N2 fixation in Thai soybeans: Effect of tillage and inoculation on15N-determined N2 fixation in recommended cultivars and advanced breeding lines

The practice of seeding soybeans following paddy rice in Thailand has encountered difficulties in seedling germination, nodulation and crop establishment. This research project evaluated the choice of a non-fixing control to quantify N₂ fixation by¹⁵N isotope dilution, and the effect of tillage regime, soybean cultivar, strain ofBradyrhizobium japonicum and P fertilization on yield and N₂ fixation after paddy rice in northern and central Thailand.Japanese non-nodulating lines Tol-0 and A62-2 were the most appropriatecontrol plants for¹⁵N isotope dilution for Thai soybeans in these soils which contained indigenous rhizobia. Cereals such as maize, sorghum and barley were also appropriate controls at some sites. The choice of the appropriate non-fixing control plant for the¹⁵N isotope dilution technique remains a dilemma and no alternative exists other than to use several possible controls with each experiment. Acetylene reduction assay (ARA) proved of little value for screening varieties on their N₂ fixing capacity.The recommended Thai soybean cultivars (SJ1, 2, 4, 5) and an advanced line 16–4 differed little in their ability to support N₂ fixation or yield, possibly due to similar breeding ancestry. The ten AVRDC (ASET) lines showed considerable genotypic control in their ability to utilize their three available N sources (soil, fertilizer, atmosphere) and to translate them into yields. None of these lines were consistently superior to Thai cultivars SJ4 or SJ5 although ASET lines 129, 209 and 217 showed considerable promise.Neither recommended Thai or ASET cultivars were affected by tillage regime. Zero tillage resulted in superior N₂ fixation and yield at two sites but conventional tillage was superior at another site. Soybean cultivars grown in Thailand were well adapted to zero tillage. Levels of N₂ fixation were similar to world figures, averaging more than 100 kg N ha^–1 and supplying over 50% of the plant's N yield. However, seed yields seldom exceeded 2 t ha^–1, well below yields for temperately-grown soybeans. It is not clear why Thai soybeans support N₂ fixation, but do not translate this into higher seed yields.     

Mycorrhizal dependence of four crops in a P-sorbing soil

The effect of methyl bromide and P fertilization on the growth of four agricultural crops in a P-sorbing soil was studied. Harvestable yields and P tissue content were significantly lower for celery (Apium graveolens L.), onion (Allium cepa L.) and pepper (Capsicum annuum L.) following fumigation. P fertilization up to 330 kg P ha⁻¹ as superphosphate could not reverse this effect. The highest P fertilizer treatment had 25.4 μg g⁻¹ bicarbonate-extractable P and 0.26 μg g⁻¹ water extractable P. This P level produced acceptable commercial yields when vesicular arbuscular mycorrhizae (VAM) were present (non-fumigated) and indicate the importance of the VAM association for these crops. In contrast, melon (Cucumis melo L.) growth was slightly greater following soil fumigation at all P fertilization rates. These findings indicate that crop VAM dependence and efficacy of VAM biotypes must be considered as essential components for P fertilization recommendations on fumigated P-sorbing soils. Formerly Gilat, now University of Kent, Kent, UK     

Bradyrhizobium technology: a promising substitute for chemical nitrogen fertilizer in Bangladesh agriculture

Nitrogen is the most limiting element in Bangladesh soils and urea is the fertilizer commonly used for supplying it. Bradyrhizobium/Rhizobium inoculant was tried as a source of N nutrition for grain legumes in a number of field experiments. The inoculants markedly increased nodule number, nodule mass, shoot weight and yield of the crops compared to uninoculated control and urea-N treatments. For soybean (Glycine max), inoculation increased yield 113 percent over the control and 49 percent over the urea treament. For groundnut (Arachis hypogaea), the increases were 36 and 11 percent; for lentil (Lens culinaris), 30 and 13 percent; and for mungbean (Vigna radiata), 47 and 7 percent. The local inoculant strains were suitable for dependable inoculant production. The inoculant technology can be used as a promising and cheap substitute of urea for growing food legume crops in Bangladesh.     

Response to inoculation and N fertilization for increased yield and biological nitrogen fixation of common bean (Phaseolus vulgaris L.)

