SciELO - Scientific Electronic Library Online

vol.36 issue1Chlorophyll content of leaf and mineral composition of passion fruits irrigated with saline water and biofertilizersReproductive parameters of suckling cows’ herds in Alentejo author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links

  • Have no similar articlesSimilars in SciELO


Revista de Ciências Agrárias

Print version ISSN 0871-018X

Rev. de Ciências Agrárias vol.36 no.1 Lisboa Jan. 2013


N2 fixation by two lupine species under different soil management systems. Transfer of fixed N2 from legume to intercropped eucalyptus

A fixação do N2 por duas espécies de tremoceiro em diferentes sistemas de gestão do solo. Transferência do N2 fixado pela tremocilha para o eucalipto em consociação

Corina Carranca1, Manuel Madeira2 and Maria Odete Torres3


1 Instituto Nacional de Investigação Agrária e Veterinária [INIAV], Centro de Engenharia dos Biossistemas [CEER], Quinta do Marquês,Av. República, Nova Oeiras, 2784-505 Oeiras, Portugal. E-mail:, author for correspondence.

2 Universidade Técnica de Lisboa, Instituto Superior de Agronomia [UTL/ISA], Centro de Estudos Florestais [CEF], Tapada da Ajuda,1349-017 Lisboa, Portugal

3 Universidade Técnica de Lisboa, Instituto Superior de Agronomia [UTL/ISA], Centro de Engenharia dos Biossistemas [CEER], Tapada da Ajuda,1349-017 Lisboa, Portugal



The rate and pattern of symbiotic fixation and N credit by white lupine as affected by soil tillage, and yellow lupine affected by soil amendment with eucalyptus harvest residues were evaluated; also, fixed N transferred from yellow lupine to the intercropped eucalyptus was assessed. Lupine species growing in sandy soils with low nutrient status fixed >94% N2 with the indigenous soil rhizobium, in two sites (Pegões, Óbidos). Soil disturbance and harvest residues incorporation did not affect the fixing capacity of both lupines. Positive N inputs (>+69 kg N ha-1) were estimated for both lupines, which is relevant in a rotation or association with non-legume crops. Nitrogen concentration in newly-planted eucalyptus was 23% lower in harvest residues amended soil and 36% greater by the addition of mineral-15N compared with organic fertilization (yellow lupine). Mineral-N use efficiency by eucalyptus seedlings was low (4%) which may be associated with the small root development (shoot:root ratio=4.3). Organically fertilized trees showed 15N enrichment thirteen times smaller than mineral-15N fertilized trees.

Keywords: Haplic Podzols, harvest residues, Lupinus albus and Lupinus luteus, N benefit, tillage



Avaliou-se a fixação simbiótica e crédito em N pelo tremoceiro doce em resposta à mobilização do solo, e pela tremocilha com incorporação no solo de resíduos de abate de eucalipto; avaliou-se também a transferência do N2 fixado pela tremocilha para o eucalipto consociado. As espécies de tremoceiro estudadas fixaram >94% N2 com o rizóbio indígena em solos arenosos, pobres em nutrientes, das áreas de Pegões e Óbidos. A mobilização do solo e os resíduos de abate não afectaram negativamente a fixação simbiótica pelos dois tremoceiros. Observaram-se benefícios em N (>+69 kg ha-1) por ambas as proteaginosas, o que é relevante em sistemas de rotação ou consociação. A concentração do N nos eucaliptos recém-plantados foi 23% menor no solo com resíduos.

Palavras-chave: Benefício em N, Lupinus albus e Lupinus luteus, mobilização do solo, Podzóis Háplicos, resíduos de abate



Grain legumes are grown worldwide and cultivated primarily for their grains which are harvested at maturity and marketed as fresh or dry products, rich in proteins for human consumption and animal feeding. However, their cultivation has been neglected for many years in the European Union (EU), which only produces about five million tonnes, especially peas (3 Mt), fababeans (1.2 Mt) and soybeans (0.8 Mt), and imports two-thirds of protein required for feed uses (Crépon et al., 2004). Grain legumes are expected to increase the eco-efficiency of agricultural production due to their ability to produce proteins of high quality without using nitrogen (N) fertilisers (Gaillard and Nemecek, 2004). In this way, it is supposed that energy use by grain legumes and N losses (nitrate, ammonia, etc.) are much lower than in non-legume crops. The current European Common Agricultural Policy favours the introduction of grain legumes in crop rotations where they are economically beneficial for farmers in several regions by their value to the next crop, but no specific financial support has been decided yet.

Legumes have crucial functions in the crop rotation via their N and non-N characteristics (pathogen inocula size, weed seed bank, soil microbial activity, soil aeration, etc.) for succeeding crops. In a grain legume-non-legume (e.g. cereal) sequence, legume N benefits are related to N carry-over of unused inorganic soil N by legume and legume residues N after the maturity phase, which may reduce the need for mineral N supply in the subsequent crop (cereal), potentially contributing to build up soil fertility.

The significance of N benefit in the shortand longterm cropping systems is determined by the crop N balance, the soil N dynamics after the legume has stopped to take up N from soil and the N demand of the succeeding crop (Criozat, 2004; Jensen, 2004; Carranca, 2013b). The N balance for legume-based systems can be both positive and negative, depending on the relationship between the fraction of N derived from the atmosphere (%Ndfa) and N removal (Carranca, 2013a). A positive N balance is most likely to occur when the proportion of N derived from N2 fixation is close to 100% (Carranca et al., 2009).

