Selecting a sunn hemp cover crop genotype for weed suppression and seed production

Final Report for LS08-205

Project Type: Research and Education
Funds awarded in 2008: $170,000.00
Projected End Date: 12/31/2011
Region: Southern
State: Florida
Principal Investigator:
Dr. Carlene Chase
University of Florida
Expand All

Project Information


Studies were conducted in 2008 and 2009 to evaluate the potential for sunn hemp as both a cover crop and as a seed crop in Griffin (Georgia), Gainesville (Florida), and Lajas (Puerto Rico). Field trials were conducted to evaluate the phenotypes of sixteen accessions of sunn hemp. In Georgia, significant differences were found among the sunn hemp accessions tested, for all morphological traits evaluated at 2 and 4 MAS. Principal component analysis showed that the first principal component accounted for 77% of the total variation while principal components 2 and 3 progressively accounted for 87 and 92%, respectively of the variation and average linkage cluster analysis grouped the original 16 accessions into well-defined phenotypes with 4 distinct groups and one outlier based on total biomass which included totaled morphological data and number of seeds produced. In Florida the 16 accessions could be separated into two distinct groups based on several vegetative and reproductive parameters. Accessions in Group 1 were tall with few branches, and produced few to no flowers and pods. These accessions included 2 (PI 234771), 3 (PI 248491), 7 (PI 295851), 14 (PI 468956), 15 (PI 561720), and 16 (PI 652939). Accessions in Group 2 flowered early (49-66 days after seeding) and seemed to be daylength insensitive. These accessions were shorter in stature, and lower in biomass than those in Group 1, but produced more branches, flowers, and seed. These included accessions 1 (PI 207657), 4 (PI 250485), 5 (PI 250486), 6 (PI 250487), 8 (PI 314239), 9 (PI 322377), 11 (PI346297), 12 (PI391567), and 13 (PI426626). In Puerto Rico, the accessions 2 (PI 234771), 12 (PI 391567), and 16 (PI 652939) were found to have potential for biomass and/or seed production.

A second field study was conducted in 2008 and 2009 to investigate the effects of seeding rates (10, 25 and 40 lb/ac) and removal of apical dominance on weed suppression and seed production. In Georgia, similar weed biomass amounts were detected regardless of sunn hemp cutting date both years and a lower seeding rate for sunn hemp as a cover crop for weed reduction is feasible. Additionally, these results show a sunn hemp grower that low seeding rates are as effective as higher rates. In Gainesville, cutting to break apical dominance had no significant effect on weed suppression and on flowering, but did induce branch formation. All three seeding rates had lower total weed biomass, but were not significantly different from one another. No seed production occurred in Florida and Georgia. In Puerto Rico, the highest biomass and seed yields were obtained with the intermediate planting density of 25 lb/ac, and apical cuttings did not improve C. juncea yields or weed suppression.

Phytotoxicity was observed with aqueous extracts of all fourteen accessions screened with a bioassay for allelopathic potential. High performance liquid chromatography provided evidence of a compound in foliar extracts with characteristics of 5-hydroxynorleucine, the amino acid derivative with allelochemical properties previously reported only in sunn hemp seeds. Sunn hemp also was shown to have the potential to serve as a catch crop for potassium. However, residue from sunn hemp seed crops was recalcitrant showing negligible decomposition over a 12-week period.

A study was also conducted to evaluate the economics of utilizing sunn hemp as a cover crop. Partial budgets were prepared for five summer fallow treatments; sunn hemp, velvet bean, cowpea, sorghum sudangrass, and tillage. These treatments are compared for weed suppression, nitrogen contribution, and the potential impacts on a following cash crop of squash. Sunn hemp was the least expensive summer fallow treatment, followed by velvet bean, cowpea, sorghum sudangrass, and tillage.

Project Objectives:

(1) Evaluate the effect of different geographical locations on biomass accumulation, flowering and seed yield of the USDA’s sunn hemp germplasm collection;

(2) Identify bee species visiting sunn hemp flowers to determine which are effective pollinators and quantify their visits;

(3) Assess the phenotypic variability of flowering and characterize the sensitivity to environmental factors;

(4) Investigate the effects of breaking apical dominance on weed suppression and seed yield, and compare the allelopathic potential of the accessions;

(5) Determine how cultural practices for sunn hemp seed production influence nitrogen accumulation, decomposition, and plant available soil nitrogen; and

(6) Evaluate the economic costs and benefits of sunn hemp domestic seed, cover, and fodder crops.


The focus of this project was to use an interdisciplinary approach to the selection of an appropriate genotype of sunn hemp (Crotalaria juncea L; Fabaceae family) for use as a seed crop as well as a summer cover crop. Sunn hemp is grown as a fiber crop in South Texas primarily for paper and cat litter production. It is also used for the manufacture of twine, cord, marine cordage, fishing nets, matting, and sacking (Purseglove, 1981; Cook and White, 1996). As a fiber crop, it is superior to kenaf (Hibiscus cannabinus) because of its nitrogen fixation capability and resistance to plant pathogenic nematodes Meloidogyne incognita, Rotylenchulus reniformis (Dempsey, 1975; Wang et al., 2001). Additionally, sunn hemp is used as silage, green manure (Ramos et al., 2001), cover crop for control of weeds (Linares et al., 2007; Sangakkara et al., 2006), prevention of soil erosion (Miller, 1967), and mitigating groundwater contamination by atrazine (Potter et al., 2007). Sunn hemp seeds are high in dietary fiber and may have potential for use as a nutraceutical (Morris and Kays, 2005).

Adoption of this potentially useful cover crop is hindered because one variety is unlikely to perform adequately in different geographical regions. Sunn hemp seed was not commercially available in 2005 and most of 2006 (Klassen, 2007). Seed prices in 2007 were prohibitively high at $6.90/lb. The limited seed supply originates from South Africa and Hawaii where short days induce photoperiodic flowering (White and Haun, 1965). These constraints present an outstanding opportunity to explore the physiology, genetics, and cultural practices that influence and control flowering of sunn hemp. The overall purpose of the studies outline below was to prepare a foundation for development of new cultivars that maintain their genetic characteristics that make them an outstanding cover crop while selecting for their ability to set and produce high-value seeds in the Southeast.

Characterizing Sunn Hemp Accessions

The USDA’s sunn hemp germplasm collection contains accessions from many geographical locations. A number of genetic traits had been evaluated, but no studies had exclusively examined flowering habits and seed set. Mainland US production of sunn hemp seed is limited by a number of proposed factors including inadequate pollination, as well as environmental constraints that affect flowering and seed production. In addition, very little was known about the biomass production capacity of the sunn hemp accessions. In order to explore the feasibility of producing seeds in the southern region it was necessary to undertake phenotypic characterization. Assessing the USDA’s accessions for desired traits under varying environmental regimes was used to delineate how this crop can be further developed into a seed/cover crop for the southern region.

Ascertaining the Role of Pollinators

The presence of effective pollinators may be one of several factors that impede seed set. We hypothesized that poor and often inconsistent pollination observed in sunn hemp grown in the southern region may be the result of insufficient bee fauna of appropriate size.

Although, sunn hemp is dependent on bee pollination, because of the structure of its blossom, relatively few species of bees have the anatomy and behavior necessary to achieve effective pollination (Westerkamp, 1997). Even in some geographical areas of sunn hemp’s native India, the plant produces little seed, evidently due to a lack of effective pollinating bees (Free, 1993). In this study we will determine to what degree native pollinators adequately pollinate sunn hemp.

