Optimal Nutritive Value of Honeylocust Seed Pods Within Temperate Silvopasture

Final report for GS09-079

Project Type: Graduate Student
Funds awarded in 2009: $9,894.00
Projected End Date: 12/31/2009
Grant Recipient: Virginia Tech
Region: Southern
State: Virginia
Graduate Student:
Major Professor:
Expand All

Project Information


Seed pods from high-sugar varieties of honeylocust (Gleditsia triacanthos L.) trees have potential as an animal feed supplement from late fall to mid-winter. However, data regarding yields and variation in nutritive characteristics of pods from improved honeylocust varieties such as ‘Millwood’ are limited. Furthermore, the degradation of pods over time as it affects the supply of high quality fodder for ruminants grazing within silvopastures is unreported. A study was conducted within an active honeylocust silvopasture to: 1) determine seedpod yields of Millwood honeylocust trees managed within silvopastures, 2) estimate the nutritional variability of seed pods among Millwood honeylocust trees, and 3) determine changes in fodder nutritive value over time. In October 2008, 2009, and 2010, seedpod yields were estimated using both visual classification and pod-harvests from representative trees. In 2008 and 2009, 12 randomly-sampled pods were collected from each pod-bearing Millwood tree to determine the variability in fodder nutritive value among trees. Additional Millwood and wild-type seedpods were placed in in-situ bags and allowed to decompose within silvopastures over time. At monthly intervals from November to March, in-situ samples were collected and analyzed for nutritive value. Levels of digestibility, neutral and acid detergent fibers, acid detergent lignin, and crude protein were characterized. Ground Millwood seedpods were comparable to whole-ear dent corn in terms of nutritive value. Both ground pods and seeds were highly digestible (78.7 and 96.3%, respectively) and low in fiber and lignin. Seeds, with over 20% CP, have potential as a CP supplement. Millwood trees displayed alternate bearing patterns with 3-yr average dry matter yields of approximately 12 kg tree-1. Since honeylocust pods represent a potential high-quality feed resource for ruminants, studies characterizing their nutritive value and degradation may influence management decisions regarding their utilization within silvopastures


ADF = Acid detergent fiber
ADL = Acid detergent lignin
CP = Crude protein
IVTD = In vitro true digestibility
MW = Millwood
NDF = Neutral detergent fiber
WT = Wild-type

Millwood honeylocust tree with high pod yield


Statement of Problem, Rational and Significance

Sustainable farms are rare in Appalachia due to severe topography, limited land use potential, lack of diversification, and small farm size (ASA, 1989). Steep slopes and rocky terrain characterize the landscape, making much of it unsuitable for crop production (Proctor and White, 1962). Historically, these lands have been cleared for pasture, despite their tendency to erode and degrade rapidly. In addition, by producing only a single product, farmers in these areas do not take advantage of the diverse, natural environmental niches of Appalachia and limit their ability to cater to various markets. Farm size also constrains the economic sustainability of single-product production systems, as typical Appalachian farms are considered too small (150 acres) to be economically viable (USDA, 1999).

Livestock production on small Appalachian farms using marginal lands is not considered sustainable (Baker et al., 1981). Appalachian farmers typically have limited resources and it is crucial to their livelihoods that those resources be appropriately and adequately utilized to secure environmental and economic sustainability. The purpose of this project is to explore ways of increasing and diversifying production from land resources and to help overcome resource limitations by incorporating trees into pasture-based production systems (silvopasture). 

Silvopasture is a land management system in which trees, forages, and livestock are simultaneously managed on the same area of land. Interactions between trees and ground vegetation result in a complex set of competitive and complementary effects, whose relative impact varies in space and time around trees (Scanlon, 1982; Sharrow, 1991). The benefits of a silvopasture system may include: (1) increased production and nutritive value of cool-season forages in mid-summer; (2) production of additional products such as nuts, fodder, and timber; (3) increased animal comfort due to shade; (4) greater soil fertility due to nutrient recycling and nitrogen fixation; and (5) reduced soil erosion and improved stream quality.

