Nitrogen fixation by red clover as related to the supply of Cobalt and Molybdenum from some Norwegian soils

Abstract

A field trial, a pot experiment and a survey of organically farmed leys were undertaken to investigate whether N fixation in red clover pastures in Norway was limited by a low supply of cobalt and/or molybdenum. Fertilization with Mo did not result in any higher production or N fixation, whereas the N yield both from established clover leys and red clover grown in pots increased slightly after application of Co to many of the investigated soils. In the organically farmed leys there was a significant and positive correlation between Co content and the N content of the red clover. As many of the investigated clover-soil systems were of those previously known to be very low in Co and Mo, and the gain in N yield obtained by extra Co supply was marginal, it is unlikely that deficiency of these trace elements is a problem of great concern in legume based forage production systems in Norway.

Keywords:

• biological nitrogen fixation, Co
• Festuca
• Mo
• N-yield
• Phleum
• Rhizobium
• Trifolium

Introduction

The trace elements cobalt (Co) and molybdenum (Mo) are both important components of Rhizobium-legume symbiosis. Cobalt constitutes the central atom in the porphyrin ring structure of the coenzyme cobalamin (vitamin B₁₂) that is essential for nodulation and bacterioid development, and Mo has a key role in the nitrogenase enzyme complex catalysing the N fixation (Marschner, 1995).

With regard to the needs of ruminants, there has been concern about the rather low Co content in roughage harvested in several regions in Norway, and the Mo content has also been variable in some investigations, and suboptimal at some sites (Frøslie & Norheim, 1983; Synnes & Øpstad, 1995; Johansen et al., 2003). In this context, whether the supply of these elements is sufficient for optimal functioning of the Rhizobium-clover symbiosis has never been addressed. We do, however, know from reports from other countries that the vigour, nodule development and N content of some legumes might be positively affected by additional supply of Co and/or Mo when soil content of these elements is low (Vrany, 1978; Marschner, 1995; Mengel & Kirkby, 2001; O'Hara, 2001).

Against this background, and because biological N fixation is of considerable importance both in conventional and organic forage production, we decided to investigate the trace element status of red clover pastures in Norway. Some results and possible implications of them are outlined in this paper.

Materials and methods

Field trial

Seven experimental fields were established in the spring of 2000 at different locations along the western coast and in the interior of central Norway (Table 1) where the supply of cobalt and/or molybdenum from the soil was expected to be low. At each site four fertilizer treatments were replicated three times on pairs of 7.0 m×1.5 m plots in a randomized block design. Within pairs, one plot was sown with a mixture of timothy (Phleum pratense L.) and meadow fescue (Festuca pratensis Huds), and the other with red clover (Trifolium pratense L.) added to the same grass mixture. In fields where no clover had been grown for the previous five years, Rhizobium trifolii from Baljväxtlaboratoriet, Uppsala, Sweden, was harrowed into the soil surface immediately before sowing. The bacterial suspension as received from Baljväxtlaboratoriet was mixed with soil from the site in a ratio of 1:500 (w/w) before application to the soil surface by hand at a rate of 0.3 g bacterial suspension m⁻².

Table 1. Volume weight, pH and the content of plant available Co in dried soil samples from the plough layer at the initiation of a series of field experiments in grass-red clover leys in central and western Norway. The mean content of Co in dried clover from first cuts on plots either supplied with or not supplied with extra Co, and the accumulated difference in N yield between the two fertilizer treatments, are also given

	Soil characteristics			Co content in clover (ppm in DM)		N yield difference accumulated over three harvests *
Field location (county) and soil type	pH	Vol.weight (kg l^−1)	Co-content (ppm)	0 g Co ha^−1	150 g Co ha^−1	(kg N ha^−1)
Måløy (sand)	7.6	1.36	0.03	0.019	0.063	38
Giske (sand)	7.4	1.19	0.03	0.023	0.162	51
Vindafjord (sand)	5.9	0.98	0.06	0.049	0.184	48
Stjørdal (silt loam)	5.9	0.97	0.18	0.061	0.210	−26
Alvdal (silt loam)	7.0	0.92	0.32	0.110	0.192	13
Fusa (loamy sand)	4.9	1.00	0.04	0.212	1.091	46
Selbu (organic soils)	5.2	0.29	0.31	0.314	0.835	31
*: The N yield was significantly (P=0.012) affected by Co fertilization.