The keel petals of sunn hemp’s papilionaceous flower enclose the sporophyll column and taper to a point, within which is confined released pollen Westerkamp, 1997). The downward movement of the keel relative to the dorsal “flag” petal (vexillum) extends the style out of the orifice at the end of the keel and expels the surrounding pollen. Effective pollinators must be sufficiently strong to depress the keel, while retrieving nectar through the base of the vexillum. The orifice of the keel is separated from the opening in the nectar chamber by about 1.5 cm. Bees must be large enough to allow pollen released from the keel to be deposited on the side or venter of the abdomen.

Based on flower morphology the most effective bees would be large females of the Megachilidae family (Westerkamp, 1997; Etcheverry et al., 2003), which have pollen collecting hairs (scopae) on the venter of the abdomen, rather than on the hind legs or propodeum as do most other bees. This location of the scopa is optimal for gathering pollen from the sunn hemp blossom. Pollen does not adhere as effectively to abdominal hairs of other large bees (Etcheverry et al., 2003), that seem to be effective pollinators, mainly Xylocopa spp. (Purseglove, 1981; Nogueira-Couto et al., 1992; Free, 1993).

Although many Crotalaria species are self-pollinated (Koul et al., 1983), C. juncea is predominantly cross-pollinated (Kundu, 1964; Free, 1993). The stigma is not responsive until it extends from the keel and comes into contact with the body of the insect. The flower may then be cross-pollinated if the insect has pollen on her abdomen from a previous visit to another blossom. Once the stigma has been stimulated, the flower then can be self-pollinated (Howard, 1919 et al. cf Free, 1993).

Molecular-Genetic Approaches to Understanding Flowering

The lack of seed set in sunn hemp is a hindrance that warrants an understanding of the fundamental physiology and growth habits that preclude efficient floral induction and seed set. We hypothesized that exploration of environmental parameters that influence the transition to flowering may lay the groundwork for marker-assisted selection and development of elite germplasm. The use of molecular-genetic techniques may expedite the creation of new cultivars from underutilized accessions, hastening the exploitation of Crotalaria as a high-value, cover/seed crop.

Sunn hemp that is commercially available in the US is propagated by seed in South Africa and Hawaii, two distinct environments with respect to temperature, photoperiod, seasonal influence, and soil type. This suggests that the successful growth of sunn hemp is influenced by a series of physiological overlaying factors that together influence the floral transition.

Over the last decade research has resulted in greater understanding of the molecular-genetic basis for flowering in Arabidopsis thaliana, the model plant system. The discoveries quickly spread to crops of economic importance, such as rice (Griffiths et al., 2003), poplar (Bohlenius et al., 2006), wheat (Nemoto et al., 2003) and even strawberry (Stewart et al., 2007). The data indicate that the central pathways consist of a common set of biochemical componentry, even in species with great divergence and varying photoperiod sensitivity. Variation comes from how these proteins interact with each other and perhaps novel interactors, yet it is well established that the core of this pathway is highly conserved among angiosperms.

The basic circuitry links light signals from the environment to coordinated gene expression through the photoperiod pathway. The pathway is comprised of a series of regulatory proteins that almost certainly exist in Crotalaria. A comprehensive understanding of the central regulation of the floral transition may be gained through study of these regulators in this crop, their expression, natural variation and any correlation with specific flowering habits.

Improving Weed Suppression with Sunn Hemp

Cover crops suppress weeds by competing for resources, inhibiting weed germination and/or growth with phytotoxins (allelopathy), and by modifying soil microclimate. We hypothesized that breaking apical dominance at early in development and minimizing biomass removal will result in a growth habit that promotes more rapid canopy closure for better weed suppression and increased flower number.

Although sunn hemp was not inherently more competitive than yellow nutsedge and smooth pigweed (Collins et al., 2007) increasing the population density caused more rapid canopy closure and more effective weed suppression (Collins et al., 2008). Wide row spacing of 90 – 105 cm recommended for sunn hemp seed production delays the rate of canopy closure and, therefore, synthetic herbicides are recommended for weed control (eg. Rotar and Joy, 1983). Since synthetic herbicides are prohibited in organic production, increasing sunn hemp’s ability to compete with weeds by changing planting arrangement and shoot morphology is proposed as a cultural alternative. Branching usually occurs above 60 cm; however, less branching and higher initiation on the stem occurs as sunn hemp population density increases (Rotar and Joy, 1983). Loss of apical dominance may explain increased branching and flower number when plants were cut to 90 cm at 100 days after seeding (Abdul-Baki et al., 2001).

Nonprotein amino acids (NPAs) are typical of the Fabaceae family. NPAs mimosine and albizzine have been demonstrated to be phytotoxic (Williams and Hoagland, 2007) and velvetbean produces the allelochemical NPA L-3,4-dihydroxyphenylalanine (L-DOPA) (Fujii et al., 1991). The NPA, 5-hydroxynorleucine, has been isolated from sunn hemp seeds (Pant and Fales, 1974) and shown to inhibit lettuce seed germination (Wilson and Bell, 1979). We hypothesized that, like L-DOPA, 5-hydroxynorleucine may also be produced in plant organs other than seeds and may partially explain weed-free zones beneath sunn hemp canopies.

Allelopathic potential of sunn hemp was demonstrated with aqueous extracts of sunn hemp leaves, which suppressed germination of livid amaranth (Amaranthus lividus) and bell pepper seeds (Adler and Chase, 2007). Soil-incorporated, ground, dried sunn hemp leaves reduced germination, plant height, and shoot dry weight of smooth pigweed (A. hybridus).

Residue Management of Sunn Hemp Following Seed Harvest

Producers will be more likely to adopt practices for sunn hemp seed production if it can provide multiple ecological services including weed suppression and nitrogen contribution. Sunn hemp used as a green manure has successfully provided nitrogen (N) to subsequent vegetable crops in experiments located in the southeastern US (Balkcom and Reeves, 2005), subtropics (Cherr et al., 2007), and tropics (Jeranyama et al., 2000). We hypothesized that planting density, clipping date, and termination date will influence plant growth and development and carbon (C) to nitrogen (N) ratios.

In many parts of the southeastern US, vegetable production is located on sandy soils with little organic matter, silt or clay. These soils have low water holding capacity and rapid water infiltration rates. High precipitation can exacerbate nitrate loss to ground water if subsequent crops do not utilize soil solution nitrate in time (Cherr et al., 2007). Soil solution is estimated to move downward through the profile of these soils 1.9 – 3.8 cm for every 0.25 cm of water applied (Simonne et al., 2005).

Sunn hemp N contribution to soil is greatest when terminated at midbloom approximately 60 days after planting (DAP) (Cherr, 2004). At midbloom, whole plant C:N ratios are typically 25:1-35:1. Around this time, biomass allocation changes from leaves to stems, the C:N ratio increases, and plant N content decreases (Abdul-Baki et al., 2001; Cherr, 2004). In contrast, at seed maturity (120 DAP), sunn hemp stems are woody and above ground plant tissue C:N ratios can exceed 45:1, thus reducing the rate of plant available N. Therefore, producing sunn hemp for seed may be an alternative for producers to avoid risk to water quality while still benefiting from a summer cover.

During the project the focus of this work changed somewhat to focus on the potential for sunn hemp to utilize soil potassium to assess its likely role as a catch crop for potassium and additionally on residue quality and decomposition rate of terminated sunn hemp grown for seed.