Lupine is an interesting seed pulse from which protein for human diet (sweet/white lupine) and fodder (yellow lupine) can be obtained. It may nodulate with slow growing soil Bradyrhizobium sp. or by inoculation with fast growing Rhizobium sp. bacteria strains, but limited research has been conducted on management options to enhance lupine ability to fix atmospheric N2. The amount of N2 fixed is controlled by the effectiveness of the rhizobium-plant symbiosis, plant sink strength, pre-existing soil N pools and fluxes, and environmental and nutritional constraints (Carranca, 2013a). Nitrogen fixation can be increased by management practices aimed to increase plant N demand, but some legumes may delay nodulation until they sense a lack of available soil N. Changes of soil management practices like the intensity of tillage will alter the edaphically chemical and biophysical factors, and therefore may indirectly influence symbiotic fixation (Pálmason et al., 1992; Carranca, 2013a). When tillage is minimized, lower rates of N mineralization and nitrification, coupled with increased N immobilization, will lead to a decrease in soil available N which stimulates N2 fixation capacity. Tillage may also destroy nodulated roots, reducing the fixation rate (Goss and De Varennes 2002). A promising development to avoid build-up of extra N in the soil may be perennially based cropping systems, intercropping, or eventually including perennials in annual cropping systems (Fuhrer, 2004; Carranca, 2013b).

Legume growth and symbiotic N2 fixing microorganisms may be inhibited by the amount and nature of allelophatic chemicals contained in residues from woody species, particularly leaves and roots (De Varennes et al., 2007; Rocha and Santos, 2007; Carranca, 2013a). For instance, essential oils in eucalyptus species, especially present in leaves show a complex composition, variable during the day and with time, and contain organic phenols, acetones, alcohols and antibiotics, among others (Ferguson and Rathinasabapathi, 2009). Plant growth and symbiotic N2 fixing microorganisms may be inhibited by the allelophatic chemicals contained in the residues of tree species (Carranca, 2013a), though Heckman and Kluchinski (1995) did not observe any allelophatic inhibition on nodulation or N2 fixation by soybean associated with heavy application to soil of oak, maple, sycamore or walnut leaves. Eucalyptus residues have substances that cause injures in the development of some leguminous plants, as well as the infection by the rhizobium. In fact, research regarding identification of allelopathic compounds in leachates of bark, fresh leaves and leaf litter of Eucalyptus tereticornis L., E. camadulensis L., E. polycarpa L. and E. microtheca L. showed a negative influence of phenolics and leachates on the germination, crop yield and nitrogenase activity of redgram [Cajanus cajan (L.) Millsp.], a pulse widely used in India (Sasikumar et al., 2001).

In this context, two field experiments were developed aiming i) to evaluate the rate and pattern of N2 fixation and N credit of white lupine (Lupinus albus L.) as affected by soil tillage, and of yellow lupine (Lupinus luteus L.) as affected by eucalyptus harvest residues management, and ii) to determine the transferred N2 fixed by L. luteus L. to the intercropped Eucalyptus globulus L. in a new plantation.


Material and Methods

Sites and experimental layout

Experiment A

A field experiment (A) was carried out in 2001/2002 and 2003/2004 at Pegões, central Portugal (38o 24’N, 8o 35’W), to evaluate the N2 fixing capacity and potential N credit of white lupine (Lupinus albus L.) in two blocks (190 m2 each), with two tillage treatments (conventional tillage (CT) and no-till (NT)), each treatment being replicated four times. The experimental field was established in a long-term use for fallow before the experiment started. Four plots (5 m x 1 m = 5 m2) were delimitated in each block for both legume and non-legume treatment, and till treatments were randomly distributed. No-till plots were not tilled at any time during the experiment, whereas CT plots were rotavated by chisel up to 20 cm depth just before seeding. The soil (Haplic Podzol) was sandy and up to 20 cm depth showed a pH(H2O) of about 6, a cation exchange capacity of 0.88 cmol(+) kg-1, and 7 g organic C kg-1 and 0.6 g Kjeldahl N kg-1. Basal dressing corresponding to 60 kg ha-1 for phosphorus (P) and for potassium (K) was applied just before tillage.

The climate of the region is of Mediterranean type, with hot and dry summers and mild autumns and winters. Average minimum and maximum air temperatures during both growth periods did not differ significantly and ranged from a minimum of 6ºC in December and January to a maximum of 23-26ºC in April, May and October. In January 2002, there were six days with temperature about 0ºC, with the formation of frost.

Monthly rainfall in 2001/2002 varied from a maximum value of 126 mm in January/02 to the lowest amount of 20 mm in February/02, and in 2003/2004 it ranged from 179 mm in October/03 to 29 mm in April/04.

A non-legume crop, oat (Avena sativa L.), was used as a reference crop to estimate the N2 fixed by the A-value method (Carranca et al., 1999b). A seeding rate of 60 plants m-2 was used for lupine and 12 g seeds m-2 for oat, and the sowing date was late September 2001 and 2003 for both crops. Legume nodulation was assessed by the native soil bacteria. In year 2003/2004, new blocks and plots were established as above. Ammonium nitrate doubly enriched with 5 atoms% 15N was annually applied in a liquid form to the plots, one week after plant emergence, at rates of 10 and 20 kg N ha-1 for lupine and oat, respectively. Molybdenum (Mo) and boron (B) were applied as topdressing in November of 2001 and 2003, at rates of 200 g Mo ha-1 and 3 kg B ha-1.

Weeds were chemically controlled by applying terbutrin before seeding the lupine. At post-emergence, weeds in each plot were controlled by hand. Glyphosate was applied in oat plots before seeding and “Oxytril M” was used at three-leaf stage. Standard methods were used for pests control during crop growth.