Honeylocust trees have gained particular interest for silvopasture systems. Honeylocust growth patterns (leafing out in late spring) and morphology (open canopy with compound leaves) are ideal for agroforestry in that they minimize competition with forages. Further, the trees have been noted as a potential source of supplemental feed for ruminants. Honeylocusts produce edible seed pods that livestock may consume after pods drop in autumn. Millwood, a high-sugar honeylocust variety produces pods that contain elevated levels of non-structural carbohydrates relative to pods from wild honeylocust trees. Millwood pods may have sugar concentrations as high as 368 g kg-1 (Gold and Hanover, 1993) and the cultivar produces significantly greater pod yields. By one estimate, an acre of well-managed honeylocust trees can produce yields of nutritious fodder comparable to an equal area of oats (Smith, 1950). However, timing of maximum nutritive value is not well defined. Moreover, the utility of pods may vary with animal species. Because cattle lack upper teeth, they may be unable to break the hard seeds and digest the seed proteins. In such case, some degradation may actually improve nutrient availability. Research regarding the nutritive characteristics of honeylocust is thus needed to further clarify this tree’s fodder potential and best management within silvopasture systems.

Project Relevance to Sustainable Agriculture

Honeylocust silvopasture production systems have great potential benefit to small farms by increasing on-farm resources through provision of supplemental energy feed for grazing livestock. Incorporating trees into pasture systems can also improve animal comfort and, depending on management, they may also diversify farm income through production of timber or cordwood. Although introduction of honeylocust trees into pasture systems would be expected to be scale neutral (i.e. applicable across the range of farm sizes) their use may be of particular value for those farming small or hilly acreages where higher-value production systems are needed to improve small farm viability. Limited resource farmers such as those in Appalachia need management alternatives to help increase short- and long-term productivity of their land. 

Integrated tree-crop forage-livestock production systems may help make farms more diverse, productive, and sustainable, while minimizing adverse off-site environmental impacts. 
Because the Appalachian Region contains headwaters of most of the rivers that provide water for the major population centers of the eastern United States, it is critical that increases in the agricultural productivity of the region not compromise water quality. Incorporating honeylocust trees into pasture production systems may actually improve environmental health because trees protect stream quality and conserve soil fertility. Trees retain and recycle on-site nutrients because they are excellent scavengers of nutrients from the subsurface flow. In addition, leaf litter from trees adds organic matter to the soil, aiding soil aggregation and reducing nutrient run-off.

Our project focused on evaluating the productivity and nutritive value of honeylocust seed pods because data on nutritive characteristics of improved varieties of honeylocust (such as Millwood) are limited. Although utility for cattle may differ from that for sheep, honeylocust pods represent a potential high-quality feed resource for ruminants, and information on this resource is needed to help guide producer management decisions regarding tree selection and utilization within silvopastures. These efforts are part of a larger project to develop diverse, sustainable, and environmentally sound tree-crop and forage-livestock production systems for farms on marginal lands with limited resources.

Project Objectives:
Project Relevance to Sustainable Agriculture

Our project has four main objectives: 
1) Determine yields and year-to-year variation of Millwood honeylocust pod production for trees managed within active silvopastures
2) Estimate the nutritional variability of pods produced from Millwood honeylocust trees managed in a silvopasture. 
3) Determine changes in husk and seed nutritive value and digestibility from three distinct honeylocust tree types over time after pod drop 
4) Characterize changes in pod development and nutritive value through the growing season and to determine their relationship to time of pod drop. 

We hypothesized that honeylocust silvopasture systems with immature trees could produce pods of sufficient yield and quality to provide financial benefit for land managers. Further, we hypothesized that nutritive value of honeylocust seed pods varies by tree type (i.e., by genetic origin) and by time (through the periods of pod development and decay), and that decay may improve digestibility of hard-coated honeylocust seeds for some livestock species. Time of peak nutritional value for different tree types also could be determined to optimize honeylocust utilization within silvopasture systems.


Click linked name(s) to expand
  • Jacob Johnson


Materials and methods:
Yield and nutritional variability of Millwood seedpods:

Visual assessment was used to estimate pod yield from Millwood trees. Measures such as percent branches bearing pods, percent cover per branch, and relative pod density were collected for all Millwood trees. Trees were ranked according to the following scoring system:
1= Low pod yield; ? 33% pod cover
2= Medium pod yield; 34-66% pod cover
3= High pod yield; ? 67% pod cover
Digital images also were collected for all Millwood trees and archived for reference. Measures were collected from three representative trees from each scoring class (nine trees, total). All pods on each representative tree were counted and used to estimate pod numbers for the remaining trees. Twenty pods from each representative tree also were collected at random and dried to determine an average mass per pod. Estimates of pods yields for unsampled trees were based on the relationship between pod numbers from sampled trees and their pod data (branches with pods, pod cover/branch, pod density). Pod production was standardized to a measure of tree size using measures collected in spring 2008. 