Click linked name(s) to expand
  • Dr. Kevin Folta
  • Dr. H. Glenn Hall
  • Dr. Alan Hodges
  • Dr. Rosalie Koenig
  • Marty Mesh
  • Dr. Jose Morales-Payan
  • Dr. John Morris
  • Dr. Marilyn Swisher
  • Dr. Danielle Treadwell


Materials and methods:

Objective 1. Evaluate the effect of different geographical locations on biomass accumulation, flowering and seed yield of the USDA’s sunn hemp germplasm collection

Griffin, Georgia

Planting: A 2-yr field experiment was conducted in 2008 and 2009 on the farm at the Westbrook Campus of the USDA, ARS, Plant Genetic Resources Conservation Unit (PGRCU), Griffin, GA. Seeds from 16 sunn hemp (Crotalaria juncea L.) accessions were scarified by placing the seeds in cheesecloth followed by immersion in boiling water for approximately 4 sec (Table 1). A slurry consisting of the cowpea type Rhizobium inoculum (Nitragen, Milwaukee, WI) was applied to the sunn hemp seeds. The seeds were then planted in jiffy pots containing potting soil and grown in a greenhouse from May through early July to account for three planting dates both years. The first, second, and third planting dates occurred on May 5th, June 5th, and June 26th in 2008 and 2009. After approximately one month of greenhouse growth, sunn hemp seedlings from each accession were transplanted by hand to the field plots. Each sunn hemp accession was transplanted into two rows spaced 61 cm apart on 1.8 m row centers. Plants were irrigated as required. The experimental design was a split plot with three planting dates assigned to the main plots and arranged in a randomized complete block with four replications. Planting dates were at about 30-d intervals. Subplots contained 16 sunn hemp accessions from several countries (Table 1).

Morphological and Reproductive Variation: Two months after seeding (MAS), 5 randomly selected plants from each plot (20 plants total per accession) were used for data collection. The number of primary and secondary branches, open flowers/plant, leaves, main stem nodes, main stem internodes, and lodged plants/accession were counted. Leaf area, plant height and width were determined also after 2 months of growth. Leaf area per plant was determined by measuring leaf circumference (cm). Plant height was determined by measuring the average height of plants from ground level using a meter stick (cm). Plant width was determined by measuring the average width of plants from the main stem to the branch tips (cm). Four months after seeding (MAS), visual evaluations were conducted to determine whether or not apical dominance had been broken. The following criteria were used for apical dominance: 0 = apical dominance is broken and difficult to determine the fate of the main branch because many primary lateral branches are growing from the center of the main sunn hemp branch (typical of early and intermediate maturing accessions); 1 = partial apical dominance because the main branch and apical meristems are visible with numerous lateral branches clustered slightly below the apical meristem which are greater than 0.30 m in length (typical of late maturing accessions); and 2 = main stem is clearly dominant with few or 0 lateral branches. If lateral branches are evident, they are less than 0.30 m in length (typical of late maturing accessions). Leaf area (cm), number of lateral branches, and whether or not there were open flowers and immature pods/sunn hemp branch. Sunn hemp maturity was also determined using the following criteria: early = the majority of lateral branches have mature pods (flowers may be visible because of the indeterminate nature of the plant; intermediate = the majority of lateral branches have open flowers and immature pods may be present; and late = the majority of lateral branches have either no flowers or buds with closed flowers or a few open flowers and no pods are present. Mature pods containing dry seed were harvested from the same 5 plants used to determine morphological traits when the pods reached the rattle stage and weekly thereafter.
Mature seed were threshed, counted, and weighed.

All morphological and reproductive data were subjected to an analysis of variance, using the PROC GLM procedure of SAS version 9.2 (SAS Inst., Cary, NC). Trait correlations were performed using PROC CORR in SAS statistical software (SAS Inst.). Principal component analysis and PC SAS procedure CLUSTER analysis were used for multivariate analysis of the data. PROC PRINCOMP was performed for all traits. Eigenvalues and the percentage of variances explained by each principal component were also determined. The similarity matrix was entered into PROC CLUSTER in SAS for cluster analysis with the unweighted paired group method using mathematic averages (UPGMA) by specifying the AVERAGE option for determining true genetic relationships among sunn hemp accessions.

Gainesville, Florida

Sixteen accessions of sunn hemp were evaluated in 2008 and 2009 at Rosie’s Organic Farm in
Gainesville, Florida to assess their vegetative and reproductive characteristics and potential for seed production in Florida (Table 1). The experimental design was a split plot with planting dates (May, June, and July) assigned to the main plots, which were arranged in a randomized complete block design with four replications. The sixteen accessions were randomly assigned to the subplots. Data were collected on plant height, leaf area, number of leaves, number of branches, plant weights, days to first open flower, and seed production.

In Gainesville data analyses were performed using SAS statistical software, versions 9.0 and 9.2 (Cary, NC). Analysis of variance was performed using the MIXED procedure and least square means were compared using the DIFF option. Square root transformations were performed on quantitative data prior to analysis. When considering the proportions of lateral branches that flowered or produced pods, and those that both flowered and had pods, PROC GLIMMIX was employed to allow for a logistic model. For each of the variables, the model under consideration included fixed effects terms for the accession, year, and planting date. An accession by planting date interaction term was also used. A random effects term for the planting block, nested within planting date, was integrated into the model. Pairwise means comparisons were conducted by using the step-down Student-Newman-Keuls method to control the false discovery rate (FDR) at the 0.05 level of significance. The FDR accounts for the number of false rejections that a test produces, so that the number of false rejections will be on average 5%. The MIXED and GLIMMIX procedures do not automatically perform this test so it was necessary to calculate the mean groupings by sequentially comparing the T-test statistics with the studentized range critical values.

In addition to univariate mean comparisons, the Hotelling’s T2 test of multivariate mean equality was also used to find similar groups of the accessions based on multiple variables. Vegetative and reproductive phenotypic parameters were assessed separately. For the vegetative parameters, the variables considered were: plant height, plant weight, leaf area, number of leaves, and number of primary branches. The reproductive phenotypic parameters were represented by the variables: seed weight, number of pods, number of immature pods, and immature pod weight. SAS code to perform the Hotelling’s test was written in PROC IML, using output from other SAS procedures including GLM and CORR. Since the Hotelling’s test was performed 120 times to compare each accession to all others, it was necessary to utilize a method to control the overall error. As before, a method was chosen to control the false discovery rate by comparing the ordered p-values of the individual tests to 0.05 divided by the ordered test number. One challenge was that for five of the sixteen accessions (accessions 2, 3, 7, 14, and 15), all or almost all of the plants failed to flower during the five-month experimental period. Such accessions had zero values for variables that included flowering and seed production data and were excluded from some statistical analyses.

Lajas, Puerto Rico

Field experiments in 2008 and 2009 at the Agricultural Experiment Station of the University of Puerto Rico-Mayaguez in Lajas (18o 03’ 07”N, 67o 03’ 35’’). Sixteen accessions of C. juncea supplied by the USDA ARS PGRCU were germinated in the greenhouse in early May, June, and July of 2008 and 2009. Three weeks after planting, the seedlings were transplanted to the field in four complete randomized blocks. Plots of each accession contained two rows of 8 plants (16 plants/plot), except in cases of low germination, where the plants were divided and planted in equal numbers in the four blocks. Plants were grown organically. To encourage root nodulation, seeds were inoculated with Rhizobium (Nitragen EL®, EMD Crop BioScience, Milwaukee, WI). To aid in the establishment of the sunn hemp transplants, we provided overhead irrigation and one weeding early in the season (in 2008) or drip irrigation and plastic mulching (in 2009). Azadirachtin (Gowan, Yuma, AZ) and the insecticide/miticide Ecotrol® (EcoSMART Technologies, Inc., Franklin, TN) were used to suppress leaf-eating beetles (Ceratoma spp and Diabrotica balteata) and the pod borers Utetheisa bella L. and Utetheisa ornatrix L.