Common bean (Phaseolus vulgaris L.) is able to fix 20–60 kg N ha^–1 under tropical environments in Brazil, but these amounts are inadequate to meet the N requirement for economically attractive seed yields. When the plant is supplemented with N fertilizer, N₂ fixation by Rhizobium can be suppressed even at low rates of N. Using the ¹⁵N enriched method, two field experiments were conducted to compare the effect of foliar and soil applications of N-urea on N₂ fixation traits and seed yield. All treatments received a similar fertilization including 10 kg N ha^–1 at sowing. Increasing rates of N (10, 30 and 50 kg N ha^–1) were applied for both methods. Foliar application significantly enhanced nodulation, N₂ fixation (acetylene reduction activity) and yield at low N level (10 kg N ha^–1). Foliar nitrogen was less suppressive to nodulation, even at higher N levels, than soil N treatments. In the site where established Rhizobium was in low numbers, inoculation contributed substantially to increased N₂ fixation traits and yield. Both foliar and soil methods inhibited nodulation at high N rates and did not significantly increase bean yield, when comparing low (10 kg N ha^–1) and high (50 kg N ha^–1) rates applied after emergence. In both experiments, up to 30 kg N ha^–1 of biologically fixed N₂ were obtained when low rates of N were applied onto the leaves.     

Evaluation of strain competitiveness in Rhizobium leguminosarum bv phaseoli using a nod + fix− natural mutant

Competition from native soil rhizobia is likely to be an important factor limiting Phaseolus vulgaris L. inoculant response in Latin America. We used UMR 1116, a nod ⁺ fix^– natural mutant of Rhizobium leguminosarum bv phaseoli strain CC511, as a reference strain to study competition for nodulation sites in this species. When P. vulgaris cv Carioca was planted in soils containing different proportions of UMR 1116 and the effective and competitive strain UMR 1899, UMR 1116 occupied more than 50% of the nodules at all inoculant ratios tested, though increasing the proportion of UMR 1899 in the inoculant did enhance the number and percentage of effective nodules and plant dry weight. Sixty two strains of bean rhizobia were tested in competition with UMR 1116. An inoculant ratio of 1:1 was used, with all strains applied to the soil rather than to seeds. Strains varied in the number and percentage of effective nodules produced in competition with UMR 1116, and in plant dry weight, and there was a strong correlation between variation in each of these traits and plant N accumulation. Seven of the strains (UMR 1073, 1084, 1102, 1125, 1165, 1378 and 1384) were identified as both superior in competitive ability and active in N₂ fixation. Site of placement of the inoculant and ambient temperature influenced strain response.Journal paper 16736, Agricultural Experiment Station, University of Minnesota, St. Paul, MN 55108, USA     

Field release of genetically marked Azospirillum brasilense in association with Sorghum bicolor L.

The agronomic impact of genetically tagged azospirilla (Azospirillum brasilense)was assessed in open field and their fluctuation were monitored in the soil/rhizosphere. Strain performance, upon inoculation of sorghum, was evaluated over a two-years period; agronomic treatments included nitrogen application (0, 80, 160 kg ha^–1), and types of inoculant (Sp245 lacZ, Sp6 gusA, Sp6 IAA⁺⁺ gusA). Grain yield was higher for inoculated seed plots than in non-inoculated ones, whereas nitrogen content, biomass of plant residues and nitrogen in plant residues gave values that were not statistically different. Root length density (RLD) of sorghum at the end of the stem elongation stage was affected only by the indole-3-acetic acid (IAA) overproducer Azospirillum strain (A. brasilense Sp6 IAA⁺⁺ gusA) with respect to the normal IAA producer (A. brasilense Sp6 gusA), being higher in the first 40 cm of depth, notwithstanding the level of nitrogen fertilization. The traceability of the released genetically modified strains enabled to monitor their ability to colonise soil and roots. Moreover, the genetic modification per se vs. the non-modified counterpart, did not affect the culturable aerobic population in soil, microfungi, streptomycetes, fluorescent pseudomonads, soil microbial biomass, or some microbial activities, all selected as important indicators.     