In October 2008, twelve pods were randomly sampled from each pod-bearing Millwood tree to determine the variability in fodder nutritive value among trees. Pods from each tree were dried (48 to 72 h at 55C) and fractionated into seed and husk components. Fractions were ground and analyzed for ADF, NDF, ADL, non-structural carbohydrates, and CP concentrations, and relationships between nutritive value and tree size and site conditions was explored.

Degradation of honeylocust seedpods:

Declines in pod dry matter and nutritive value over time were determined by exposing honeylocust pods to environmental conditions. Thirty pods were collected from six trees for each of the three honeylocust tree types (Millwood, Millwood-scion, and wild-type honeylocust) in 2008. However, due to low bearing in 2009, only Millwood and wild-type pods were tested. These pods were randomly selected and composited (n=180 or n=120) by tree type. Nylon mesh in situ bags with a surface area slightly larger than the surface area of six to eight pods, a pore diameter of ~4 mm, and ~5 pores per cm2, were used to hold pods during the period of exposure to environmental conditions. The large pore diameter was used to ensure exposure sunlight, moisture, insects, and other factors influencing decomposition. Six to eight pods were placed in each in situ bag, and bags were laid flat on the ground surface under trees in the silvopasture. At monthly intervals from November to March, in-situ samples (four replications per treatment) were collected. Pods from each of the bags were dried (48 to 72 h at 55C) and fractionated into seed and pod components. Percent seed damage will be determined by measuring characteristics such as cracks, holes, or insect damage. Seed samples will be split, with half saved as whole seeds. The remainder of the seed samples and the pod husk were ground (separately) to pass a 1-mm screen. Whole and ground seed and husk samples were analyzed for IVTD. Ground samples also were analyzed for NDF and ADF, ADL, and CP concentrations. Because of low bearing and poor pod fill in 2009, whole seed sufficient seed for testing were available only in 2008. We used these to compare digestibility of whole and ground seeds and to determine the degree to which nutrient losses from seed degradation (e.g., due to cracking and leaching or insect herbivory) was offset by increased hard seed permeability and nutrient availability during digestive processes.

Pod development and drop study:

This study was eliminated due to limited pod development and availability in 2009.

Sampling honeylocust trees for yield estimates

Research results and discussion:
Nutritional variability of Millwood seedpods:

Both ground pods and seeds were highly digestible (787 and 963 g/kg, respectively). Pod IVTD values ranged from 649 and 874 g/kg, and seed IVTD ranged from 900 to 991 g/kg. Pod NDF, ADF, and ADL values were generally low and averaged 273, 193, and 63 g/kg, respectively. Pod NDF concentrations ranged from 184 to 438 g/kg, ADF from 131 to 321 g/kg, and ADL from 40 to 108 g kg/. Few trees produced pods with high fiber and low digestibility. Seeds in 2008 contained 132 g kg NDF, 75 g/kg ADF, and no measurable ADL. Seed NDF concentrations ranged from 68 to 160 g kg-1 and ADF concentrations ranged from 47 to 101 g/kg. Seeds had greater CP concentrations than pods; values for the two fractions averaged 204 and 62 g/kg, respectively. The range of CP values ranged from 155 to 255 g kg-1 for seeds and 31 to 101 g/kg for pods. Estimated values for IVTD, NDF, ADF, ADL, and CP of whole seedpods (pods and seeds together) in 2008 were 833, 235, 161, 63, and 99 g/kg respectively. Although seeds were only 29% of DM, they supplied 60% of total CP within whole seedpods.

Millwood yield variation:

Nearly all Millwood trees displayed some variation of alternate bearing pattern. Those few trees having similar yields from year to year had relatively low pod yields in all three years. Tree DBH was weakly correlated with seedpod yields during productive years. Yields, averaged over all 72 study trees (including non-bearing trees), were 13.4, 0.9, and 12.3 kg/tree, for 2008, 2009, and 2010, respectively. A 13-to 15-yr-old Millwood silvopasture planted with 170 trees/ha and producing yields similar to those in this study would bear approximately 2.3 Mg/ha of pods (DM basis) in a productive year and 0.14 Mg/ha of pods in a low-yielding year (Table 4.5). Based on a 3-yr average of 1.51 Mg/ha, and using an average value of oat grain for 2008 – 2009 ($3.15/bushel; USDA, 2009), Millwood trees could generate over $325 ha/y1 in feed supplement equivalent.