Experiment B

A field experiment (B) with yellow lupine (Lupinus luteus L.) was laid out at Óbidos west Portugal (39o 20’N, 9o 13’W) in 2004/2005 and 2005/2006, in an area where the soil (Haplic Podzols associated with Haplic Arenosols) was amended with eucalyptus (Eucalyptus globulus L.) harvest residues (leaves, branches, twigs, bark), to evaluate the N2 fixing capacity of legume and its potential credit N. The trial was designed in two blocks with two rates of eucalyptus residues (0 and 43 t ha-1). Each residue treatment was replicated four times. This site was also in a long-term use for fallow before the experiment. The soil (0-20 cm) was sandy and showed a pH(H2O) of 6.9, very low cation exchange capacity (0.58 cmol(+) kg-1), 3 g organic C kg-1 and 0.2 g Kjeldahl N kg-1. Harvest residues from a neighbour eucalyptus clear-felled site consisted of 20% leaves, 74% branches+twigs and 6% bark, and were chopped into pieces of 10 cm length and incorporated by harrowing in the soil at 0-20 cm depth, before sowing the yellow lupine. Amounts of mass and nutrients in harvest residue components are shown in Table 1. Basal dressing corresponding to 60 kg P ha-1 and 60 kg K ha-1 was applied after incorporation of harvest residues into the soil. After harvest residues incorporation in the soil, the C/N ratio in the top soil layer (0-10 cm) increased from 15 to about 26.



The climate at Óbidos site is of Mediterranean type, but tempered by the ocean proximity. Mean annual air temperature at the site, taking into account data from a meteorological station which existed in the neighborhood (Caldas da Rainha) is 15.2ºC (Reis and Gonçalves, 1981), ranging from a monthly average of 10.4ºC (January) to 19.8ºC (August). The 30-yr mean rainfall is about 610 mm, but less than 10% occurs between May and September. Relative air humidity is about 80% along the year, both at 9 A.M and 6 P.M.

Lupine was late sown in January 2005 at a rate of 60 plants m-2. Re-seeding of legume occurred naturally from shattering pods and plant germination took place by November 2005 for the second-year experiment.

Oat (Avena sativa L.) was used as a reference crop to estimate N2 fixed by the legume using the A-value method and was sown annually (January and November in 2004/2005 and 2005/2006, respectively) at a seeding rate of 12 g seeds m-2, in each amendment treatment and four replications. Root nodulation in legume was assessed by natural soil bacteria. Plots were marked for legume (3 m x 3 m) and oat (5 m x 1 m) in the two soil amendment treatments. In 2005/2006, new plots were marked for each soil amendment treatment. Ammonium nitrate doubly enriched with 5 atoms% 15 was annually applied in the liquid form to each plot, one week after plant emergence, at rates of 10 and 20 kg N ha-1 for lupine and oat, respectively.

Experiment C

A field trial (C) was laid out for one year (2005/2006) in the Óbidos site to evaluate the symbiotically fixed N2 transferred from yellow lupine (L. luteus L.) to the associated eucalyptus (Eucalyptus globulus L.) in a new plantation (3 m x3 m spacing), by comparison with sole eucalyptus in a forest system. The layout of the experiment was designed in two blocks with two soil amendments (0 and 43 t eucalyptus residue ha-1) and plots with three N fertilizer treatments (organic, mineral and control), which were replicated four times. Harvest residues collected in a neighbour eucalyptus clear-felled site consisted of 20% leaves, 74% branches+twigs and 6% bark, and were chopped into pieces of 10 cm length and incorporated in the soil by harrowing at 0-20 cm depth one year before (November 2004). The mass and nutrient amounts in residues are shown in Table 1. Basal dressing corresponding to 60 kg P ha-1 and 60 kg K ha-1 were applied after residue amendment.

In October 2005, one eucalyptus seedling was planted in the centre of each sub-plot (3 m x 3 m) containing the associated fodder lupine (organic fertilizer) or as a sole crop fertilized with mineral 15N. Control trees were also planted in each soil amendment as unfertilized sole crops. Lupine undergone natural re-vegetation and plant emergence occurred in November 2005.

One week after legume emergence, ammonium nitrate doubly enriched with 5 atom% 15N was distributed in each legume sub-plot, in the liquid form at a rate of 10 kg N ha-1, avoiding the area of direct tree influence (about 30 cm from the seedling). At the same time, the same fertilizer type was applied at a rate of 20 kg 15N ha-1 in 1 m x 1 m areas surrounding the sole trees planted. Trees in organic sub-plots (in both soil amendment plots) were not fertilized with mineral 15N.

Sampling, measurements and laboratory procedures

Experiments A and B

Complete plant samples (legume and oat) were taken from 15N labelled areas (1 m2) in each sub-plot, at different legume growth stages: 1-vegetative, 2-bloom, 3-pod formation, and 4-pod-filling or physiological maturity to establish the phenological phase of maximum fixation capacity. Plant material was separated into different plant organs (shoots, pods (when present) and visible roots (plus nodules in case of legume), which were washed with deionised water, dried at about 75ºC for 24 h to determine the DM yield (kg DM ha-1), and ground (<0.5 mm), avoiding any cross contamination, to evaluate the N concentration (g kg-1) and %15N enrichment by dry combustion in an automatic N analyzer-mass spectrometer (Universidad Autonoma de Madrid, Spain). Nitrogen yield (kg N ha-1) was calculated as the product of DM yield and N concentration. Percent N derived from the atmosphere (%Ndfa) was estimated by the Avalue method (Carranca et al., 1999b) using the reference value of 0.3663% for air natural 15N abundance.

