Gliricidia sepium (Jacq.) S t e u e a selected bibliography. Addis Ababa, Ethiopia, InternationalLivestock Centre for Africa 12 p. Ugalde, L.A. 1979. Devrripci6n y evaluaci6n de las prkticas agroforestales en la cuenca piloto de La Suiza, Cant6n de Tunialba. Turrialba, Costa Rick UNU/ CATIE. 3 1P. Vastey, J. DE. 1962. Estudios sobre propagaci6n de especies forestales por estacas. Tesis Mag. Agr. Tunialba, Costa Rica, IICA, 67 p. Withington, D., N. Glover and J.L. Brewbaker (eds). 1987. Gliricidia sepium (Jacq.) Walp.: Management and Improvement Nitrogen Fixing Tree Association, Waimanalo, HI. 255 p.

RECEIVED: 5/14/91 REVISED: 6/5/92 ACCEPTED: 6119/92