Plant height, plant diameter, branching, and leaf area were determined 2 and 4 months after planting. Days to flowering, seed yield and biomass yield were assessed. Accessions were considered to have flowered when there were flowers in 50% of the plants in a plot. Potential pollinators visiting the flowers were observed and collected. The data collected were submitted to analysis of variance for a split-split-plot design (planting time x accession x time of evaluation).

Data were analyzed using Infostat to identify the potential of the accessions for biomass and seed production in Puerto Rico. Data from accessions which had poor germination or from those which failed to thrive in the field, resulting in insufficient or inaccurate samples, were removed from the data set before analysis. These removed accessions were: PI 295851, PI 337080, PI 346297, and PI 561720. Additionally, the accessions PI 234771 and PI 248491 (in 2008), and PI 468956 and PI 652939 (in 2009) were taken out of the data set during one year of the experiment.

Objective 2. Identify bee species visiting sunn hemp flowers to determine which are effective pollinators and quantify their visits

Observations of Sunn Hemp Bee Pollinators

Field visits were made to Rosie’s Organic farm in summer 2008 and 2009 to identify bee species pollinating sunn hemp flowers. Bees were identified that visited sunn hemp and interesting behavior was observed, but the small number of bees precluded the quantitative studies that had been initially proposed.

Artificial Pollination Experiments

Fertilization of sunn hemp has been described to be self-incompatible (Kundu, 1964; Free, 1993). Other reports claim that the plant becomes self-compatible after mechanical stimulation, by brushing pollen on the stigma (Howard, 1919 et al. cf Free, 1993). We conducted experiments to verify these reports. Pollinators were excluded from flowers with mesh bags, and the plants were divided into four groups: 1. Control, no treatment. 2. Keel pulled down so that the style was pushed out and then allowed to retract. 3. Keel pulled down so that the style was pushed out, the flower’s pollen brushed on the stigma, and then the style was allowed to retract. 4. Same as treatment 3, except a mixture of pollen from other plants was brushed on the stigma. We also tested the possibility of cryptic self-incompatibility, where the pollen from the late opening anthers might be compatible but not the pollen from the early opening anthers. Pollen from the oldest flower on a raceme was used to hand-pollinate the remaining more distal flowers. Additionally, crosses among plants within accessions, using mixtures of pollen from several plants, were made.

Objective 3. Assess the phenotypic variability of flowering and characterize the sensitivity to environmental factors, such as photoperiod and temperature

The induction of flowering in plants is controlled by a number of genes that have been previously identified and characterized. In plant species that have been studied, the photoperiod pathway consists of a set of photosensory receptors that monitors the ambient light environment and transduces signals to an internal oscillator that conditions day/night responses. This circuit integrates with other inputs to regulate the accumulation of CONSTANS (CO) a protein central to the photoperiod pathway. CO is a transcriptional regulator that is controlled on many levels (Valverde et al., 2004). This protein, when conditions are appropriate, regulates the accumulation of downstream proteins (such as FLOWERING LOCUS T (FT), LEAFY (LFY), and SUPPRESSOR OF CO OVEREXPRESSION 1 (SOC1) that ultimately remodel the meristem—leading a transition from vegetative to floral habits.

3a) Tests of photoperiodic gene expression in Crotalaria using heterologous probes

The conservation of the regulator genes (described above) between species allows use of existing molecular tools to monitor gene expression in species where such genes have not been yet defined. For instance, the Arabidopsis CO gene has been used to study CO mRNA accumulation during heat stress in poinsettia (R. Schnelle, K. Folta, unpublished). The same resources, already present in the lab, were used to monitor mRNA accumulation trends in Crotalaria to determine how this species fits, or perhaps does not fit, with accepted paradigms. Plants from the ‘Tropic Sun’ cultivar were grown in chambers under long day and short day conditions at a constant temperature to isolate the effects of photoperiod. Leaf tissue was isolated from mature plants and mRNA was prepared using standard protocols and analyzed by northern blotting. CO transcript accumulation was monitored using the Arabidopsis CO probe. These studies were conducted to determine if Crotalaria contains genes that have a central regulatory tendency similar to other previously studied species.

3b) Isolation of photoperiodic regulators

Using polymerase chain reaction (PCR) with degenerate primers designed against photoperiodic regulatory genes, we will next amplify these targets from Crotalaria mRNA that has been reverse transcribed. The amplification products will be sequenced. Crotalaria-specific primers will then be designed and used to amplify DNA products from the 16 accessions and ‘Tropic Sun’. These will be cloned and sequenced. These activities will define the DNA sequence of critical floral regulators in the genus. More importantly, they will allow sequence variations to be correlated with specific flowering habits observed in the previously described experiment (see 3a) and objective 1 of this proposal.

Objective 4. Investigate how cutting sunn hemp to break apical dominance affects weed suppression and compare the allelopathic potential of the accessions

4a) Effect of breaking apical dominance

This study was conducted at the Plant Science Research and Education Unit (PSREU) in Citra, Florida and at similar research farm facilities in Griffin, Georgia and Lajas, Puerto Rico. The effect of breaking apical dominance on branching, weed suppression, flowering, and seed production was assessed in comparison with a non-treated control. A commercially available cultivar obtained from Kauffman Seed Inc., Haven, KS (imported from South Africa) was grown at a seed production density (10 lb/acre), a cover cropping density (40 lb/acre), and an intermediate density (25 lb/acre). In Georgia and Florida, seeds were planted in May in both 2008 and 2009. In Puerto Rico, seeds were planted on June 19 in 2008 and on April 22, 2009. Plot size was 12 ft wide by 25 ft long in Gainesville, with an inter-row spacing of 18 inches. Intra-row spacing varied with planting density. The growing points of plants at each density were either left uncut (control) or cut at three, four, and five weeks after planting (at 5, 6, and 7 weeks after planting in Georgia only) to break apical dominance. This 3×4 factorial of three sunn hemp densities and four cutting times was arranged in a randomized complete block design with four replications.

Weeds were identified and counted using randomly placed quadrats at eight and 12 weeks after planting. In Gainesville, photosynthetically active radiation penetrating the canopy and nondestructive leaf area index will be measured. Other data collected included: plant height, number of primary and secondary branches, shoot biomass, number of flowers, number of pods, and seed yield. Pods were harvested at rattle stage.

4b) Comparison of the allelopathic potential of the accessions

Sunn hemp seeds have been demonstrated to contain the phytotoxic constituent 5-hydroxynorleucine, which inhibits lettuce seed germination. In Gainesville, bioassays of the phytotoxic effects of aqueous extracts from leaves, stems, and roots of 14 sunn hemp accessions on lettuce seed germination were conducted using a procedure modified from Adler and Chase (2007). Thin layer chromatography and high performance liquid chromatography were used to detect and quantify 5-hydroxynorleucine.