Nitrogen fixation (15N dilution) with soybeans under Thai field conditions

Controlled environment and field studies were conducted to determine relationships between various measurements of N₂ fixation using soybeans and to use these measures to evaluate a number ofBradyrhizobium japonicum strains for effectiveness in N₂ fixation in Thai soils.¹⁵N dilution measurements of N₂ fixation showed levels of fixation ranging from 32 to 161 kg N ha⁻¹ depending on bacterial strain, host cultivar and location. Midseason measures of N₂ fixation were correlated with each other, but not related measures taken at maturity. Ranking ofB. japonicum strains based on performance under controlled conditions in N-free media were highly correlated with rankings based on soybean seed yields and N₂ fixation under field conditions. This study showed that inoculation of soybeans with effectiveB. japonicum strains can result in significant increases in yield and uptake of N through fixation. The most effective strains tested for use in Thai conditions were those isolated from Thai soils; however, effective strains from other locations were also of benefit.     

Effect of seed inoculation,mycorrhizal infection and organic amendment on wheat growth

Summary The effect of seed inoculation withAzotobacter spp. orAzospirillum spp., and garbage amendment (0.5%), on the growth of wheat was studied in a field experiment under sub-tropical conditions. Two levels of N fertilizer were applied, the usual field rate (150 kg N ha^–1) and half this amount. Tillering of plants, dry matter contents and nitrogenase activity were determined 30, 60 and 90 days after sowing. At the end of the experimental period, spore numbers and percentage of mycorrhizal infection were observed in the rhizosphere and root systems of plants. Straw and grain yields were also determined. The results of this study showed that seed inoculation and/or organic amendment stimulated plant growth, nitrogenase activity and mycorrhizal infection. This was more noticeable withAzotobacter than withAzospirillum. Inoculation withAzotobacter together with 1/2 N dose and organic amendment was the most effective application (19.75 and 10.70 t ha^–1 were recorded for straw and grain yield, respectively).     

Associative effect ofAzospirilium brasilense withRhizobium japonicum on nodulation and yield of soybean (Glycine max)

Summary Azospirillum was associated with nodules of soybean. In general, seed inoculation with a broth culture ofAzospirillum brasilense alone significantly increased nodulation and grain yield of soybean grown in pots in unsterilized soil with different levels of urea ranging from 0 to 80 kg N/ha. This trend was significantly reproducible in a second experiment when a carrier based inoculant of the bacterium was used for seed inoculation.Inoculation withRhizobium japonicum andA. brasilense in combination generally increased grain yield in both the experiments, although the data were not significant.     

Effects of P-fertilization and inoculation by two vesicular-arbuscular mycorrhizal fungi on growth and nodulation ofCalopogonium caeruleum

The growth response ofCalopogonium caeruleum, a leguminous covercrop in plantation agriculture, to inoculation with two vesicular-arbuscular mycorrhizal (VAM) fungi was investigated in five phosphorus (P)-deficient soils supplied with various levels of rock phosphate. Significant shoot yield increases over the uninoculated controls were obtained in most sterilised or unsterilised soils at all applied P levels, although the inoculant VAM fungi differed in their effectiveness in the soils used. Responses in mycorrhizal root infections, P and nitrogen (N) concentrations in tops and plant nodulation varied. The results are discussed in relation to the edaphic environment of the mycorrhizal association.     

Effects of phosphate fertilization,lime amendments and inoculation with VA-mycorrhizal fungi on soybeans in an acid soil

Soybeans [Glycine max (L.) Merr. cv. Essex] were grown in nonsterile acid (pH. 5.2) infertile Wynnville silt loam (Glossic Fragiudult) in a glasshouse. The effects of P fertilization and lime were determined by inoculation with two VAM-fungi (VAMF): Glomus fasciculatum (Gf) and Glomus etunicatum (Ge). An important factor affected by the interaction between applied lime (soil acidity), applied P, and VAMF inoculation was the soil Al. Five application rates of P as KH₂PO₄ and three rates of lime were tested. Potassium was equalized with KCl (muriate of potash). P-efficiency (g seed/mg P kg^-1 soil) by vesicular-arbuscular mycorrhiza (VAM) was maximal at 20 mg P kg^-1 soil at all lime and VAMF treatments. VAMF inoculation increased plant survival and protected the soybeans from leaf scorch, thereby substituting for the effects of lime and P. The Ge inoculum was superior in ameliorating leaf scorch in the nonlimed soil. The Gf inoculum required more lime and P than the Ge inoculum to increase seed yield relative to the noninoculated controls containing only native VAMF. Both inocula increased root Al uptake and extractable soil Al in the acid soil without apparent adverse effects on root or shoot. The ability of the VAMF inocula to enhance the efficiency of applied P and decrease seed Cl concentration was increased by lime. Seed yield (Y) was negatively related to seed Cl concentration (X) where Y=aX^-b. Both VAMF inoculation and lime application reduced this negative relationship and may have increased the tolerance to both Cl and soil Al.     