Millwood and wild-type pod degradation:

Pod nutritive value changed very little in 2008, but a general pattern of decreasing digestibility and increasing fiber concentrations with time was observed for both pod types. Decreases in pod IVTD and corresponding increases in pod fiber and lignin concentrations were greater in 2009 than in 2008, most likely due to wetter environmental conditions. Millwood pods were more digestible and less fibrous, despite being more resistant to degradation than pods from WT honeylocust. The decrease in digestibility from November to March was over two-fold greater for WT pods compared to MW pods. In general, pod CP did not change over winter in either year. Further, differences in CP between tree types, although significant, were usually small. Honeylocust seeds were largely resistant to short-term degradation. Although the differences were minute, ground MW seeds were less fibrous, more digestible, and contained less CP than WT seeds. Ground honeylocust seeds were nearly 5 fold more digestible than whole seeds. Evidence of increased seed digestibility with time of exposure was limited, despite bug infestations and cracking.

Change in most pod mineral concentrations (Ca, Mg, and Cu), as well as the relative differences between pod types, varied from year to year and no consistent patterns were evident. Concentrations of macro-minerals, such as P and K, decreased over winter in both 2008 and 2009. Whole honeylocust seedpods have greater K and Ca concentrations than more conventional feeds with comparable nutritive quality. Seedpod concentrations of P, S, and Mg fall below the maintenance requirements of a dry pregnant cow.Thesis data tables for honeylocust nutritive value and yield results, and comparison with other feedstuffs

Participation Summary

Project Outcomes

Project outcomes:

Our data show that Millwood honeylocust trees can supply substantial value to a silvopastoral grazing system in terms of a late fall/early winter nutrient supplement.

Economic Analysis

Although we have not conducted a thorough analysis, we can present some initial estimates of returns in “adolescent” Millwood silvopastures. First, we assume 13-to 15-yr-old Millwood silvopasture planted with 170 trees/ha would produce yields similar to those in this study (approximately 2.3 Mg/ha of pods (DM basis) in a productive year and 0.14 Mg/ha of pods in a low-yielding year. Based on a 3-yr average of 1.51 Mg/ha, and using an average value of oat grain for 2008 – 2009 ($3.15/bushel; USDA, 2009), Millwood trees could generate over $325 ha/y1 in feed supplement equivalent. 

This estimate has both optimistic and conservative elements. Optimistically, and to be realized by the producer, this resource would have to be completely “captured” (i.e., eaten). It is unlikely that this would happen, but we do not have a working sense of how much would be captured in these systems – there is room for work on this. However, our yield averages include non-bearing trees, and we had only about 1/3 of the landscape planted to Millwood because these trees were limited in number. The original planting included many wild-type trees to create the silvopastoral effect.


We anticipate two publications on honeylocust trees from Jacob Johnson’s MS thesis. In addition, Jacob should have a general review about the use of trees in sustainable cropping systems along with another chapter on black walnut production. These will be submitted by the end of the year as he has taken on a new job and is decompressing from his MS work. 

These data also may be used in an extension publication to be developed with an economist. The research for the site routinely is used for class work and for farm visits.

Farmer Adoption

Producers in South Africa are looking at systems using honeylocust trees as part of an overstory crop for sheep production. They already have honey locust from 50 years ago but were looking for better pod types, so we have supplied some seed for them to test.

Getting US producers to try these systems will be slower, likely because of the paradigm shift in perspective. For example, many producers, when approached with this idea rather readily comment that they’ve spent their life cutting out trees…

We believe that greater adoption can come as we promote these efforts through extension meetings and perhaps with other grants where recognized leaders in the community can put in these systems in areas of high visibility.


Areas Needing Additional Study

Feeding studies that give a better picture of animal performance – both with small and large ruminants – are needed to understand the ability of different types of livestock to utilize the pods (and browse) from these trees. Interestingly, sprouts from wild-type honeylocust that were thinned out of the system were a favorite feedstuff by sheep during the growing season.

The ecology of these systems needs further understanding. Little information is available on how incorporating honeylocust trees affects forage production and animal behavior.

Limitations for adoption are another key issue for getting any trees into pasture systems. Honeylocusts may be more challenging than others because it has fewer high-value outputs than other potential silvopasture trees such as black walnuts and pines. The thorny nature of the plant may be a challenge unless thornless grafts are available, and study will be needed to develop markets and opportunities for industry to produce thornless, high-sugar varieties.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.