Objective 5. Determine how cultural practices for sunn hemp seed production influence nitrogen accumulation, decomposition, and plant available soil nitrogen

Potential for Sunn Hemp (Crotalaria juncea L.) to Utilize Soil Potassium

In organic production systems, efficient cycling of nutrients is critical to minimizing the high costs of compliant fertilizers. Potassium (K) is absorbed in plants in larger amounts than any other nutrient except nitrogen, yet it generally receives less attention than nitrogen and phosphorus in many crop production systems. Cover crop species with the capacity to uptake K that could be available to subsequent income-producing crops would be beneficial to farmers. To determine the capacity of sunn hemp K uptake and the influence of K on above-ground biomass production, three rates of potassium fertilizer [45 kg/ha (LOW), 90 kg/ha (MID) and 179 kg/ha (HIGH)] as potassium magnesium sulfate (22% K2O) were compared to a control of 0 K (ZERO), and treatments were randomized and replicated three times. Sunn hemp was seeded to 28 kg/ha on 15 May, 2008 at the Plant Science Research and Education Unit in Citra, FL.

Residue Quality and Decomposition Rate of Terminated Sunn Hemp Grown for Seed

A 2-year field trial was conducted in the Organic Unit at the UF-IFAS Plant Science Research and Education Unit in Citra, FL to ascertain if managing sunn hemp for seed production would provide additional N benefits. A short-day cultivar imported from South Africa by Kauffman Seed, Inc. was seeded to 11, 28, and 45 kg/ha. Cutting the main stem to break apical dominance was performed as described for Objective 4a. Treatments were arranged in a randomized complete block design and replicated four times. One half of each plot area was terminated at midbloom and the second half was terminated at seed maturity. Litter bags were prepared from sunn hemp residue and removed weekly for six weeks following termination by mowing. Plant height, biomass, tissue carbon and nitrogen, soil nitrogen, soil temperature and moisture, and crop carbon and nitrogen at termination and during decomposition comprise the scope of data collection.

Objective 6. Evaluate the economic costs and benefits of sunn hemp domestic seed, cover, and fodder crops

Partial budgets were written for a summer squash crop following five summer fallow treatments in Florida. Three leguminous cover crops (sunn hemp, velvet bean and cowpea), one non-leguminous cover crop (sorghum sudangrass), and a tillage treatment to manage weeds were compared. The costs of the cover crop as well as the ecosystem services that these cover crops provide were taken into consideration using data from published literature.

Research results and discussion:

Objective 1. Evaluate the effect of different geographical locations on biomass accumulation, flowering and seed yield of the USDA’s sunn hemp germplasm collection

Griffin, Georgia

Planting Dates: Sunn hemp planting dates produced significant differences in all morphological traits except for the number of open flowers at 2 MAS. Only planting dates 1 and 2 were analyzed due to late season precipitation which interfered with optimum data characterizations in planting date 3. Planting date 1 produced significantly more secondary branches, leaf number, leaf area, nodes, internodes, and larger plants than planting date 2 at two MAS. Sunn hemp accessions in planting date 1 produced significantly higher leaf area and more primary lateral branches than planting date 2 at 4 MAS. Sunn hemp accessions in planting date 1 also produced significantly more seed weighing more than those produced in planting date 2. In general, planting date 1 produced superior biomass when compared to planting date 2.

Sunn Hemp Accessions: Among the sunn hemp accessions tested, significant differences (P < 0.01) were found for all morphological traits evaluated at 2 MAS. The accession, 426626 from Pakistan produced the widest plants (43 cm), most leaves (2653), nodes (1474) and internodes (737) than any other sunn hemp accession and produced taller plants (119 cm) than many accessions at 2 MAS. However, seven sunn hemp accessions produced greater average branching (78) and number of leaves (2009) including PI 207657 (Sri Lanka), PI 250485 (India), PI 250486 (India), PI 250487 (India), PI 314239 (former Soviet Union), PI 322377 (Brazil), and PI 391567 (S. Africa) than most of the other sunn hemp accessions. Only PI 391567 produced a leaf area exceeding 40 cm. In fact, all but PI 322377 produced an average of 1015 nodes, 427 internodes, with plant heights and widths averaging 112 and 36 cm, respectively. The accession, PI 561729 produced more leaf area (58.3 cm) than all other sunn hemp accessions, but it also produced the fewest primary lateral branches (15), nodes (438), and internodes (219). Only PI 314239 was earlier maturing than all other accessions at 4 MAS with a leaf area of 14.06 cm and about 4 primary lateral branches greater than 0.91 m (Table 5). Ten accessions including PI 337080 (Brazil), PI 391567, PI 652939 from the U.S.A., PI 346297 (India), PI 207657, PI 426626, PI 322377, PI 250485, PI 250486, and PI 250487 produced intermediate maturity because the majority of the lateral branches had open flowers and immature pods 4 MAS. These accessions produced leaf areas and primary lateral branches greater than 0.91 m averaging 17 cm and 4, respectively. All other accessions produced late maturing plants with leaf areas and lateral branches greater than 0.91 m averaging 31 cm and 4, respectively. Overall, the majority of the sunn hemp accessions produced open flowers/branch, exhibited apical dominance because a visible main branch, apical meristem, and several lateral branches (greater than 0.91 m) were clustered slightly below the apical meristem, and all but PI 248491 from Brazil produced immature pods.

Significant differences were observed for all reproductive traits also. The accession, PI 314239 produced the most and heaviest seeds (3737 and 137 g, respectively) followed closely by PI 391567 (3351 seeds weighing 110 g) and PI 322377 (2785 seeds weighing 95.7 g). The accessions, PI 207657, PI 250485, PI 250486, PI 250487, PI 314239, and PI 426626 produced the next highest yield averaging 2113 seeds weighing 79 g. Low seed producing accessions included PI 337080, PI 346297, PI 652939, PI 234771, and PI 561720 averaging 463 seeds weighing 17 g/accession. Only PI 248491, PI 295851, and PI 468956 did not produce seed.

Principal Component Analysis: Principal component analysis showed that the first principal component accounted for 77% of the total variation. When principal components 2 and 3 were added progressively, the cumulative amount of variation accounted for was 87 and 92%, respectively. The first principal component was most correlated with primary lateral and secondary branches, leaf number, and total seed weight, while the second principal component, accounting for 10% of the variation, was primarily related to plant height, number of nodes, internodes, and open flowers. The third principal component explained 5% of the variation and was composed mainly of plant height, width, total seed number and weight. Primary lateral branches was significantly correlated with secondary branches (r-square = -0.19**), number of flowers (r-square = 0.22**), plant width (r-square = -0.17*), leaf area (r-square = -0.19**), node (r-square = 0.40***) and internode (r-square = 0.40***) numbers. Secondary branches were significantly correlated with all morphological and reproductive traits. Leaf number and total seed weight was significantly correlated with all traits as well, except primary lateral branches. Plant height was significantly correlated to all traits except primary lateral branches and leaf area. Number of nodes and internodes significantly correlated with all traits but number of flowers and plant width.

Average linkage cluster analysis grouped the original 16 accessions into well-defined phenotypes with 4 distinct groups and one outlier based on total biomass which included totaled morphological data and number of seeds produced (Group 1 – 6 accessions with low biomass and 0 – 660 seeds, Group 2 – 2 accessions consisting of intermediate biomass producers and 728 – 806 seeds, Group 3 – 2 accessions with high biomass and 3351 – 3737 seeds, Group 4 – 5 accessions with high biomass and 1734 – 2785 seeds, and the outlier – 1 accession producing very high biomass but only 1847 seeds). Accessions clustered in Group 1 are more closely related genetically than those in Groups 2, 3, 4, and the outlier. Using the distance values indicated in Figure 1, the groupings at any similarity level can be identified. For example, PI 248491 and PI 561720, which originated from Brazil have a phenotypic distance index of 0.0579, which indicates their close morphological and reproductive similarities.