Short-term survival of rhizobia on arrowleaf clover seed sown at different depths

Seed of arrowleaf clover (Trifolium vesiculosum Savi) were inoculated with a streptomycin resistant mutant ofRhizobium leguminosarum biovartrifolii and planted on the surface of a Norwood fine sandy loam and at 10 and 25 mm depths. Populations of rhizobia declined from an excess of 10,000 seed⁻¹ immediately after inoculation to less than 100 within three to four days after sowing on the soil surface when water was the peat inoculant adhesive. Gum arabic as the adhesive promoted the survival of rhizobia. Populations of rhizobia on surface sown seed declined much more rapidly than on seed buried in soil. Although, the soil was nearly air dry, rhizobia on buried seed survived at populations exceeding 1,000 seed⁻¹. The maximum soil temperatures ranged between 21 and 36°C over the sampling time and did not seem to have a major influence on short term survival of rhizobia. Delayed germination of seed due to the higher temperature would indirectly influence the number of viable rhizobia present at germination.     

15N-Determined effect of inoculation with N2 fixing bacteria on nitrogen assimilation in Western Canadian wheats

Summary A two-year field study was undertaken using¹⁵N isotope techniques to differentiate between stimulation of N uptake and N₂ fixation in Western Canadian cultivars of spring wheat (Triticum aestivum L. emend Thell) and durum (T. turgidum L. emend Bowden) in response to inoculation with N₂-fixing bacteria. Bacterial inoculation either had no effect or lowered the % N derived from the fertilizer and the fertilizer use efficiency. Despite the depression of fertilizer uptake, inoculants did not alter the relative uptake from soil and fertilizer-N pools indicating that bacterial inoculation did not alter rooting patterns. Nitrogen-15 isotope dilution indicated that N₂ fixation did occur. In 1984, % plant N derived from the atmosphere (% Ndfa) due to inoculation with Bacillus C-11-25 averaged 23.9% while that withAzospirillum brasilense ATCC 29729 (Cd) averaged 15.5%. In 1985, higher soil N levels reduced these values by approximately one-half. Cultivar x inoculant interactions, while significant, were not consistent across years. However, these interactions did not affect cultivars ‘Cadet’ and ‘Rescue’. In agreement with previous results, ‘Cadet’ performed well with all inoculants in both years while ‘Rescue’ performed poorly. Among 1984 treatments, the N increament in inoculated plants was positively correlated with % Ndfa but no such correlation existed in 1985. N₂ fixation averaged over all cultivars and strains was 17.9 and 6.7 kg N fixed ha⁻¹ in 1984 and 1985, respectively. Highest rates of N₂ fixation were estimated at 52.4 kg N ha⁻¹ for ‘Cadet’ in 1984 and 31.3 kg N ha⁻¹ for ‘Owens’ in 1985, both inoculated with Bacillus C-11-25, an isolate from southern Alberta soils. Inoculation with either ofAzospirillum brasilense strain Cd (ATCC29729) or 245 did not result in as consistent or as high N₂ fixation, suggesting that these wheats had not evolved genetic compatability with this exogenous microorganism. These agronomically significant amounts of N₂ fixation occurred under optimally controlled experimental conditions in the field. It is yet to be determined if N₂ fixation would occur in response to bacterial inoculation under dryland conditions commonly occurring in Western Canada. Contribution from Agriculture Canada Research Station, Lethbridge, Alberta, Canada.     

