Contributions of legume cover crop root systems to soil carbon pools in organic systems using different termination strategies

Final Report for GS12-112

Project Type: Graduate Student
Funds awarded in 2012: $10,997.00
Projected End Date: 12/31/2013
Grant Recipient: NCSU
Region: Southern
State: North Carolina
Graduate Student:
Major Professor:
Expand All

Project Information


In the field study, legume root decomposition and N release were not affected by termination approach or soil inorganic N levels.  All roots decomposed similarly at the Goldsboro, NC site, but crimson clover roots decomposed fastest at the Kinston, NC site.  At both sites, hairy vetch roots released N fastest and crimson clover roots released N slowest.  In the incubation, legume fine roots decomposed and released N faster than coarse roots, but species and N addition did not affect N release.  Our results indicate that under the humid summer conditions of the Southeast, legume roots decompose and release N rapidly.


The purpose of this project was to investigate the decomposition and N release dynamics  of legume cover crop roots, and how this may be affected by cover crop management and root morphology.  In particular, we investigated how popular and novel termination methods interact with three winter annual leguminous cover crop species’ root morphologies to ultimately result in root decomposition and N release.  Soil contains more carbon (C) than plants, animals and the atmosphere combined. This Soil organic carbon (SOC) is a critical part of agricultural soils due to its ability to improve soil physical properties such as water holding capacity and aggregate stability, leading to deeper and more prolific root systems (Alcántara, 2011) and creating a positive feedback loop where larger diameter roots contribute further to increasing SOC. With the need to adapt to a changing climate, and potential payment for C storage looming on the horizon, farmers are particularly interested in learning more about management strategies that will help contribute to SOC sequestration. Organic systems can have a particularly important role in C storage due to the many C-based inputs utilized.  Winter annual legume cover crops, such as Austrian winter pea, crimson clover, and hairy vetch, are typically planted in late fall and terminated in spring before summer crops are planted.    Cover crop roots generally remain in the soil profile following termination, making them an important potential contributor to long-term SOC.

While most cover crop research has focused on shoot material, roots deserve further attention considering that C contributions from root biomass can be significantly large.  Shoots generally have lower C to nitrogen (C:N) ratios and quantities of recalcitrant materials than roots, leading to faster decomposition than root material (Bird et al. 2008, Wang et al. 2010), and suggesting that root-derived C may contribute greatly to overall soil C stored after a season of cover crop growth.  Puget and Drinkwater (2001) reported hairy vetch shoots to have a higher N content (4.5 vs. 2.8%), lower C:N ratio (9.7 vs. 11.9), and lower lignin concentration (5.2 vs. 17%) than roots, favoring more rapid shoot decomposition.  Buchanan and King (1993) report similar findings with crimson clover.  Combined, available data indicate that legume cover crop shoots may decompose more rapidly than roots (Fujii and Takeda, 2010). This observation has been verified in field studies demonstrating that 50-52% of root derived cover crop C can remain in the soil, compared to only 4-13% of the quickly-decomposed shoot C, at the end of the growing season  (Puget and Drinkwater, 2001; Kong and Six, 2010). Over the long term, partially-decomposed root material has been shown to be preferentially stored within aggregates, giving this material increased physical protection as root derived particulate organic matter (Alcantara, 2011). Further, the continuous nature of root carbon inputs from exudates and fine root turnover leads to a constant input of root carbon into the soil environment (Puget and Drinkwater 2001).

Leguminous cover crops are an important feature of organic systems and can provide significant quantities of plant residues that build up SOC and furnish essential plant nutrients, especially N, for subsequent crops.  In fact, the national organic standard singles out the role of cover crops in these systems, stating that growers should “Manage crop nutrients and soil fertility through rotations, cover crops, and the application of plant and animal materials” (NOP, 205.203 (b).  Investigating the decomposition and and N release of legume cover crop roots under different  management scenarios will provide producers with pertinent information to better utilize leguminous cover crops.