Gainesville, Florida

Field observations suggested that accessions could be separated into two distinct groups based on size and daylength sensitivity. Analysis of the vegetative and reproductive data provided evidence for one group of short-day accessions and one group of day-neutral accessions. The short-day accessions were taller, with higher shoot biomass, later branching and flowering than the day-neutral accessions, and little or no seed. The day-neutral accessions flowered early and produced viable seed in summer. Of these accessions, PI 314239 and PI 322377 produced the most seed, demonstrating potential for use for seed production in Florida. These accessions had the lowest shoot biomass, making them less desirable as a cover crop. Future work will focus on developing day-neutral cultivars of sunn hemp that retain the cover crop attributes of the commercially available short-day sunn hemp varieties, but are capable of producing seed in Florida.

Puerto Rico

The accessions represented a wide variety of morphological and phenological attributes. The accessions identified for producing high biomass were characterized as tall and narrow, the most prominent of which were ‘Nigeria’ (PI 234771), ‘São Paulo’ (PI 295851),and ‘Tropic Sun’ (PI 468956); however, these same accessions were also identified for their photoperiodic sensitivity, resulting in a greater number of days to flowering when exposed to longer day lengths. The accessions ‘Guizo de Cascavel’ (PI 248491), ‘IAC-1’ (PI 561720), and ‘Texas 374’ (PI 652939) produced moderately high levels of biomass and reflected similar parallels in respect to flowering and photoperiod. A relationship was also noted between these accessions and poor seed production, often resulting in little to no seed yield. Consequently, while these accessions would be ideal for use as a green manure, owing to their biomass production capacities, the extent of this study showed them to lack the potential for the development of a strong C. juncea seed production industry. The exceptions were ‘Texas 374’ and ‘Nigeria’, which produced average seed yields comparable to the other accessions when exposed to the longer photoperiods which were present during the June and July plantings in 2008. In the case of ‘Nigeria’, this accession failed to produce sufficient plants for harvest during the 2009 trials. While ‘Texas 374’ produced a harvest in 2009, seed yields from each planting time were significantly lower than average. Nonetheless, the results of this study provide sufficient evidence for both the ‘Texas 374’ and ‘Nigeria’ accessions to merit further research.

Although the accession ‘T’ai-yang-ma’ (PI 391567) produced only average levels of biomass with respect to the other accessions studied, its low sensitivity to photoperiod and distinctive seed production characteristics make it another good candidate for future research. Interestingly, while the other accessions responded to the environmental stresses and pest pressures in the July planting of 2009 with reduced seed yields, ‘T’ai-yang-ma’ appeared to thrive, out performing all of the other accessions during the same trial. Additionally, this accession was one of the highest seed producers among all three trials in 2008.

Future studies should include trials of these three accessions (‘Texas 374’, ‘Nigeria’, and ‘T’ai-yang-ma’) during seasons with shorter photoperiods and under different environmental conditions than were defined in the parameters of this research. Furthermore, crosses between these accessions should be studied to assess the potential for developing a hybrid variety with the favorable characteristics of each of these accessions.

Objective 2. Identify bee species visiting sunn hemp flowers to determine which are effective pollinators and quantify their visits

Observations of Sunn Hemp Bee Pollinators

We identified bees visiting sunn hemp flowers and observed interesting behavior, but the small number of bees precluded the quantitative studies that had been proposed. Large Megachile and Xylocopa (large carpenter bees) species have been reported to visit sunn hemp and to be the effective pollinators (Purseglove, 1981; Nogueira-Couto et al., 1992; Free, 1993; Westerkamp, 1997; Etcheverry et al., 2003; Chaudhury et al.). Our observations agree with these earlier findings. The exotic Megachile sculpturalis, the giant resin bee, has the size, anatomy, and behavior to be a most effective pollinator. It was first recognized in Florida by one of us during this study and was a frequent visitor to our sunn hemp plots during June. At least one native Megachile species was seen, but visited too infrequently and was too small to be an effective pollinator. Honey bees were frequent nectar and pollen robbers, but were not effective pollinators.

Artificial Pollination Experiments

Of the four treatments used [(1) Control, no treatment, (2) Keel pulled down so that the style was pushed out and then allowed to retract, (3) Keel pulled down so that the style was pushed out, the flower’s pollen brushed on the stigma, and then the style was allowed to retract, (4) Same as treatment 3, except a mixture of pollen from other plants was brushed on the stigma] only treatment 4 resulted in seed pod formation. No evidence was found in support of cryptic self-incompatibility, where the pollen from the late opening anthers might be compatible but not the pollen from the early opening anthers. No pods were formed when pollen from the oldest flower on a raceme was used to hand-pollinate the remaining more distal flowers. Crosses among plants within accessions, using mixtures of pollen from several plants, were made, which resulted in seed pod formation in eight (89% average of number of flowers pollinated) out of nine accessions tested.

(4a) Investigate the effects of breaking apical dominance on weed suppression and seed yield


Sunn hemp cutting date had no significant effect on weed suppression in 2008, but significant differences for grass type weeds at 4, 8, and 12 WAP and for yellow nutsedge at 8 and 12 WAP did occur when compared to the weedy control in 2009. In comparison to the weedy control in 2009, all three seeding rates had lower grass dry weights at 4, 8, and 12 WAP. Yellow nutsedge dry weights were lower with all seeding rates than with the weedy control at 8 and 12 WAP.

Removing apical dominance by cutting below the meristematic region at all three cutting dates (5, 6, 7 WAP) had no effect on branching in 2008 and 2009. However, seeding rates affected sunn hemp branch numbers in both years. During 2008, the lowest sunn hemp seeding rate (10 lb/ac) produced plants with 1 to 2 more branches than the intermediate or highest seeding rates. In 2009, the lowest sunn hemp seeding rate produced plants with 2 more branches than the highest seeding rate. Low numbers of flowers were observed in both years and no pods were set.


In 2008, cutting sunn hemp to break apical dominance had no significant effect on weed biomass, but in 2009 cut plots had significantly less weed biomass than those left uncut. In 2008, the lowest seeding rate of 10 lb/ac had significantly higher weed biomass than the other seeding rates. In 2009, with the added weedy fallow, all seeding rates had significantly less weed biomass than the weedy fallow. By 12 WAP, the highest seeding rate of 40 lb/ac had the lowest weed biomass. Cutting did not significantly impact the amount of photosynthetically active radiation (PAR) measured below the canopy. Lower seeding rates had higher percentages of PAR penetrating the canopy at 3 and 6 WAP, but by 8 WAP all plots were well-established and there were no significant differences among the treatments. In both 2008 and 2009 cutting induced branching. In 2009 removing the apical meristem at the later dates increased the number of branches per plant. Plants in the lower seeding densities in both years had significantly more branches than plants in the cover crop seeding rate treatment (40 lb/ac). In 2008 and 2009 cutting had no significant effect on flower production. In both years, the lowest seed rate (10 lb/ac) had significantly more flowers than the other two seeding rates. Although flowering was profuse, seed set was minimal. This may have been due to the lack of an effective naturally-occurring pollinator in fall. Commercially available bumblebee hives were provided in an effort to promote pollination. Although bumblebees were observed visiting sunn hemp flowers they did not result in pod set.

Puerto Rico

Puerto Rico was the only location at which seeds were produced. In general, the difference in time of planting between the two years of the experiment played a significant role in the production of biomass and seed. Seed production was considerably lower with the April planting date in 2009 than with the June planting in 2008. The commercially available variety used in this study was a short day variety. In 2009, sunn hemp would have been exposed to approximately two months of increasing daylength before the summer solstice, when the photoperiod would have begun to decrease. In contrast, the 2008 trial was planted one day just prior to the summer solstice.