Immediately after sowing, a solution of either CoCl₂×6H₂O (150 g Co ha⁻¹), of Na₂MoO₄×2H₂O (600 g Mo ha⁻¹) or a solution of both CoCl₂×6H₂O and Na₂MoO₄×2H₂O (150 g Co ha⁻¹+600 g Mo ha⁻¹) was applied on the pairs of plots. The fourth treatment consisted of pairs of plots where no trace elements were added.

In the first and second year of the ley, the fields were fertilized with 60 kg N ha⁻¹ in spring and 40 kg N ha⁻¹ after the first cut. P, K and S were also supplied in quantities consistent with local practice. On plots fertilized with cobalt chloride immediately after sowing, the N source in spring both years was NitraCo™ (calcium nitrate containing 0.02% Co from Hydro Agri). The other plots were supplied with corresponding quantities of calcium nitrate.

The first cut was taken between early heading and full heading of timothy, and the second cut about eight weeks later. The content of N in the total yield (all plots and harvests), and of Mo and Co in grass and clover separately (composite samples pooled over replicates) was analysed. Nitrogen fixation in the clover-Rhizobium symbiosis was estimated by the N-difference method (Nesheim & Øyen, 1994).

In the final analyses of yields, N fixation and plant elemental content, 7 fields and 3 harvests were included. The data for yield, N yield and N fixation were subjected to analysis of variance (ANOVA) with field, harvest and fertilizer treatment as class variables, and for the test of the effect of treatment, the field treatment interaction was used as an error term. The content of Mo and Co in plant samples was also analysed by ANOVA with species and fertilizer treatment as class variables.

Pot experiment

On the basis of results from a previous survey of mineral status in leys in western Norway (Synnes & Øpstad, 1995), soils from the plough layer at locations where the plant samples were very low in Co and/or Mo were collected. Red clover was grown in these soils for 13 weeks from germination in October until harvest in January. There were 6 plants per pot containing 1 l of soil mixed with 0.25 l Ticon perlite (expanded volcanic rock). The perlite contained 0.04 ppm Mo and 0.02 ppm Co. The air temperature in the acrylic greenhouse was 18–20°C, and the natural daylight was supplemented with light from high pressure sodium lamps (180 μmolm⁻²s⁻¹ at plant height, 18 h photoperiod). Altogether 35 mg N was added to each pot during the first four weeks after germination. Other macronutrients, B and Cu were supplied as a compound fertilizer (Hydro-PK™ 5–17) every second week until harvest. Mn and Zn (as sulphates) and Fe (as FeEDTA) were also added to the solution of the compound fertilizer; 5 ml of a suspension of Rhizobium trifolii was injected just after sowing into the upper 5 cm of the soil in each pot.

The experiment was split into one Co investigation and one Mo investigation conducted within the same greenhouse compartment. Six soil types (limed and fertilized peat moss, and mineral soils from five farms) were included in each of them. In the first, all pots were supplied with 2.65 mg Na₂MoO₄×2H₂O, but only half of the pots with 1.06 mg CoCl₂×6H₂O. In the second investigation, all pots were supplied with 1.06 mg CoCl₂×6H₂O, but only half of the pots with 2.65 mg Na₂MoO₄×2H₂O. There were three pots per treatment (combination of soil and Mo/Co fertilization), and each treatment was replicated three times.