Decomposition rate is controlled by many factors, including soil N availability, and size of the decomposing material.  In N-limited (low soil N) systems, the amount of N present for microbes to use as a nutrient source during the decomposition process controls their rate of decomposition, with N shown to be a limiting factor in fine root decomposition for tree species (Lin et al. 2010).  Since most agricultural systems are N-limited, any increase of N to the soil via N-rich materials, such as incorporated legume shoots, will stimulate decomposition of available carbon substrates, including roots. In this project, we  conducted a field study utilizing two cover crop termination methods that are either commonly employed by organic growers (incorporating residue), or that are novel, with organic producers desiring further study (roller-crimping). These methods place N-rich shoot biomass at different locations in the soil profile. For example, incorporation may enhance decomposition due to increases in N from the shoots.  Roller-crimping leguminous cover crops, however, may slow decomposition since N is concentrated in a mulch on the surface rather than being in contact with soil decomposer microorganisms. We hypothesized that incorporated residues will increase root decomposition and N release rate compared to roller-crimped legumes due to an increase in N available to the decomposer microbes.

Root morphology, particularly diameter, has been shown to influence decomposition and nutrient release (Fujimaki et al. 2008). Since nutrient release from legume roots of different size classes is not reported in the literature, morphological characterization would be valuable . Further, the effect of soil N status on decomposition of different root sizes has also not been investigated. Thus, a second major objective of this project was to assess root morphology and conduct a laboratory-based decomposition study over an N gradient to provide information pertinent to producers seeking to predict how N fertility in their fields will affect long-term root C persistence.   We hypothesized that smaller roots would decompose and release N faster than larger roots with any increase in N inputs.

This research has generated data for evaluating the residence time of legume cover crop root biomass and root N release dynamics, which can give producers a better idea as to which legume cover crop roots are most effective at persisting in soil and releasing N.  Furthermore, this project’s emphasis on investigating the long-term residence time of legume cover crop root biomass and root N release dynamics in agricultural fields is in accordance with the SARE program objective of promoting stewardship of natural resources and maintaining the profitability of sustainable agriculture.

Project Objectives:

  1. Determine the effect of winter annual leguminous cover crop termination method (incorporation vs. roller-crimping) on soil carbon by altering available soil N at rooting depth.
  2. Determine how cover crop root size (diameter) affects decomposition across a soil N gradient created in a lab incubation study.


Materials and methods:

Objective 1: Root decomposition was investigated at sites in Kinston and Goldsboro, NC at the Center for Environmental Farming Systems (CEFS).  A randomized complete block design was utilized at both sites with a total of four blocks per site.  Experimental plots within blocks measure 20 x 50 ft. (6.09 x 15.24 m) at Kinston and 20 x 45 ft. (6.09 x 13.71 m) at Goldsboro.  Cover crop species included Austrian winter pea, crimson clover, and hairy vetch, while termination methods include incorporation (disking) and roller-crimping.   There were four replications for each treatment (cover crop x termination method).  Roots from Austrian winter pea, crimson clover, and hairy vetch were collected prior to termination at both Goldsboro and Kinston.  Roots were soaked, rinsed thoroughly, and air-dried on greenhouse benches for two days.  Subsamples of air dried material were used to measure moisture content, ash-free-dry-weight, lignin, C and N.   Clean, air-dried roots were placed in 9X9cm2 1-mm diameter nylon litter bags.  This study contained six litter bag collection periods: 2, 4, 6, 8, 12, and 16 weeks after burial.  At each collection date, mass loss was determined by combustion in a muffle furnace.  Carbon and N were measured and soil core samples were collected every two weeks to measure for N content at burial depth since we hypothesized that higher soil N status in disked plots would accelerate decomposition.


Objective 2:  Determine if legume cover crop root decomposition is affected by root particle size and soil inorganic N content in a controlled incubation in an environmental growth chamber.  Field collected crimson clover and hairy vetch roots were separated into fine (< 1 mm diameter) and large (> 1 mm diameter) fractions and incubated in soil collected from Goldsboro, NC study site with (150 kg/ha) and without N addition. Urea was used as N source.  Roots were collected 1, 2, 3, 4, 8, and 12 weeks after burial in cups placed in the growth chamber. Root decomposition and N release were determined in addition to soil inorganic N.   

