Wilson, A.A. 1991. Browse Agroforestry Using Honeylocust.

The Forestry Chronicle. 67(No.3):232-235.




     The efficiency of establishing honeylocust (Gleditsia triacanthos L.) trees in operating pastures is being tested at the Springtree Agroforestry Project.  Existing electric fences are used as fence rows and to provide protection for honeylocust which produce high nutrient pods for animal consumption.  High production cultivars have been grafted to seedling nursery stock and planted out.  Annual pod production can be self‑harvested by sheep and cattle as a supplementary feed source.  When properly spaced, honeylocust do not significantly reduce understory grass production.  The literature on honeylocust as an agroforestry species is reviewed, and is used to develop financial rates of return for the program.  Potential internal rates of return for pasture honeylocust plantings, calculated using a variety of production and cost assumptions, show net gains ranging from 9% to 24%.  The formation of the Honeylocust Research Group is described and future research needs discussed.










     Agroforestry, or tree crop systems, have long been advanced as an answer to the dual problem of soil erosion and declining farm profits(Smith 1950; Felker and Bandurski 1979; Williams and Merwin 1983a; McDaniels and Lieberman 1979; Gold and Hanover 1987).  However, few such projects have been implemented.  Major obstacles exist to the commercial introduction of agroforestry in developed countries.  Field tests have been limited and profitability of new species has not yet been adequately demonstrated, especially in harvesting and marketing.  As with traditional orchards, agroforestry requires considerable time lag between initial cash outlays and financial returns.  Implementing agroforestry usually involves allocating land from traditional crops that offer present cash return.  Finally, agroforestry requires farmers to adopt non‑traditional methods for which there are few successful demonstrations.


     Recent developments in agriculture may, however, increase the likelihood of introducing agroforestry projects in developed countries.  Some of these factors include soil erosion, productivity reductions in agriculture, increasing cost of petro‑chemical inputs, and environmental concerns related to the use of chemicals in traditional agriculture(Gold and Hanover 1987).  A positive factor facilitating the introduction of pasture tree crops is the improved pasture management techniques utilizing electric fences. These electric fences can provide the necessary protection for inexpensively establishing pasture trees.





     The Springtree Agroforestry Project has been designed with these factors in mind.   Honeylocust (Gleditsia triacanthos L.) trees are planted directly into currently operating pastures and hayfields.  The trees produce pods which the livestock then harvest themselves under the trees.  Such a system offers several advantages. Harvesting and processing costs are nil. Honeylocust pods provide a complementary feed source in the fall when seasonal grass production is declining.  Trees are introduced to working farms without interruption of present cash flows.   Although not easily quantified, additional benefits from introducing honeylocust to pastures include: reduction of water runoff and topsoil erosion, shade for livestock, a productive pollen and nectar source for bees, a more diversified and aesthetically pleasing pasture environment, and timber upon project termination. 


      The Springtree Agroforestry Project is being implemented on a 50‑hectare farm in central Virginia where sheep and cattle are raised commercially.  A comprehensive pasture management program includes an intensive rotational grazing system.  This system, now being recommended by many state extension agents and farm journals, replaces the traditional continuous grazing practice where livestock have constant access to a large pasture(Murphy,et.al. 1986, and Murphy 1987).


     Electric fences subdivide the pastures for intensive grazing and rotation.  These fences provide an inexpensive method for introducing tree crops to the pastures by providing protection for young trees from livestock browsing and rubbing.  A single or double wire is attached to the fence in a semi‑circle around each tree and is supported with a fiberglass or metal stake in front of each trunk.  High production honeylocust cultivars have been grafted to seedling nursery stock and planted out. Trees are planted 0.5 to 1 m from the electric fence and 6 to 8 m apart along the fence row, although shading may require some later thinning. 





General Characteristics of Honeylocust


     Honeylocust is adaptable to a wide range of climate within the temperate zone and is winter hardy to temperatures of below ‑34C(Funk 1957).  It also thrives in a variety of soils, although it grows poorly on shallow, gravely, or heavy clay soils.  Honeylocust is also quite drought‑resistant (the natural range includes areas with 51 cm annual precipitation) and can survive severe storms.  Its leaves and flowers appear late in the spring and are rarely damaged by late frosts.  Although honeylocust naturally carries thorns dangerous to livestock and tractor tires, thornless trees can be produced by grafting scions taken from the thornless upper branches of the desired cultivars.