The optimal treatment combination with respect to both biomass and seed production proved to be a planting density of 25 lb/ac with no apical cut. Although planting sunn hemp at 40 lb/ac and applying apical cutting treatment at 5 WAP resulted in similar seed yields, this treatment combination would incur a higher seed cost and as well as a labor cost for the apical cutting treatment. Similarly, the assessment of biomass production did not reflect any effect of apical cutting and a planting density of 40 lb/ac did not produce a significantly greater amount of biomass than at 25 lb/ac. In the absence of cutting, the 10 lb/ac planting density produced significantly less biomass and seed than the 25 lb/ac density. Furthermore, while apical cutting did have an effect on height, these differences were not reflected in biomass and seed production, nor were they significant in weed suppression. Within the parameters of this study, planting sunn hemp at 25 lb/ac proved to be both productive and cost efficient.

(4b) Comparison of the allelopathic potential of the sunn hemp accessions

A method was developed to rapidly screen sunn hemp tissue for water-soluble allelochemicals using lettuce bioassays. Extracts from sunn hemp leaves, stems and roots contained high levels of compounds inhibitory to seedling growth of the test species. When sunn hemp leaf tissue was screened, high levels of allelochemicals were present in all fourteen accessions collected from seven countries and PI 250486 had the greatest potential. Inhibitory potential of water eluates was not reduced if the eluates were treated with acid and/or heated to boiling. The inhibitory compound(s) bound to weak anion exchange resin and could be eluted using acid. Thin layer chromatography analyses suggested that fractions with high inhibitory potential contained compounds that could be stained with ninhydrin and are UV positive. Leaf eluates have been submitted for analysis by High Performance Liquid Chromatography combined with mass spectrometry of leaf eluates to determine and quantify a compound with characteristics of hydroxynorleucine, the previously reported amino acid derivative with allelochemical properties.

Our results suggest that allelochemicals in sunn hemp are evolutionarily important in this species as they are found in all genotypes screened. Chemicals with a high degree of allelopathic activity could be used as leads for designing novel herbicides. Phytochemical analytical methods developed in this work will be valuable to formulate natural product extracts that may be useful for organic control of weeds.

Objective 5. Determine how cultural practices for sunn hemp seed production influence nitrogen accumulation, decomposition, and plant available soil nitrogen

Potential for Sunn Hemp (Crotalaria juncea L.) to Utilize Soil Potassium

The results of this study were reported in Treadwell et al. (2009). They found that K content in sunn hemp stem tip tissue (top 15 cm) increased linearly with rate (y = 0.004x + 1.98; r-square = 0.12). However, K content also declined linearly with time (y = –0.44x + 3.57; r-square = 0.73). Total K in aboveground dry plant tissue increased with increasing K rate (y = 0.005x + 1.92; r-square = 0.58) at 6 WAP. Potassium accumulation in aboveground tissues By 6 WAP sunn hemp already contained 68 kg/ha of K, which indicated that in addition to serving as a good N green manure, it could serve as a catch crop for K making the element available for a subsequent crop.

Residue Quality and Decomposition Rate of Terminated Sunn Hemp Grown for Seed

Biomass accumulation in 2008 increased from 240 kg/ha at 4 weeks after planting (WAP) to 25243 kg/ha at 20 WAP. In 2009 biomass was 214 kg/ha and 15, 061 kg/ha by 20 WAP. Seeding rate and cutting to remove apical dominance had little influence on residue quality at termination, but did influence biomass production and biomass C:N early in the season. Decomposition of sunn hemp residue was negligible over a 12-week period. Additions of compost or manure may help to increase the rate of decomposition of sunn hemp residue.

Participation Summary

Educational & Outreach Activities

Participation Summary

Education/outreach description:


Cho, A.H. 2010. Investigation of sunn hemp as a cover crop and a seed crop in Florida. MS Thesis, University of Florida. Gainesville, Florida.

Halbrendt, J.C. 2010. Here comes the sunn: A comprehensive analysis of the cover crop sunn hemp (Crotalaria juncea L.) and its potential for seed production and weed control in Puerto Rican agricultural systems. M.S. Thesis, University of Puerto Rico, Mayaguez.

Conferences posters and presentations

Cho, A. H., C. A. Chase, D. D. Treadwell, and R. L. Koenig. 2009. Apical dominance and planting density effects on weed suppression by sunn hemp. WSSA Abstracts, #23, 1 page.

Cho, A. H., C. A. Chase, D. D. Treadwell, and R. L. Koenig. 2009. Investigating sunn hemp seed production to promote cover crop adoption. Proceedings Florida Weed Science Society, p. 7.

Cho, A.H., C.A. Chase, D.D. Treadwell and R.L. Koenig. 2010. Effects of seeding rate and apical dominance on flowering and weed suppression in sunn hemp. Proceedings Southern Weed Science Society.

Cho, A.H., C.A. Chase, R.L. Koenig, and D.D. Treadwell. 2010. Evaluation of Sixteen Accessions of Sunn Hemp (Crotalaria juncea) for Seed Production Potential in Florida. Abstracts of the ASHS Southern Region 70th Annual Meeting. HortScience 45: 486-518.

Cho, A.H., C.A. Chase, D.D. Treadwell, and R.L. Koenig. 2010. Evaluation of Phenotypic Characteristics of Sixteen Accessions of Sunn Hemp in Florida. HortScience 45(8)(Supplement)—2010 ASHS Annual Conference. S240-S241.

Cho, A.H., A.W. Hodges, and C.A. Chase. 2011. Economic analysis of summer cover crop fallows for organic weed management in Florida. Oral Presentation at the Southern Weed Science Society Meeting. January 24-26, San Juan, Puerto Rico.

Cho, A.H., A.W. Hodges, and C.A. Chase. 2011. Quantifying the costs and benefits of cover crop use for weed management. Poster presentation at the Weed Science Society of America. February, Portland, Oregon.

Halbrendt, J., Morales-Payan, J.P., Martinez Garrastazu, S., Brunner, B. and. Flores-Lopez, L. 2009. Effect of planting density and apical cutting of the cover crop Crotalaria juncea on weed suppression in southern Puerto Rico. Abstr. Society for Agricultural Sciences of Puerto Rico (SOPCA) 35:D5.

Halbrendt, J., J. P. Morales-Payan, S. Martinez-Garrastazu, B. Brunner, L. Flores-Lopez, M. Vega-Almodovar, and J. Toro. 2008. Onset of flowering of 16 accessions of the potential green manure and cover crop Crotalaria juncea grown in Puerto Rico. Abstr. Soc. for Agric. Sci. of Puerto Rico (SOPCA) 34:51.

Halbrendt, J., J.P. Morales-Payan, S. Martinez Garrastazu, B. Brunner, L. Flores and J. Toro. 2010. Crotalaria juncea green manure biomass and seed yield as affected by planting density and apical cutting in an organic system. 46th Annual Meeting Caribbean Food Crops Society.

Halbrendt, J., J.P. Morales-Payan, S. Martinez Garrastazu, B. Brunner, and L. Flores. 2010. Pests in organically managed Crotalaria juncea in southwestern Puerto Rico. 46th Annual Meeting Caribbean Food Crops Society.

Halbrendt, J., J. Pablo Morales-Payán, S. Martinez Garrastazú, B. Brunner, L. Flores y J. Toro. 2011. Plagas en sistemas de producción ecológica de la planta de cobertura Crotalaria juncea en Puerto Rico. Poster presented at the 4th Agroecology Symposium of the University of Puerto-Rico-Utuado.