Research results and discussion:

Field Study: Soil inorganic N dynamics and legume cover crop root decomposition

Disking and roller-crimping termination approaches resulted in different levels of soil inorganic N in legume cover crop treatment plots at both study sites.  In Goldsboro, disked treatment plots averaged across cover crop species had higher levels (P < 0.0001) of soil inorganic N than roller-crimped plots until week 12 of the study (Fig. 1).  Soil inorganic N peaked in both disked and roller-crimped plots 6 weeks after cover crop termination (~ 4 weeks after litterbag burial) in Goldsboro, which falls within the general time frame of peak soil inorganic N following legume cover crop termination by disking and roller-crimping in the Southeastern U.S (Varco et al., 1989; Parr et al., 2014).  At this peak in Goldsboro, soil inorganic N in disked plots was approximately 81% higher than in roller-crimped plots.  In Kinston, soil inorganic N was also higher (P < 0.0001) under disking compared to roller-crimping, but for a much shorter duration reaching its peak just 2 weeks after cover crop termination (at litterbag burial).  At this peak in Kinston, soil inorganic N was 50% higher in plots terminated by disking.  Soil inorganic N levels in legume cover crop treatment plots at both sites was also influenced by cover crop species.  In Goldsboro, soil inorganic N levels were generally highest in HV plots through week 8 (P < 0.0001) and lowest in CC plots through week 4 (P = 0.0004) when averaging across termination approaches (Fig. 2).  At its peak in Goldsboro, soil inorganic N in HV legume cover crop treatment plots was 27% and 74% higher than in AWP and CC legume cover crop treatment plots, respectively.  In Kinston, differences in soil inorganic N content in legume cover crop treatment plots were only observed at root burial (~ 2 weeks after cover crop termination) during which time CC legume cover crop treatment plots had lower N (P = 0.009) than AWP (Fig. 2).        
In the month following cover crop termination in Kinston, rainfall was approximately three times higher than it was in the month following termination in Goldsboro (Table 1), which may have accelerated cover crop N release and subsequent leaching in both disked and roller-crimped plots in Kinston.  High rates of NO3- leaching from legume cover crop residues have been observed in both disked and surface mulched systems where soil moisture content exceeds soil water holding capacity (Moller et al., 2009; Campiglia et al. 2011).  This leaching effect would explain the similar levels of soil inorganic N measured in legume cover crop treatment plots terminated by disking and roller-crimping after week 2 at the Kinston site.  In Goldsboro, total rainfall was relatively low compared to Kinston, and leaching of NO3- was likely not a major event, which would account for the sustained higher levels of soil inorganic N in disked plots compared to roller-crimped plots over a longer period of time in Goldsboro. 

Although disking and roller-crimping generated different levels of soil inorganic N at both sites, termination did not affect root decomposition at either Goldsboro (P = 0.79) or Kinston (P = 0.82) study sites (Fig. 3).  Thus, our hypothesis that legume cover crop root decomposition would proceed more rapidly when cover crops are terminated by disking rather than roller-crimping was not supported despite the higher levels of soil inorganic N measured in disked plots at multiple sampling dates at both sites.  Recous et al. (1995) observed that microbial decomposers in the soil do not necessarily partake in luxury consumption of N under elevated N conditions.  This finding, along with our observations, indicates that factors other than soil inorganic N played a more pivotal role in driving legume cover crop root decomposition. 

Averaging across termination methods, roots from all species decomposed at similar rates (P = 0.752) in Goldsboro, while CC roots decomposed faster (P = 0.004) than AWP and HV roots in Kinston (Fig. 4).  After 16 weeks, only 15% of initially buried CC root litter remained at the Kinston site.  Approximately 12% and 23% of initial root litter remained after 16 weeks at both Goldsboro and Kinston sites, respectively when averaging across all treatments.  These rates of root decomposition are higher than those reported by Buchannan and King (1993) for CC root decomposition over the same period of time at another North Carolina location.  Climatic factors, particularly temperature and soil moisture, are a major determinant of plant residue decomposition (Gijsman et al., 1997; Berg and McClaugherty, 2008), and may help explain this variability as total rainfall in the Buchannan and King (1993) study was lower than the rainfall total at both Goldsboro and Kinston sites.


At both study sites, root litter decomposition was most rapid early in the investigation at both sites with greater than 38% and 48% of initially buried AWP, CC, and HV root mass decomposing within two weeks after litterbag burial in Kinston and Goldsboro, respectively (Fig. 4).  Initial rapid rates of mass loss are expected in plant litter decomposition studies, and are most often attributed to microbial consumption of easily decomposable materials such as simple carbohydrates, hemicelluloses, and proteins (Berg et al., 1987; Gunnarsson and Marstorp, 2002).  Dodd and Mackay (2011) reported slower rates of decomposition (30% mass loss over 30 days) for white clover (Trifolium repens L.) roots in a New Zealand pasture system, which may be related to the more temperate climatic conditions in New Zealand compared to the Southeastern U.S.  