Pod Production


     Although incomplete, available data on honeylocust pod production indicate high yields in some cases and considerable variation between southern and northern regions of the temperate zone.  Annual yields of 180 kg from mature trees have been reported from South Africa and New Zealand (Loock 1947, Lennard 1980).  Ten‑year old trees grown in the southern United States at the Auburn University research project yielded 43 kg per tree(Table 1).  Honeylocust have produced considerably lower yields in the shorter growing season of the middle and northern United States(Detwiler 1947).                                     


Table 1. Honeylocust production: Auburn University study

(Scanlon 1980)[1]


            1942    1943    1944    1945    1946    1947    Average

 Tree age  (5)   (6)   (7)     (8)     (9)    (10)                 

Calhoun[2]12.0      0     14.7    28.9    10.0    20.9     14.4

Millwood  26.4      0     66.2    17.9    81.6     5.4     32.9


[1]Average dry kg per tree for 5‑ to 10‑year‑old trees.

[2]Cultivars selected for high pod production and sugar content.


     The yields at Auburn University were achieved from fertilized grafted trees, and some fertilizer applications would undoubtedly be required to achieve the higher yields, especially on marginal lands often recommended for tree crops.   A potential problem is the biennial nature of honeylocust pod production(Table 1).  This can be offset to some extent by planting different cultivars, and possibly by selective pruning, although the cost effectiveness of such pruning has not been determined(Williams 1980).  While male honeylocust are not required for pod formation, they are desirable for full seed development.  Selected thornless male cultivars are available, and a ratio of 1 male to 10 to 30 females is recommended.   High production cultivars are available, and grafting to seedling nursery stock and transplanting honeylocust is not difficult or expensive.


Harvesting by Livestock

     Honeylocust pods are readily harvested by livestock from under the trees.  Ripe pods drop gradually throughout the fall, permitting animals to feed on them over several months.  In the Auburn University study, honeylocust pod meal as feed for dairy cows was found to be palatable and was substituted for oats on a pound for pound basis(Atkins 1942).  Honeylocust pods contain large amounts of sugar and starch, comparable with corn (Zea mays L.)(Table 2).  Approximately 5% of the honeylocust pod by weight is digestible protein.  Hence the digestibility of the honeylocust seed within the pod is critical to the economic viability of the project. 


     Although cattle eat honeylocust pods, they do not digest the seed(Lennard 1980).  For cattle to get the full benefit of the protein value, the pods must first be harvested and then ground into meal.  Sheep, however, apparently do digest the honeylocust seeds. Two of the three studies reviewed corroborate our own experience that sheep do digest honeylocust seed when eaten in the pod(leRoux 1959a, Felker and Bandurski 1979, Small 1983 a and b).  The contradictory findings may be due to differences in test design, in sugar content of pods, and possibly in sheep breeds.



Table 2. Estimated nutrient and ethanol yields for corn and

honeylocust(Williams and Merwin 1983b)


                          Corn                       Honeylocust

                          -‑‑‑‑‑‑‑‑----‑‑  x103 -------‑‑‑‑‑‑‑‑‑


Protein (kg/ha)            0.3                          0.4‑2.3


Sugar and starch(kg/ha)    4.5                          2.6‑l5.0


Ethanol (1/ha)             2.8                          1.5‑9.0



Grass Production


     Evaluation of pasture agroforestry requires determining the effect of trees on grass production.  Depending on the tree and grass species selected, as well as the soil, climate, and grazing conditions, this effect may be positive or negative.  For example, in a study conducted at the University of Tennessee grass and cattle (weight gain) productivity were greater in pasture under the light shade of black walnuts trees(Zarger and Lutz 1961). 


   Casual observations of field workers suggest that pasture grasses grow better under honeylocust. The tree's open canopy throws a light shade that may encourage grass production in hot weather.  Grasses and legumes grow well right up to the trunk of the tree(leRoux 1959b, Lennard 1980). Honeylocust leaf out late in the spring and leaves die early in the fall, encouraging cool weather grass production.  In addition, the tree's small leaflets are easily absorbed into the pasture grasses during leafdrop. 


     A 17‑year study was conducted at Virginia Polytechnic Institute to measure grass production under a honeylocust plantation(Zarger and Lutz 1961).  The study found lower grass quality and output (9% on unfertilized plots and l5% on fertilized plots) for pasture under honeylocust.  However, because of design problems, the study's conclusions are open to qualification.  First, tree spacing in the test plots was too dense and the project was terminated before the effect of proper thinning could be tested.  Second, cattle were allowed to graze indiscriminately between the various test plots, and during hot weather they probably remained longer in the shaded plots.  The likely result was heavier grazing and greater sod injury and soil compaction in the shaded areas, although none of these factors was measured.  If the necessary adjustment for proper tree spacing and eliminating over grazing are made, honeylocust shade would probably not reduce grass production.  For example, at Auburn University, 6,272 kg/ha of Lespedeza sericea was harvested each year from under honeylocust with a density of 85 per ha. or a 11‑m spacing(Moore 1948).