Fixed N2 (kg N ha-1) was calculated as the product of %Ndfa and N yield. Potential contribution of legumes to soil N was calculated as the subtraction of the amount of total N2 fixed in the legume (kg N ha-1) from the amount of N yield (kg N ha-1) in the pods (seeds included) (Carranca et al., 1999a). A positive difference indicated a potential gain to soil N. Nitrogen harvest index (%NHI) was also estimated as the ratio between N in the pods+grain and aboveground plant N (Lecoeur and Sinclair, 2001).

Experiment C

One complete tree was harvested from each sub-plot at legume physiological maturity (June 2006) to establish the symbiotically fixed N2 transferred from lupine to the intercropped eucalyptus. The control trees were also harvested at the same time. Trees were separated into aboveand belowground plant material which was washed with deionised water and dried at about 75ºC for 48 h to evaluate dry matter yield (g DM tree-1). Plant samples were then ground (<0.5 mm), avoiding any cross contamination. In each plant sample, N concentration (g kg-1) and %15 abundance were determined as mentioned above. Plant N (mg tree-1) was calculated as the product of DM yield and N concentration. 15 transferred from lupine to the planted trees was evaluated by the atom% 15 excess determined in the tested trees minus the 15 in the control trees (unfertilized).

Percent N derived from mineral fertilizer (%Ndff) in the sole eucalyptus tested was determined as the ratio between atom% 15 excess (a.e.) in these trees and the %15 a.e. in the mineral N fertilizer (4.6337% a.e.) times hundred (Carranca et al., 1999b). 15 recovered in these trees (mg 15 tree-1) was estimated by the product of %Ndff and plant N (mg N tree-1). Fertilizer N use efficiency (%FUE) was estimated by the ratio between 15 recovery and mineral 15 fertilizer input (20 kg N ha-1, i.e., 2 g N m-2 tree-1) times hundred.

Statistical analysis

Analysis of Variance by General Linear Model using the Statistica 6.0 program was performed to evaluate the effects of factors (year, tillage, amendment, date of sampling, plant organ, fertilizer treatment) on measured variables. Means were compared using the Bonferroni’s test, at p<0.05.



Experiment A

Crop yield

Average yield (aboveground, pods, roots) of white lupine during the growth cycle was not significantly affected by soil disturbance (average 2.3 t DM ha-1) but was affected by climate, with a significantly higher biomass in 2003/2004 (2.6 t DM ha-1) compared to year 2001/2002 (2.2 t DM ha-1) (Table 2).



Crop N

Soil disturbance did not affect the average legume N concentration (28 g kg-1 DM) and %Ndfa (94%) which was mainly dependent on the year weather conditions, with a higher rate in 2003/2004 (99% Ndfa) compared to 2001/2002 (89% Ndfa). A significant reduction of %Ndfa was found in date 2 of 2001/2002 (86% Ndfa), increasing thereafter to the pod-filling in both years (100% Ndfa) (Table 2). Nitrogen concentration was significantly higher in shoots (33 g N kg-1 DM) and pods (29 g N kg-1 DM) compared to visible roots+nodules (22 g N kg-1 DM), and declined throughout the growth cycle from an average value of 37 g N kg-1 of total plant DM at vegetative phase to 21 g N kg-1 total DM at pod-filling stage. Nitrogen concentration was also significantly greater in year 2003/2004 (29 g kg-1DM) compared to year 2001/2002 (26 g kg-1DM).

Unlike total N, percent N derived from the atmosphere did not differ in shoots and visible roots+nodules (93% Ndfa), but reached 100% in pods (Table 2). Average fixation capacity in year 2003/2004 was significantly higher (70 kg N ha-1) than in 2001/2002 (61 kg N ha-1) with evidence of preferential allocation to the aboveground (shoots+pods+grain) biomass (98 kg N ha-1 year-1) compared to 7 kg N ha-1 year-1 in the visible roots+nodules (Table 2). Fixed N2 was substantially smaller in the first and second growth stages (29.5 kg N ha-1), compared to the pod-filling phase (115.5 kg N ha-1) sustained by the increase of crop yield and fixation rate.

N credit and harvest index

Because of the focus on aboveground biomass N, the contribution of grain legumes to N budget in rotation is generally underestimated. In this study, we included root N for the N balance. Lupine roots contained about 7% of aboveground N and accounted for 7% of total N2 fixed in the plant, and thus the potential N input was estimated by including shoots and visible roots+nodules biomass N. Nitrogen yield in pods at filling stage (9 kg N ha-1) was subtracted to estimate the credit N of white lupine. An apparent positive soil N input of +96 kg N ha-1 (93% of crop N) was estimated. At physiological maturity, a smaller N benefit by the sweet legume is expected since a higher N mobilization to the grain must occur. The apparent N harvest index determined at pod-filling stage in this pulse was only 9.4%, though the %NHI at maturity should result in a higher value.

Experiment B

Crop yield

Average DM yield of yellow lupine was significantly higher in 2005/2006 (4.4 t ha-1) compared to 2004/2005 (2.8 t ha-1) (Table 3), and in average about 1 t ha-1 greater than the white lupine biomass (Table 2). The incorporation of eucalyptus harvest residues in the soil did not affect the biomass produced by the fodder crop in both study periods (3.8 t DM ha-1) (Table 3), with a similar response of sweet lupine to tillage, but with a lower rate (Table 2).



Crop N

Incorporation of harvest residues did not affect the yellow lupine N concentration (27 g N kg-1 DM) and N2 fixation capacity (103 kg N ha-1) (Table 3). Fixation rate was also unaffected by the incorporation of harvest residues in the soil (99%). Greater N2 fixed was measured in shoots and pods (66 kg N ha-1 year-1), but lower than in white lupine (96 kg N ha-1 year-1).