Legume Cover Crop Root N Release Dynamics

Legume cover crop termination approach did not affect the rate at which roots released N at either study site.  Roots followed similar trends in N release at both sites with HV roots releasing N fastest and CC roots releasing N slowest (Fig. 5).  Hairy vetch roots released N at a faster rate than CC roots in both Goldsboro (P ≤ 0.013) and Kinston (P ≤ 0.028) up to week 12.  Throughout the study, CC roots never released N at a faster rate than AWP and HV roots at either site.  In Goldsboro, AWP and HV roots released N at similar rates, while HV roots released N faster than AWP roots at weeks 2 and 8 in Kinston.  These observed trends in root N release are likely explained by the initial higher N content in HV and AWP roots compared to CC roots at burial (Table 2).  Other studies have reported that legume cover crop shoots with higher initial N content release N faster than shoots with lower initial N content (Luna-Orea et al., 1996; Lawson et al., 2012), and it is reasonable to expect that a similar trend would be observed for legume cover crop root N release. 

In legume cover crop-based corn production systems in the Southeastern U.S., corn can be planted at cover crop termination in roller-crimped systems (Parr et al. 2011; Parr et al. 2014), or immediately after seed bed establishment in disked systems.  From a nutrient management perspective, it is most critical that the legume cover crop releases a large enough portion of its N to meet corn demands during its period of rapid N consumption, beginning at the six leaf stage (Magdoff 1991).  In this study, CC and HV root N release rates closely mirrored the rates of CC and HV shoot N release described in previous studies (Wilson and Hargrove, 1986; Stute and Posner 1995), which was unexpected since legume cover crop shoots generally release nutrients faster than roots (Buchanan and King, 1993; Puget and Drinkwater, 2001).  Within two weeks after cover crop termination in legume cover crop treatment plots, HV roots released 45-50% of their N at both sites, while AWP and CC roots released 19-37% and 24-29.5%, respectively.  After 4 weeks, these figures rose to 62-65%, 49-57%, and 39% for HV, AWP, and CC, respectively.  

In a Pennsylvania field study, Puget and Drinkwater (2001) estimated that HV standing root biomass immediately before termination was approximately 900 kg biomass ha-1.  If this biomass estimate is applied to our study, it would indicate that HV roots contained approximately 25 and 23 kg N ha-1 at Goldsboro and Kinston sites, respectively, at root burial.  Using this 900 kg biomass ha-1 estimate, during the first month after HV root burial in this study, approximately 16 and 14 kg N ha-1 would have been released and made available for corn uptake during its period of peak N demand.  After 16 weeks, HV roots would have released approximately 22 and 16 kg N ha-1 at Goldsboro and Kinston sites, respectively.  These estimates of HV root-derived N contributions to a subsequent corn crop fall far short of meeting the 150 -175 kg N ha-1 required for optimal grain yield.  However, it is conceivable that HV root N contributions to subsequent corn crops may be higher than these estimated values. Puget and Drinkwater’s (2001) root biomass estimate of 900 kg ha-1was based on standing root biomass prior to termination, and did not account for root turnover earlier during the HV life cycle.  Nor did it include roots less than 0.5 mm diameter.  Legume cover crop fine roots turnover and release N rapidly compared to coarse roots, and it is essential that they are included in estimates of N release from legume cover crop roots.  

Incubation Study: Root morphology analysis

In the 16-week root morphology study, fine roots ranged from 79% to 81% of  total roots across species, but differences between species in terms of total root length, proportion coarse and fine roots, and total root surface area were not observed (P > 0.05) (Table 3).  

Root litter chemistry and decomposition dynamics

Only CC and HV root decomposition and N release were studied in the incubation due to insufficient quantity of AWP roots.  Initial root litter chemical characteristics for field-collected CC and HV coarse and fine root fractions used in the incubation are reported in Table 4.  The initial CC and HV C/N ratios determined in this study resemble values for CC and HV roots reported in other investigations (Puget and Drinkwater 2001; Gardner and Sarrantonio 2012) ranging from 12.4-17.9 and CC coarse roots having a higher (P > 0.05) C:N ratio than other root fractions.  The higher C/N ratios measured for coarse roots compared to fine roots have also been observed in other plant species (Fujimaki et al. 2008; Rasmussen et al. 2010).  Crimson clover and HV coarse roots also had higher (P > 0.05) C content (37.4-37.6%)   compared to CC and HV fine roots (28.5-30.6%). 