Economic Evaluation


     The pod production and nutrition data developed above can be used to determine the economic value of honeylocust plantings.  Naturally, project costs will be incurred largely during the initial years of the project while financial returns occur only in the later years. Therefore, separate dollar cost and revenue flows must be discounted back to the present to account for the time value of money.  A measure of internal rate of return (similar in principle to cost‑benefit ratios) is used to compare these dollar flows.  The internal rate of return is the highest possible percentage return for a given dollar investment(cost of grafted trees, tree protection, maintenance, and fertilizer) and revenue(measured by the market value of pod production).


     Rates of return were calculated using the Auburn University production data and assuming 85 graphed Millwood honeylocust cultivars per ha.  To compensate for any possible reduction in grass yield, pod production was estimated conservatively to remain at the level reached after 10 years.  Because of their similar nutritional values, the price of oats was used to give a dollar value to the honeylocust pod production.  To judge the sensitivity of the internal rates of return to lower production and to lower oat prices, rates of return were also calculated assuming one‑half the Auburn University production figures and a 30% reduction in oat price.  Because of the uncertainty of fertilizer needs, calculations were made with and without the cost of fertilizer.


     Rates of return indicate a net gain from honeylocust plantings for each set of assumptions used.  Calculated rates of return range from 25% for fertilized trees under the medium production assumptions to 13% for fertilized trees under the low production assumption(Table 3).  Annual pod production of 113 kg. achieved in New Zealand gave a rate of return of over 30%.  Clearly, the profitability of livestock grazing improves with the addition of honeylocust to the pasture.


Table 3. Internal rates of return for pasture honeylocust plantings.


                                          1988 Oat      Discounted

                                           price       oat price[1]

Auburn University yields assuming           25%            20%

fertilizer cost (See Table 1)


One‑half Auburn University yields           20%            16%

assuming no fertilizer cost


One‑half Auburn University yields           13%             9%

assuming fertilizer cost


[1]Internal rates of return calculated using 70% of the 1988 oat

      price to value the honeylocust pod production.




     Before honeylocust is planted on a large scale, several important research questions remain to be answered. 


     (1) Pod Production Data.  Internal rates of return indicate that honeylocust plantings can be economically justified even for pod yields considerably below those demonstrated on 10‑year‑old trees at Auburn University.  However, additional information is needed on pod production, nutrition, and cultivar performance from different sites, especially in the middle and northern United States and in Canada.  Toward this end the author has organized the Honeylocust Research Group to systematize data collection and encourage propagation of selected cultivars among professional and amateur researchers.


     (2) Cultivars.  Cultivars with high pod production and sugar content already exist; however, given the considerable genetic diversity found in wild honeylocust, it seems probable that further improvements are possible(McDaniel 1980). One important objective for future selection and breeding is reduction of the tendency for biennial production.  Honeylocust provenance plantations leading toward breeding and hybridization work have been established by Dr. Michael Gold at Michigan State University(Gold 1985).


     (3) Cultural Techniques.  More experience with a variety of culture techniques is needed, specifically fertilizer combinations, alternative tree spacing patterns, and pruning as a possible method for reducing the habit of biennial production. Work is also needed to determine the optimal ratio of male to female trees in a pasture.


     (4) Insect Problems. Insects on honeylocust include the mimosa webworm (Homadaula anisocentra), a seed‑feeding weevil larvae (Amblycerus robiniae), and the European hornet (Vesta crabro)(Funk 1957). The effect of these insects on pod production is not currently known.  However, as has been the case with other fruit tree species, extensive monoculture plantings may increase insect problems. The mimosa webworm has caused defoliation in some urban honeylocust plantings, and a spraying program of (Bacillus thuringiensis) for this insect may prove profitable for pasture honeylocust.


     (5) Pasture Grasses.  Good pasture grass growth under honeylocust seems likely; however, experimentation with different grass and legume varieties under honeylocust shade is needed.


     (6) Seed Digestibility.  The ability of sheep to digest honeylocust seeds needs further confirmation and any differences for particular breeds noted.


     (7) Farmer Participation.  Many of the traditional objections to agroforestry projects can be met by establishing honeylocust in pastures in use.  Farmer participation should be encouraged by low initial costs(using existing fences), continued cash flow from livestock operations, and elimination of harvesting and processing costs.  Pasture honeylocust can be established in a traditional or intensive grazing system with little adjustment, allowing farmers to continue normal grazing operations. Honeylocust are now being introduced to working farms in New Zealand and Australia, where farm journals carry advertisements from nurseries offering seedling and grafted honeylocust(Smith 1985).