The effect of inoculation with Azospirillum brasilense on growth and nitrogen utilization by wheat plants

Azospirillium brasilense is a rhizosphere bacteria that has been reported to improve yield when inoculated on wheat plants. However, the mechanisms through which this effect is induced is still unclear. In the present work, we have studied the effects of inoculating a highly efficient A. brasilense strain on wheat plant grown in 5 kg pots with soil in a greenhouse, under three N regimes (0, 3 or 16 mM NO₃ ^–, 50 ml/pot once or twice-a -week), and in disinfected or non-disinfected soil. At the booting stage, the inoculated roots in both soils showed a similar colonization by Azospirillum sp. that was not affected by N addition. The plants grown in the disinfected soil showed a higher biomass, N content and N concentration than those in the non-disinfected soil, and in both soils the inoculation stimulated plant growth, N accumulation, and N and NO₃ ^– concentration in the tissues.At maturity, the inoculated plants showed a higher biomass, grain yield and N content than the uninoculated ones in both soils, and a higher grain protein concentration than the uninoculated. It is concluded that in the present experiments, A. brasilenseincreased plant growth by stimulating nitrogen uptake by the roots.     

Plant growth stimulation of wheat (Triticum aestivum L.) by inoculation of salinity tolerant Azotobacter strains

Five salinity tolerant Azotobacter strains i.e., ST3, ST6, ST9, ST17 and ST24 were obtained from saline soils. These Azotobacter strains were used as inoculant for wheat variety WH157 in earthen pots containing saline soil under pot house conditions, using three fertilizer treatment doses i.e., control (no fertilizer, no inoculation), 90 Kg N ha⁻¹ and 120 Kg N ha⁻¹. Inoculation with salinity tolerant Azotobacter strains caused significant increase in total nitrogen, biomass and grain yield of wheat. Maximum increase in plant growth parameters were obtained after inoculation with Azotobacter strain ST24 at fertilization dose of 120 kg N ha⁻¹ and its inoculation resulted in attaining 89.9 cms plant height, 6.1 g seed yield, 12.0 g shoot dry weight and 0.7 % total nitrogen. The survival of Azotobacter strain ST24 in the soil was also highest in all the treatments at 30, 60 and 90 days after sowing (DAS). However, the population of Azotobacter decreased on 90 DAS as compared to counts observed at 60 DAS at all the fertilization treatments.     

Soil and litter respiration rates in different microhabitats of a mixed oak-conifer forest and their control by edaphic conditions and substrate quality

Summary Seven isolates belonged toA. brasilense and 3 belonged toA. lipoferum. Isolates having more denitrifying capacity fixed less nitrogen in nitrogen free semi-solid malate medium. One strain ofA. lipoferum having high nitrogen fixing capacity with negative test for denitrification was tested as inoculant to supplement the nitrogen need of a wheat crop in field condition with different doses of N with and without the inoculant. While control without nitrogen yielded 1260 kg/ha the yield in inoculated treatment was 2070 kg/ha resulting in significant increase. In a treatment receiving 40 kg N/ha the grain yield was 2370 kg/ha as against yield of 3110 kg/ha in a similar treatment receiving fertiliser plus inoculant. Thus increase in yield was about 30%. Further the treatment receiving 80 kg N/ha yielded 2970 kg/ha as against yield of 4150 kg/ha in a treatment receiving inoculant alongwith the above dose of the fertiliser. Thus increase in yield due to application of inoculant was about 36%. Similarly, the uptake of N in different treatments was augmented due to inoculation of seeds with the culture.     

Improvement of growth,under field conditions,of wheat inoculated with <Emphasis Type="Italic">Pseudomonas chlororaphis</Emphasis> subsp.<Emphasis Type="Italic"> aurantiaca</Emphasis> SR1

Effects of inoculation of wheat (Triticum aestivum L.) with the rhizobacterium Pseudomonas chlororaphis subsp. aurantiaca strain SR1 (termed SR1) were studied at an experimental field site in Río Cuarto, Argentina. Treatments involved SR1 inoculation with or without nitrogen/phosphorus fertilization. Inoculation produced a significant increase in plant height and root length in early growth stages. Inoculation plus fertilization with 40 kg ha⁻¹ urea/30 kg ha⁻¹ diamonic phosphate (“50% dose”) gave a yield increase of 636 kg ha⁻¹ relative to control, and an increase of 472 kg ha⁻¹ relative to fertilization with 80 kg ha⁻¹ urea/60 kg ha⁻¹ phosphate without inoculation. SR1 inoculation without fertilization, compared to control, produced increases of 6% in weight of 1,000 grains, 13% in number of spikes per plant, and 30% in number of grains per spike. Inoculation plus 50% dose fertilization also improved these parameters. Results of the study indicate that inoculation of wheat with SR1 improves various growth and yield parameters, and allows reduced dosage of nitrogen/phosphorus fertilizers in the field.     