Fixation increased since the vegetative phase (57 kg N ha-1) to the pod formation (144 kg N ha-1), than decreasing to the maturity phase (83 kg N ha-1). This seasonal variation showed a similar trend as in white lupine, though the N2 fixed at bloom in the yellow lupine was more related to the value measured at pod-forming in sweet lupine. Fixation rate by L. luteus at maturity (99% Ndfa) was slightly higher than by L. albus at pod-filling (95% Ndfa). Maximum fixation (100% Ndfa) by fodder legume was obtained at pod formation, earlier than in the sweet pulse (podfilling).

Credit N and N harvest index

Potential N benefit from L. luteus was estimated including the aboveand visible belowground biomass N at maturity and amounted to +69 kg N ha-1 (50% of crop N) showing a smaller fertilizer N value than sweet lupine at pod-filling stage, but still a high N input for the associated or subsequent crop in rotation.

Nitrogen harvest index was similar to the N credit (50%).

Experiment C

After eight months of growth, soil amendment did not affect the control eucalyptus seedlings which only accumulated 8 g total DM tree-1, and had a N concentration of 10.4 and 4.8 g N kg-1 DM, respectively in the top and root organ, and a natural 15N abundance of 0.006% and 0.015% a.e., respectively in the aboveand belowground plant material (data not shown). Dry matter yield (top+root) of fertilized 8-month old eucalyptus did not vary significantly with fertilizer type (organic or mineral) or with the 2-year soil amendment treatment, and presented a mean value of 32 g total DM tree-1 (Table 4), four times greater than control trees.



Tree N also did not vary significantly with fertilizer treatment (276 mg N tree-1) whereas woody residue in soil reduced significantly the exported N (188 mg N tree-1) compared to 366 mg N tree-1 in unamended sub-plots (Table 4). These results are mainly a consequence of tree biomass since the N concentration responded positively to mineral N fertilization (10.4 g kg-1 DM), in contrast with the organic treatment (7.5 g kg-1 DM), and only slightly to soil amendment (8.1 g kg-1 DM) in comparison with untreated sub-plots (9.8 g kg-1 DM). Residues in soil under organic fertilization did not significantly affect the total C assimilated (8.6 g C tree-1) (Table 4).



Fixation capacity and potential soil N input by lupine species

Results of the present study confirm that systems involving legumes represent a sustainable alternative to conventional agriculture by capture of atmospheric N2, being an adequate strategy for organic fertilization and rotation efficiency improvement. In fact, fixation rate by sweet lupine was high (>89%). The fodder crop with shattering pods was supplied with N2 from the air considerably better than the sweet pulse (Table 3), as already observed by Merbach and Schilling (1980). Both studied lupine species, for human diet and fodder production fixed N2 from the atmosphere very efficiently (94-99% Ndfa) in symbiosis with the indigenous rhizobium in coarse textured soils, poor in organic matter (OM) and available N.

The difference observed in %Ndfa for the two-year experiment was mainly due to the low value (86% Ndfa) determined in the second sampling date in year 1 (flowering stage, in January), which may be explained by the occurrence of six days of extremely low soil temperature with frost. The lower fixation rate by sweet pulse in the growth stage 2 and 3, respectively in January and February (86-92%), when the daily mean air temperatures were as low as 5-6ºC, agrees with the value of 92% Ndfa measured by Pálmason et al. (1992) for sweet-blue lupine cropped at the mean temperature of 10ºC. In fact, low temperature reduces plant growth and controls the geographical distribution of soil rhizobia (Breitbarth et al., 2006), which are mesophiles and mostly do not grow below 10ºC and nodule growth is restricted (Carranca et al., 2009; Carranca, 2013a).

In the present study, fixation rate increased from the vegetative phase (one month after sowing) to the pod formation/filling stage (100% Ndfa) in both lupine species, in much larger rates than those observed for other grain legumes (Vicia sativa L., Vicia faba L., Ciccer arietinum L.) cropped under similar rain fed ecological conditions, which fixed 30-82% Ndfa (Carranca et al., 1999a). Fixation rates observed in the present study are also higher than those reported by Howieson et al. (1998) for L. albus L. (68-85% Ndfa in Australia, 44-92% in Germany, and 80% in France), by Chalk et al. (1993) for L. angustifolius L. (79% Ndfa), and by Halvorson et al. (1992) for L. lepidus and L. latifolius (60% Ndfa). The high fixation rates observed in experiments A and B may were favoured by poor soil N and the presence of efficient indigenous soil bacteria with the tested local lupine varieties.

Merbach and Schilling (1980) reported that the yellow lupine fixed N2 up to the ripeness, but the white lupine on the contrary finished fixation in the flowering phase by a possible competition of growing seeds and nodules for the photosynthates, and legume should be supplied with mineral N at this phenological stage. This trend was not observed in the present study (experiment A, see Table 2) as the rate of fixation by white lupine significantly increased from the flowering phase (92% Ndfa) to the pod-filling stage (100% Ndfa), and the NHI was very low (9.4%), showing that N mobilization in the plant to the grain may be later than this phase.

Disturbance associated with soil tillage did not reduce the symbiotic fixation by sweet pulse (94% Ndfa) because microbial communities were apparently not greatly affected, though a slightly higher N2 fixation capacity (6% greater) was measured in conservative practice in line with the accumulation of plant biomass (Table 2); also, no great soil disruption occurred in the sandy soil by tillage since aggregation in this soil is negligible. Moreover, N fluxes resulting from mineralization were probably not relevant to inhibit fixation since the soil only contained 12 g OM kg-1. Horn et al. (1996) and Shah et al. (2004) did not find any significant effect of tillage system on the rate of N2 fixation by chickpea (Ciccer arietinum L.) and lentil (Lens culinaris Medikus), respectively.