     In summary, planting honeylocust along existing pasture fence lines, as done at the Springtree Agroforestry Project, is technically and economically feasible.  If adopted, this agroforestry system could help reduce soil erosion by increasing the profitability of livestock grazing as an alternative to traditional row crops.




Atkins, O.A. 1942. Yield and sugar content of selected

   thornless honeylocust. Ala. Agr. Ext. Stu. 53

   Ann. Rep.:25‑26.


Davies, D.J. 1982. Forage trees: tree lucerne and other

   multipurpose leguminous species. D.Noel(Ed.)

   Tree crops: the 3rd component. Proc. First Australasian

   Conference on Tree and Nut Crops: 107‑120. Subiaco,



Detwiler, S.B. 1947. Notes on honeylocust. USDA Soil

   Conserv. Serv.(mimeo).


Felker, P. and R.S. Bandurski. 1979. Uses and potential

   uses of leguminous trees for minimal energy input

   agriculture. Econ. Bot. 33:172‑189.


Funk, D. T. 1957.  Silvical characteristics of honeylocust.

   USDA For. Serv., Cent. States Forest Exp. Sta. Misc.

   Rel. 23.


Gold, M.S. and J. W. Hanover. 1987. Agroforestry systems

   for the temperate zone. Agroforestry Systems 5:109‑121.


Gold, M.S. 1985. Honeylocust: genetic variation and

   potential uses as an agroforestry species.  Ph.D.

   Thesis. Michigan State Univ.


Lennard, T. 1980. The story of honey locust. J. N.Z. Tree

   Crops Ass. 5:14‑17.


leRoux, P.L. 1959a. Honey locust tree useful sources of

   fodder for stock. Farming in S. Afri. 35:20‑21.


leRoux, P.L. 1959b. Advantages and disadvantages of

   honeylocust. Farming in S. Afri. 35:8.


Loock, E.E. 1947. Three useful leguminous fodder trees.

   Farming in S. Afri. 2:7‑12.


MacDaniels, L. and A.S. Lieberman. 1979. Tree crops: a

   neglected source of food and forage from marginal lands.

   Biosci. 29:173‑175.


McDaniel, J.C. 1980. A plant breeder looks at some American

   tree crops: Morus, Gleditsia, and Diospyros. In: Tree Crops for     Energy Co‑production on Farms: 113‑118. US Solar Energy Res.

   Inst. Golden, Colorado.


Moore, J.C. 1948. The present outlook for honeylocust in

   the south. North. Nut Growers Ass. 39th Ann. Rep. pp.



Murphy, B. 1987. Greener Pastures on Your Side of the

   Fence: Better Farming with Voisin Grazing Management.

   Arriba Publishing Co., Colchester, Vermont. 232 pp.


Murphy, W.M., J.R. Rice and D.T. Dugdale. 1986. Dairy farm

   feeding and income effects of using Voisin grazing

   management of permanent pastures.  Am. J. Alt. Agric.



Scanlon, D.H. 1980. A case study of honeylocust in the

   Tennessee Valley Region.  In: Tree Crops for

   Energy Co‑production on Farms: 21‑31. US Solar Energy

   Res. Inst.  Golden, Colorado.


Small, M., 1983a. Sheep can digest 90% of the whole

   honeylocust seeds ingested. Agrofor. Rev. 3:6.


Small, M., 1983b. Honeylocust pods and the digestion of

   protein by sheep. Agrofor. Rev. 4:6‑7.


Smith, J.R. 1950. Tree Crops: A Permanent Agriculture.

   Derin‑Adair, New York, NY. 408 pp.


Smith, R., 1985. Contented cow leads to value of honey

   locusts. Growing Today. 2:3‑4.


Williams, G. 1980. Tree crops for energy production in

   Appalachia. In: Tree Crops for Energy

   Co‑production on Farms: 7‑20.  US Solar Energy Res.

   Inst. Golden, Colorado.


Williams, G. and M.L. Merwin. 1983a. Energy‑ and

   soil‑conserving perennial crops for marginal land in

   temperate climates. In: W. Lockeretz(Ed.)

   Environmentally Sound Agriculture: 317‑332. Praeger Co.,

   New York, NY.


Williams, G. and M.L. Merwin. 1983b. Potential tree crops

   for the production of staple nutrients and alcohol fuel

   on hill lands. In: D.B. Hannaway(Ed.) Foothills

   for Food and Forests: 365‑367.  Timber Press. Beaverton,



Zarger, T.G. and A.J. Lutz. 1961. Effects of improved

   thornless honeylocust shade trees on pasture

   development.  TVA Div. of For. Relat., Tech. Note No.

   34. 12 p.