Soil inorganic N addition did not affect (P > 0.05) CC and HV coarse and fine root decomposition in this incubation.  Crimson clover and HV fine roots decomposed faster (P < 0.05) than both CC and HV coarse roots, but within coarse and fine root fractions, species decomposed at similar rates.   The general pattern of decomposition for coarse and fine root fractions was similar. However, fine roots decomposed faster than coarse roots for both species.  Total root mass loss was characterized by a sharp decline during the first week of the incubation as approximately 41% and 62% of coarse and fine roots, respectively, decomposed when averaged across N treatments.  This rapid rate of decomposition was followed by a three week period (weeks 1 to 4) in which significant root mass loss was not observed.  A gradual and increasing trend in root decomposition for all treatments was observed during the final 8 weeks of the study.  When averaged across N treatments, 9 to 14% of the initial HV and CC fine roots remained at the end of the incubation period, while 12 to 19% of HV and CC coarse roots remained (Fig. 6).  

Root N release and soil inorganic N fluctuation

Initially, the release of root N proceeded most rapidly for CC and HV fine roots.  During the first week of incubation fine roots released approximately 40 to 50% of their N, while coarse roots released only 11 to 30%.  These rapid early rates of N release are on par with those reported by Lindsey et al. (2013) for incubated weed residues.  However, coarse and fine root N release progressed similarly and more slowly during subsequent weeks; such that N release was not observed (P > 0.05) from weeks 1 to 4 before gradually increasing for the remainder of the incubation.  After 12 weeks, CC and HV coarse roots had released 66-77% of their N, while CC and HV fine roots released 77-82% of their N, respectively.         

Release of N among CC and HV coarse and fine root fractions corresponded to an increase (P < 0.05) in soil inorganic N levels for all treatments during the incubation period.  The largest gain in soil inorganic N occurred during the first two weeks of the incubation (Fig. 7), which coincided with the fastest rate of root N release.  After 2 weeks of incubation, soils containing coarse and fine root fractions increased in soil inorganic N content by approximately 21 and 29%, respectively.  Malpassi et al. (2000) observed a similar increase of 30% after 7 days in an oat (Avena sativa L. `Ogle') and rye (Secale cereale L. `Rymin') root incubation study.  At the conclusion of this 12-week incubation, there was a net increase in soil inorganic N of approximately 39 and 29% for coarse and fine roots.  Net immobilization of N was not observed for any treatments. 

Participation Summary

Educational & Outreach Activities

Participation Summary:

Education/outreach description:

A manuscript for the field study portion of this research has been submitted to Renewable Agriculture and Food Systems, while a manuscript for the incubation study was submitted to Plant and Soil. We are still awaiting responses from both journals and will be sure to inform SARE once this research has been accepted for publication.

Project Outcomes

Project outcomes:

There is strong interest in reducing atmospheric levels of CO2 to mitigate the threat of global climate change, and having a more accurate assessment of the C sequestration dynamics associated with cover cropping systems is critical to achieving this goal.  There is the possibility that future legislation may create “carbon markets” in which farmers can create C credits for adopting certain conservation agriculture principles within the context of a cap-and-trade system57.  One component of such a system may entail quantifying the amount of C sequestered by specific cropping systems.  Under such a scenario, it would be in the interest of farmers cultivating legume cover crops to have a firm understanding of the decomposition dynamics associated with legume cover crop roots.

The rapid rate of root N release in this study underscores the importance of quantifying legume cover crop root contributions to soil N pools.  From a production standpoint, it is essential that farmers utilizing legume cover crops have an understanding of total legume cover crop (shoots plus roots) N contributions that can be used by subsequent crops.  Such an understanding allows producers to better gauge yield expectations of summer crops and plan for additional fertilization, if needed.  Our results indicate that HV, in particular, may be an especially appealing cover crop option to farmers in the Southeastern U.S. due to its relatively large contribution to soil N pools in the weeks following termination, a portion of which is likely derived from HV roots, which release N at a faster rate than roots from other popular cover crops such as CC.   


Areas needing additional study

While it is evident that roots, especially HV, released N rapidly, it is important that root-derived N be quantified to determine if its reservoir is large enough to play a role in furnishing N for subsequent summer crops. To accurately quantify legume cover crop root-derived N, more effective methods to determine total root biomass on which estimates of root N are based should be employed that account for fine roots and root turnover.

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.