However, former authors observed a greater N accumulation in no-till (66 kg N ha-1) comparing with conventional tillage (44 kg N ha-1) system, while the latter found an apparently greater crop N yield under tillage than under no-till. Incorporation of harvest residues from Eucalyptus globulus L into the soil, at Óbidos site (experiment B), showed no significant effect on the fixation rate by L. luteus (99% Ndfa) with the indigenous soil rhizobium (Table 3), and only a 3% reduction of fixed N2 was noted in presence of those residues. This indicates that allelopathic compounds of those residues were not able to depress the leguminous growth and symbiosis, following the trend reported by Heckman and Kluchinski (1995) for soil amended with oak, maple, sycamore or walnut leaves, where no indication of an allelopathic inhibition on nodulation or N2 fixation by soybean (Glycine max. (L.) Merr.) was observed. Our results are not in agreement with effects of allelopathic compounds in leachates of bark, fresh leaves and leaf litter of E. tereticornis L., E. camadulensis L., E. polycarpa L. and E. microtheca L., which showed a negative influence on the germination, DM yield and nitrogenase activity of redgram [Cajanus cajan (L.) Mills] (Sasikumar et al., 2001). Findings of the present study suggest that E. globulus L. harvest residues may be innocuous for the indigenous rhizobium and legume germination and growth.

A great reduction in shoot N concentration during the growth cycle of sweet lupine (41 to 26 g N kg-1 shoot DM, from vegetative to pod-filling phases, respectively) (data not shown) was associated with a dilution effect caused by the crop development and the withdrawn of N to the reproductive organs (29 g kg-1 pod DM) after the onset of the flowering stage. A decrease in N concentration was also observed in roots during the growth cycle (32 to 12 g kg-1 DM) mainly by the dilution effect, roots decay, and nodules detachment and necrosis. Similarly to L. albus L., shoot N concentration in yellow lupine decreased with the plant age from the vegetative phase to maturity (37 to 10 g kg-1 DM) (data not shown), whereas the pod N increased from pod-filling to the harvest (32 to 62 g kg-1 DM, respectively). Also, the N concentration decreased from 39 to 16 g kg-1 DM in roots during the growth cycle.

This trend is in agreement with Howieson et al. (1998) who reported a decrease of shoot N from 44 at vegetative to 15 g N kg-1 DM at ripening stages by L. angustifolius L.. The distribution of N in shoots, roots and pods has important implications for the %NHI and fertilizer N value of legume residue for the associated crop or the subsequent crop in a rotation. Nitrogen harvest indices were 9% and 50%, respectively for the white and yellow lupine species, though in the case of the former the estimation was done at pod-filling stage and it should result in a higher NHI value when calculated at maturity stage.

According to Jensen and Hauggaard-Nielsen (2003), the incorporation of residues from grain legumes contributes to maintain soil N level, but the overall benefit is much less than from legume-based pastures because of high NHI in pulses. But because of the focus on aboveground biomass N, the contribution of grain legumes to the N budget in rotation is generally underestimated. In addition, maximum contribution to the soil N pool occurs when the legume is 100% symbiotically dependent, since in this situation soil N is not taken up by the legume. Nitrogen output in the pods+grain of L. luteus at maturity (68 kg N ha-1) was used to evaluate the N credit of this crop. Visible roots+nodules contained about 4% of aboveground N and accounted for 4% of total N2 fixed in the crop.

The potential soil N input by this pulse was estimated by including the aboveand belowground visible roots-nodules at the end of ripeness and amounted to +69 kg N ha-1 (50% of crop N) showing a positive fertilizer N value for associated crop or subsequent crop in rotation. Considering the pods+grain at podfilling stage of white lupine, a soil N input of +96 kg N ha-1 (93% of crop N) was estimated, but a smaller gain of N to the soil at crop maturity is expected due to a higher %NHI in the crop at this phase, caused by a greater N mobilization in the plant from shoots to pods. If crops are grown for the sole purpose of providing N in organic farming, the N contribution of white pulse at pod-filling (100% Ndfa) may reach +110 kg N ha-1 (see Table 2) and +144 kg N ha-1 (100% Ndfa) by the fodder lupine at pod formation (see Table 3). These values are of the same magnitude or even greater than the maximum values reported by Howieson et al. (1998), Jensen et al. (2004) and De Varennes et al. (2007) who found a soil N benefit by white lupine in Europe and Australia varying from +32 to +96 kg N ha-1, depending on the growth phase, and do not agree with the N deficit found in North America (–41 kg N ha-1), which can be explained by the exclusion of fallen leaves in the estimate of N2 fixation at maturity (Howieson et al., 1998). These au-thors also reported a net N input by L. angustifolius L. (+95 kg N ha-1) similar to that estimated in the present study for L. albus L. at pod-filling, whereas Chalk et al. (1993) reported an annual input to the soil N by L. angustifolius L. of only +34 kg N ha-1.

Mayer et al. (2004) found a large variation (+7 to +64 kg N ha-1) of potential N derived from sweet pulse residues (stems, leaves, visible and decayed roots, and rhizodeposits). Estimates of N balances based on standard root harvesting data may be confounded by neglecting decayed roots, detached nodules and rhizodeposition (exsudates and products of root and nodule necrosis), which may add significant amounts of N to the soil (22-68% of the total plant N for pulses), as supported by Fuhrer (2004), Yasmin et al. (2006), De Varennes et al. (2007), Mathieu et al. (2007) and McNeill and Fillery (2008). We assume that N inputs by study pulse crops can exceed the aforementioned values, as decayed roots and rhizodeposits were not accounted for in the estimation of N benefit.