At harvest the plants were cut at the soil surface, dried, weighed and analysed for total N content. The plant contents of Co and Mo were analysed in one of the replicates. Nodules from the same replicate were also dissected from roots, rinsed in deionised water and dried. All nodules within the same fertilizer treatment constituted one compound sample which was analysed for Co and Mo. The data for DM yield and N content for each experiment separately were subjected to ANOVA with fertilizer treatment and soil type as class variables, whereas the data for plant trace element content were subjected to one-way ANOVA with treatment as class variable and soil types as replicates.

Survey in organically farmed grass-clover leys

As part of the research programme ‘Mineral contents in plants and mineral supply for ruminants in organic agriculture’ (Strøm et al., 2003) red clover from 22 organic farms in central, eastern and western Norway was collected and analysed for Co, Mo and N at the time of the first cut in 2001. Samples from three different leys within each farm were taken, and the stage of phenological development of the predominant grass species and red clover at harvest was determined. The proportion which the clover constituted of the plant stand was determined by sorting and later drying of subsamples of the gross yield from three 7 m×1.4 m plots. Soil samples from the same plots were taken in spring of the same year and analysed for pH, mineral and carbon content, and texture. The relationship between the different soil characteristics, the clover content of the leys, and the plant content of Mo, Co and N was later investigated by correlation and regression analyses.

Plant and soil analyses

All plant samples to be analysed for elemental content were dried at 60–70°C after harvest and then ground in a Cyclotec 1093 sample mill (Foss Tecator) with a grinding ring and impeller guaranteed not to contaminate the samples with any Mo and Co. The content of Mo and Co was analysed at SGAB Analytica in Luleå, Sweden by plasma-mass-spectroscopy (ICP-SMS) after extraction with HNO₃ and H₂O_2, and the total N content by standard Kjeldahl methods at Chemical Analytical Laboratory at Holt Research Centre in Tromsø, Norway. The pH and content of easily soluble macronutrients in the soil were analysed at the same laboratory, and the content of soluble Co and Mo at Jordforsk laboratory, the Norwegian Centre for Soil and Environmental Research at Ås, Norway.

Results and discussion

The results from the present investigations did not indicate that the growth and N fixation of red clover was limited by the supply of Mo. Additions of the element to soils expected or known to be low in Mo did not cause any higher clover production or N content, either in the field or in the pot experiment (Table 3, data not shown).

All clover samples from the organically farmed sites contained more than 0.2 ppm Mo, a critical level for deficiency according to the authors cited in Johansen et al. (1997). Plants not supplied with any extra Mo contained less than 0.2 ppm in three of the soils in the pot experiment and at one of the sites in the field trial. Additional supplement of Mo to these soils and a subsequent increase in plant uptake of Mo did not cause any higher N fixation or plant production. It thus seems that if a critical level for deficiency exists, it must be well below 0.2 ppm, even in relatively young plants (at early bud stage in the pot experiment). The total range on organic farms was 0.3–14.8 ppm, and the content in clover plants was significantly and positively correlated with the pH and the content of Mo in the soil. In all the present investigations, the content of Mo was higher in red clover than in grasses (Table 2, data not shown).

Table 2. The content of Co and Mo in red clover and in a mixture of timothy and meadow fescue grown and harvested together in a series of field experiments. The samples were harvested between early and full heading of timothy, and means for the first cut at 7 fields/locations are given

	Co content (ppm in DM)*		Mo content (ppm in DM)*
Species, fertilizer treatment	0 g Co ha^−1	150 g Co ha^−1	0 g Mo ha^−1	600 g Mo ha^−1
Timothy/fescue, 1st cut 2001	0.04	0.09	1.54	4.76
Timothy/fescue, 1st cut 2002	0.02	0.11	0.81	3.28
Red clover, 1st cut 2001	0.08	0.37	3.87	14.75
Red clover, 1st cut 2002	0.11	0.38	2.58	8.88
*: Significant differences (P<0.05) between species and fertilizer treatments for all harvests.