Morris, J.B., Chase, C.A., Cho, A.H., Koenig, R.L., Morales-Payan, J.P. 2010. Earliness, morphological, and reproductive variation among 16 sunn hemp (Crotalaria juncea L.) accessions in Griffin, GA. Association for the Advancement of Industrial Crops Conference. Ft. Collins, CO. Sept. 18-22.

Morris, B., C.A. Chase, A.H. Cho, R. Koenig, J.P. Morales-Payan. 2010. Principal component analysis for morphological, seed reproductive, and phenology traits in 16 sunn hemp (Crotalaria
juncea L.) accessions. HortScience 45(8) (Supplement)—2010 ASHS Annual Conference. S182 (Poster).

Treadwell, D.D., C. Chase, A. Cho, M. Alligood, and J. Elsakr. 2009. Potential for sunn hemp (Crotalaria juncea L.) to utilize soil potassium. Florida State Horticultural Society Abstracts

Treadwell, D.D., C.A. Chase, A.H. Cho, M. Alligood. 2010. Residue Quality and Decomposition Rate of Terminated Sunn Hemp Grown for Seed. HortScience 45(8) (Supplement)—2010 ASHS Annual Conference. S138 (paper). Video available at:


Brunner, B., S. Martinez, L. Flores, and P. Morales. Crotalaria. Hoja Informativa. Proyecto Agricultura Organica Z-NRCS-007/Z-208. Departamento de Cultivos y Ciencias Agroambientales. Estacion Experimental Agricola de Lajas.


Cho, A.H., C.A. Chase, D.D. Treadwell, R.L. Koenig, J.B. Morris, and J.P. Morales-Payan. 2011. Apical Dominance and Planting Density Effects on Weed Suppression by Sunn Hemp in Florida. Weed Technology (In preparation)

Morris, J.B., C.A. Chase, D.D. Treadwell, R.L. Koenig, A.H. Cho, J.P. Morales-Payan, and T. Murphy. 2011. Effect of Sunn Hemp (Crotalaria juncea L.) Cutting Date and Planting Density on Weed Suppression in Georgia, USA. Weed Technology (In review)

Cho, A., Hodges, A.W. Hodges, and C.A. Chase. 2011. Economic Analysis of Summer Fallows for Organic Nutrient and Weed Management in Florida. HortTechnology (In Preparation)

Treadwell, D.D., C. Chase, M. Alligood, and J. Elsakr. 2009. Potential for sunn hemp (Crotalaria juncea L.) to utilize soil potassium. Proc. Florida State Horticultural Society Meeting 122:243-246.


A workshop was held on May 27, 2010 to inform stakeholders and extension faculty about the progress of the work and to discuss key concepts and ideas originating from the work and to obtain feedback from the participants.

Questions addressed:

1. Does it make sense to use sunn hemp for seed production vs cover crop?
2. What are the benefits and constraints associated with pollination when growing sunn hemp?
3. What are the limitations in chopping /cutting sunn hemp when terminating either a cover crop or a seed crop?
4. What are the pros and cons of having different planting dates for sunn hemp?

Workshop Conclusions

Seed: Seed production is challenging. Accessions that will produce seed on the mainland will require a compromise on biomass since plants tend to be of smaller stature. Figure out farmers who want to produce sunn hemp for seed and those that would like to produce it for biomass.

Pollination: Honey bees visit sunn hemp flowers but are essentially pollen robbers since they are able to collect pollen without pollinating the flowers. Good pollination requires big, heavy bees. The “perfect” bee is heavy and long so that the next flower they visit is pollinated. Future studies should consider whether encouraging native bees such as the carpenter bee (Xylocopa spp.) or the giant resin bee Megachile sculpturalis is feasible and appropriate since the former may be considered by some to be a pest and the latter is an exotic species.

Cutting for Biomass/Mulch: For green manure, sunn hemp should be chopped and turned under at midbloom. Seed crop residue was very fibrous and was difficult to chop and incorporate. The very slow decomposition of seed crop residue means that it may serve as a good mulch, but not be good green manure.

Planting Dates: Accessions that are less sensitive to daylength will set seed in Florida and Georgia. Earlier planting dates will result in greater biomass. However, other considerations are that flowering and the presence of the effective pollinators must be synchronized and harvest of the pods should occur during a dry period since fungi can ruin the seed when dried pods are exposed to rainfall. Some short day cultivars produced seed in Puerto Rico, but planting date is critical.

Project Outcomes

Project outcomes:

Sunn hemp accessions in the USDA-ARS collection varied in their response to daylength. Accessions that were less sensitive to daylength flowered in summer and were more appropriate for seed production in Florida and Georgia. When planted in May, June or July flowering coincided with the presence of effective pollinators resulting in seed production. However, short-day accessions flowered in fall later than day-neutral accessions and little to no pod set occurred in Florida and Georgia. The location of Puerto Rico at a lower latitude than Florida and Georgia means than shorter daylengths occur during summer and some short-day accessions were better adapted to seed production there than on the mainland. Therefore, if the market demands the taller sunn hemp varieties with larger biomass production potential, seed production in the Caribbean is recommended. Seed production in Florida and Georgia would be feasible with accessions that are less sensitive to daylength, but these will yield shorter plants with less biomass. For all locations, ensuring that flowering and the occurrence of effective pollinators such as Megachile spp. and Xylocopa spp. are synchronized will be critical. Pod borers were a problem in Puerto Rico and will have to be managed.

Planting at a density of 25 lb/ac was as effective as 40 lb/ac at all locations for biomass production and weed suppression. A lower seeding rate will mean lower cost of establishment for a sunn hemp cover crop, which may boost adoption by growers.

Sunn hemp was demonstrated to be an effective catch crop for potassium. Residues from sunn hemp grown for seed were found to be very recalcitrant and did not decompose readily. Sunn hemp may, therefore, have the potential for use as organic mulch and in the sequestration of carbon.

Economic Analysis

Objective 6. Evaluate the economic costs and benefits of sunn hemp domestic seed, cover, and fodder crops

Economic analyses of five summer fallow treatments in Florida preceding a cash crop of summer squash were conducted. The five treatments were sunn hemp, velvet bean, cowpea cv. Iron Clay, sorghum sudangrass, and tillage. Costs were estimated for each summer fallow treatment, including the cost of seed, inoculant, implementation, management, and termination. Benefits were calculated in terms of contributions to the following cash crop of summer squash in the form of biologically fixed nitrogen and reduced weed pressure. Cho (2010) found that when economic benefits were assigned to nitrogen accumulation and weed suppression that the use of cover crops was a more affordable option than a clean fallow with tillage. Results showed that total production costs were minimized by cover crops, even though implementation costs were higher than for tillage.


Areas needing additional study

Further evaluation of promising accessions is recommended with the end result of developing foundation seed to facilitate commercial sunn hemp seed production. A better understanding of sunn hemp floral biology is also needed to facilitate sunn hemp breeding and improving pollination. Of interest, also, is whether pod set can be maximized by augmenting the populations of effective pollinators such as Xylocopa species. For the short-term, a day-neutral sunn hemp accession from the existing USDA collection can be used to develop a variety that can be made available to growers for commercial production of certified seed. However, for the longer term, an improved variety should be bred that combines the high biomass attributes of the short-day accessions with the ability to set seed prior to frost that is typical of the day-neutral accessions and possessing weed- and nematode-suppressive attributes. Residue left after harvesting seeds is recalcitrant and if it is to be incorporated it will require addition of compost or manure to increase the rate of decomposition. Other ways of utilizing the residue should be investigated such as the production of biochar.

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.