Establishment of Bradyrhizobium japonicum for soybean by inoculation of a preceding wheat crop

On fields with no history of soybean (Glycine max (L.) Merr.) production, inoculation alone is often inadequate to provide for adequate nodulation the first time this crop is grown. The objective of this study was to determine if inoculation of spring wheat (Triticum aestivum L.) seed with Bradyrhizobium japonicum would lead to an increase of B. japonicum numbers in the soil, and improve nodulation of a subsequent soybean crop. In the greenhouse, wheat seed inoculation increased B. japonicum numbers from undetectable numbers to >9000 g^–1 of soil, whereas the numbers of introduced B. japonicum declined in unseeded pots. In the field, inoculation of wheat seed increased B. japonicum numbers in the soil from undetectable levels to >4000 g^–1 the following year. When soybean seed was inoculated, but grown in soil devoid of B. japonicum, nodules formed only near the point of seed placement. The heaviest nodulation, and widest distribution of nodules in the topsoil were found whenB. japonicum was established the year before by wheat seed inoculation, plus soybean seed inoculation. Wheat seed inoculation the year before growing soybean, combined with proper soybean seed inoculation, should provide for abundant nodulation the first time soybean is grown on a field.     

Nitrogen transfer from nodulating soybean to maize or to nonnodulating soybean in intercrops: the 15N dilution method

In 1985, 1986 and 1988, maize (Zea mays L.) was monocropped or intercropped with nodulating or nonnodulating soybean (Glycine max [L.] Merr.). In addition, nodulating soybean and nonnodulating soybean were each monocropped and grown as a mixture. In 1985 and 1986, treatments were grown at 0 and 60 kg N ha^–1 and in 1988, the treatments were grown without N fertilizer, on N-depeted soil and on non-N-depleted soil. ¹⁵N enriched N was applied to soil in all the aforementioned treatments to test for N transfer from nodulating soybean to non-N₂-fixing crops by the ¹⁵N dilution method.The ¹⁵N dilution method did not show the occurrence of N transfer in 1985 and 1986, but the N sparing effect was evident from the total N uptake of nonnodulating soybean, dwarf maize and tall maize, in 1986. In 1988, maize and nonnodulating soybean seed yields and seed N yields were higher on non-N-depleted soil than on N-depleted soil. On N-depleted soil, the ¹⁵N dilution method indicated N transfer from nodulating soybean to maize and to nonndulating soybean. At a population ratio of 67% nodulating soybean to 33% nonnodulating soybean, N transfer was also seen on non-N-depleted soil in 1988.     

Incorporation of Plant Residues into Soil Organic Matter Fractions With Grassland Management Practices in the North American Midwest

Disturbed grassland soils are often cited as having the potential to store large amounts of carbon (C). Fertilization of grasslands can promote soil C storage, but little is known about the generation of recalcitrant pools of soil organic matter (SOM) with management treatments, which is critical for long-term soil C storage. We used a combination of soil incubations, size fractionation and acid hydrolysis of SOM, [C], [N], and stable isotopic analyses, and biomass quality indices to examine how fertilization and haying can impact SOM dynamics in Kansan grassland soils. Fertilized soils possessed 113% of the C possessed by soils subjected to other treatments, an increase predominantly harbored in the largest size fraction (212–2,000 μm). This fraction is frequently associated with more labile material. Haying and fertilization/haying, treatments that more accurately mimic true management techniques, did not induce any increase in soil C. The difference in ¹⁵N-enrichment between size fractions was consistent with a decoupling of SOM processing between pools with fertilization, congruent with gains of SOM in the largest size fraction promoted by fertilization not moving readily into smaller fractions that frequently harbor more recalcitrant material. Litterfall and root biomass C inputs increased 104% with fertilization over control plots, and this material possessed lower C:N ratios. Models of incubation mineralization kinetics indicate that fertilized soils have larger pools of labile organic C. Model estimates of turnover rates of the labile and recalcitrant C pools did not differ between treatments (65.5 ± 7.2 and 2.9 ± 0.3 μg C d⁻¹, respectively). Although fertilization may promote greater organic inputs into these soils, much of that material is transformed into relatively labile forms of soil C; these data highlight the challenges of managing grasslands for long-term soil C sequestration.