In contrast to the situation with Mo, there were indications that the supply of Co from some soils was too low to sustain the potential for N fixation (Tables 1 and 3). In the pot experiment, the yield of clover N was on average about 10% higher after Co fertilization, and in the field trial the N yield of the swards increased slightly after an addition of extra Co at six of the seven investigated sites. It should, however, be noted here that when the N fixation at the experimental units was calculated as the difference in N yield between the two neighbouring suplots with and without red clover (cf. Materials and methods section, and Nesheim & Øyen, 1994), the difference in N fixation between Co treatments was not statistically significant.

Table 3. Yields of dry matter and N, and content of Co and Mo in red clover grown for 13 weeks with and without extra supply of the respective trace elements. Means for plants grown in six different soils are given

			Co content (ppm in DM)		Mo content (ppm in DM)
Fertilizer treatment	Yield of stems+leaves (g per pot)	Nitrogen in stems+leaves (g per pot)	Stems+leaves	Nodules	Stems+leaves	Nodules
No Co added	21.9	0.58 (*)	0.18*	4.0
Extra Co added	23.7	0.64	0.72	8.0
No Mo added	20.6	0.54			1.22*	14.0
Extra Mo added	21.5	0.58			10.40	10.0
(*), *: Significant differences (P=0.081 and P<0.05) between fertilizer treatments.

According to Mengel & Kirkby (2001), soils containing at least 0.1 ppm Co meet in most instances the Co demand of the Rhizobium-legume symbiosis. Without any extra supplement, three of the soils in the present pot experiment and four of the soils in the field trial contained less than 0.1 ppm (data not shown, Table 1). The increment in N yield after the addition of Co was, however, not restricted to these soils only.

In the survey of organic leys, the soil at 31 out of 66 investigated sites contained less than 0.1 ppm Co, and the soil at 9 sites less than 0.05 ppm Co. There was a significant and positive correlation between the N content and the Co content of the clover at the 66 sites (see regression (1) below). Although it is not possible to draw any conclusions about a true causal relationship, this correlation does indicate that the supply of Co might not be high enough to sustain the demands of the N-fixating symbiosis in all soils.

(Adj R² for the model was 0.312, and all parameters were significant (P<0.03)).

Simple pairwise correlation analyses on the results from the survey further revealed that the content of Co in the clover decreased with increasing stage of phenological development, with decreasing content of clay and Co, and increasing pH in the soil.

Conclusion

The supply of Mo from the investigated Norwegian soils appeared to be sufficient to meet the demand of red clover-Rhizobium symbiosis. There were on the other hand indications that the supply of Co was insufficient to sustain the actual potential for symbiotic N fixation. As many of the investigated clover-soil systems were those previously known to be very low in Co, and the gain in N yield obtained by extra Co supply was marginal, we suggest that Co deficiency is not a problem that merits any significant concern in the further development of legume based forage production systems in Norway.

Acknowledgments

This work was financed by the Norwegian Research Council, The Agricultural Agreement between Norwegian farmers and the Government, and by the Norwegian Agricultural Extension Service. Anne Stine Ekker is acknowledged for her technical assistance.

Additional information

Notes on contributors

Anne Kjersti Bakken *

Bakken, A. K., Synnes, O. M. and Hansen, S. (The Norwegian Crop Research Institute, Kvithamar Research Centre, NO-7500 Stjørdal, Norway, Sunnmøre Local Experimental Group, P.O. Box 1312, NO-6001 Ålesund, Norway, and Norwegian Centre for Ecological Agriculture, NO-6630 Tingvoll, Norway). Nitrogen fixation by red clover as related to the supply of cobalt and molybdenum from some Norwegian soils.

Notes

Bakken, A. K., Synnes, O. M. and Hansen, S. (The Norwegian Crop Research Institute, Kvithamar Research Centre, NO-7500 Stjørdal, Norway, Sunnmøre Local Experimental Group, P.O. Box 1312, NO-6001 Ålesund, Norway, and Norwegian Centre for Ecological Agriculture, NO-6630 Tingvoll, Norway). Nitrogen fixation by red clover as related to the supply of cobalt and molybdenum from some Norwegian soils.