Transfer of symbiotically fixed N2 from yellow lupine to eucalyptus seedlings

Nitrogen concentration in the newly-planted E. globulus L. was 19% lower in harvest residues amended area (Table 4) in comparison with trees planted in unamended subplots, which may be associated with soil N immobilization caused by the relatively high soil C/N (26) after harvest residues incorporation. On the other hand, N concentration in mineral N treated eucalyptus seedlings was significantly higher than in organically (leguminous) fertilized seedlings (see Table 4), for the average effect of soil amendment, but exported N did not differ significantly between both treatments (276 mg N tree-1). In organically fertilized seedlings, the 15N a.e. coming from labelled fallen leaves, roots, nodules and root rhizodeposits of associated fodder legume was about thirteen times smaller than the 15 a.e. in mineral treated seedlings (Table 4). Nevertheless, the N transferred from legume to associated seedlings showed a potential benefit of slow N transfer from L. luteus to the associated young eucalyptus for medium term productivity.

The low FUE (4%) estimated for eucalyptus is in line with the values observed in drip-fertigated orchards of 8-month-old orange and pear trees (6%), planted in similar environmental conditions, by Menino et al. (2007) and Neto et al. (2008). These low FUE values can by explained by the small root development in transplanted plants, where the ratio between the aboveand belowground biomass was 4.3 in eucalyptus and 1.5 in fruiting trees. Despite the low FUE values observed for Eucalyptus seedlings, it is expected a positive Lupinus effect at a later stage of forest plantations. In fact, in a field experiment, close to the site of the present study, foliar N and P concentrations were greater in 5-yr old trees which were associated with Lupinus (Madeira et al., 2010) during the early plantation phase, following results observed in tropical regions for mixed forest plantations (J. Leonardo Gonçalves, personal communication).

Although there is a high N input by Lupinus species, their use to improve tree growth in forest plantations is questionable under Mediterranean conditions. In fact, in the aforementioned field experiment, the intercropped Lupinus negatively affected the early eucalyptus growth either when the harvest residues were incorporated or removed from the soil (Madeira et al., 2010), where the understory biomass in treatments with yellow lupine was 2.1-7.5 times greater than in other treatments. This early tree growth suppression agrees with results observed in other studies where leguminous cover reduced tree growth, especially when the drought season was more severe (Malik et al., 2001; Mendham et al., 2004). Such trend may be related to competition between trees and understory vegetation for water resources, as eucalyptus root mass, at early growth stage, is mostly located in the 30-cm top soil layers (Jones et al., 1999) and therefore may be more sensitive to summer drought in Mediterranean conditions. However, the negative effect of Lupinus on tree growth decreased with time, which may be associated with reduced competition between trees and the cover crop, as tree roots may have explored deeper layers (Fabião et al., 1987; Madeira et al., 2010). In fact, at middle rotation (five years after plantation) of the aforementioned field study, a positive effect of legumes was observed, in the absence of harvest residues, which may be ascribed to inputs of N through leguminous and P due to superphosphate application (Madeira et al., 2010).

Although small or nil Lupinus effects are reported on tree growth under Mediterranean conditions, it is expected a positive Lupinus effect at a later stage of forest plantations by improving at least nutrient availability and soil quality (Madeira et al., 2010).

Therefore, sharing of N sources between the intercropped N2 fixing plants and the non-fixing trees may contribute to a better use of N natural resources and to counteract nutrient removal by harvesting.

Also at the early forest plantation phase, legumes plants may reduce leaching of N and base cations which tend to occur after tree harvesting.



Under Mediterranean conditions, white and yellow lupine species behaved as beneficial legumes when cropped in sandy soils, low in OM and available N.

Tillage showed little impact on fixation rate by these pulses, but no-till slightly increased total amount of fixed N2 by increasing plant biomass. Eucalyptus harvest residues in soil did not depress the legume growth, and activity or efficiency of symbiotic bacteria, and only induced an apparent reduction on symbiotic dependence. Although rhizodeposits were not accounted for the determination of soil N benefit by the two lupine species in the present study, relevant N inputs were found (>+69 kg N ha-1), which are of great importance in a legume/cereal rotation or an association with non-legume crops, particularly when legume was cropped for organic farming and fixation was 100% (fixation at pod formation/pod-filling stages). Symbiotically transferred N2 from the fodder legume to newly planted E. globulus L. was apparently favourable for the sustainability of forest systems.

Results emphasized the need of continuing the study to confirm the positive effect of using the fodder lupine as organic supplier for the eucalyptus planted in soils with a low OM status, especially looking into the influence of different rooting patterns in the early forest plantation development. Also, studies on possible benefits of soil amendment with harvest residues from eucalyptus plantations should proceed in order to evaluate the physical and chemical changes in soil, and its long-term influence on symbiotic N2 fixation.



This study was supported by the Portuguese government through the PIDDAC 166/01 project, and the European Union through POCTI/AGG/42616/01 and POCTI/AGG/42825/01 projects. Centro Experimental de Pegões and CELBI company are acknowledged by providing experimental areas at Pegões and Óbidos, respectively. Both João Pedro Pina and Mário Abrantes are also acknowledged for helping on the installation and management of experimental site at Óbidos. Acknowledgements are also due to the staff of soil laboratory of Instituto Superior de Agronomia for analyses of harvest residues and soil from Óbidos site, and staff of ex-LQARS for analyses of soil from Pegões site.


Bibliographic References

Breitbarth, E.; Oschlies, A. and La Roche, J. (2006) Physiological constraints on the global distribution of Trichodesmium – effect of temperature on diazotrophy. Biogeosciences Discussion, 3: 779-801.         [ Links ]

Carranca, C. (2013a) Legumes: Properties and symbiosis. In: Camisão, A.H. and Pedroso, C.C. (Eds.).- Symbiosis: Evolution, Biology and Ecological Effects. New York, Nova Science Publishers, p. 67-94.         [ Links ]

Carranca, C. (2013b) The role of legumes for sustainability: Research needs for future prospects. In: Camisão, A.H. and Pedroso, C.C. (Eds.). - Symbiosis: Evolution, Biology and Ecological Effects. New York, Nova Science Publishers, p. 177-188.         [ Links ]

Carranca, C.; De Varennes, A. and Rolston, D.E. (1999a) Biological nitrogen fixation by fababean, pea, and chickpea under field conditions estimated by the 15 isotope dilution technique. European Journal of Agronomy, 10: 49-56.         [ Links ]

Carranca, C.; De Varennes, A. and Rolston, D.E. (1999b) Biological nitrogen fixation estimated by the 15 dilution, natural 15 abundance and N balance in a subterranean clover-grass sward under mediterranean conditions. European Journal of Agronomy, 10: 81-89.         [ Links ]

Carranca, C.; Torres, M.O. and Baeta, J. (2009) White lupine as a beneficial crop in Southern Europe. I Potential for N mineralization in lupine amended soil and yield and N2 fixation by white lupine. European Journal of Agronomy, 31: 183-189.         [ Links ]

Chalk, P.M.; Smith, C.J.; Hamilton, S.D. and Hopmans, P. (1993) Characterization of the N benefit of a grain legume (Lupinus angustifolius L.) to a cereal (Hordeum vulgare L.) by an in situ 15 isotope dilution technique. Biology and Fertility of Soils, 15: 39-44.         [ Links ]

Crépon, K.; Carrouée, B. and Schneider, A. (2004) Grain legumes and the environment: how to assess benefits and impacts? In: Abstracts of the International Workshop on the Methodology for Environmental Assessment of Grain Legumes. Swiss Federal Research Station, European Association for Agriculture and Nature Grain Legumes Research. Zurich, Agroscope FAL Reckenholz, p. 9-10.         [ Links ]

Criozat, Y. (2004) Session B. Grain legumes in crop rotations. Assessing the role of grain legumes in crop rotation: some agronomical concepts that can help! In: Abstracts of the International Workshop on the Methodology for Environmental Assessment of Grain Legumes. Swiss Federal Research Station, European Association for Agriculture and Nature Grain Legumes Research. Zurich, Agroscope FAL Reckenholz, p. 22-23.         [ Links ]

De Varennes, A.; Torres, M.O.; Queda, C.; Goss, M.J. and Carranca, C. (2007) Nitrogen conservation in soil and crop residues as affected by crop rotation and soil disturbance under Mediterranean conditions. Biology and Fertility of Soils, 44: 49-58.         [ Links ]

Fabião, A.; Madeira, M. and Steen, E. (1987) Root mass in plantations of Eucalyptus globulus in Portugal in relation to soil characteristics. Arid Soil Research and Rehabilitation, 1:185–194        [ Links ]

Ferguson, J.J. and Rathinasabapathi, B. (2009) - Allelopathy: How plants suppress other plants (online). University of Florida, IFAS Extension, HS 944, 3 p. Available at .         [ Links ]

Fuhrer, J. (2004) Session C. Nitrogen cycle of grain legume crops. Environmental aspects of the nitrogen cycle in legume-based cropping systems. In: Abstracts of the International Workshop on the Methodology for Environmental Assessment of Grain Legumes. Swiss Federal Research Station, European Association for Agriculture and Nature Grain Legumes Research. Zurich, Agroscope FAL Reckenholz, p. 30-31.         [ Links ]

Gaillard, G. and Nemecek, T. (2004) Environmental assessment of grain legumes: The role of a sound system analysis. In: Abstracts of the International Workshop on the Methodology for Environmental Assessment of Grain Legumes. Swiss Federal Research Station, European Association for Agriculture and Nature Grain Legumes Research. Zurich, Agroscope FAL Reckenholz, p. 13-14.         [ Links ]

Goss, M. and De Varennes, A. (2002) Soil disturbance reduces the efficacy of mycorrhizal associations for early soybean growth and N2 fixation. Soil Biology & Biochemistry, 34: 1167-1173.         [ Links ]

Halvorson, J.J.; Franz, E.H.; Smith, J.L. and Black, R.A. (1992) Nitrogenase activity, nitrogen fixation, and nitrogen inputs by lupines at Mount St. Helen. Ecology, 73: 87-98.         [ Links ]

Heckman, J.R. and Kluchinski, D. (1995) Soybean nodulation and nitrogen fixation on soil amended with plant residues. Biology and Fertility of Soils, 20: 284-288.         [ Links ]

Horn, C.P.; Dalal, R.C.; Birch, C.J. and Doughton, J.A. (2007) - Nitrogen fixation in chickpea as affected by planting time and tillage practice. 1996. Available at         [ Links ]

Howieson, J.G.; Fillery, I.R.P.; Legock, A.; Sjkorski, M.M.; Stepkowski, T.; Minchuir, F.R. and Dilworth, M.J. (1998) Nodulation, nitrogen fixation and nitrogen balance. In: Gladstones, J.S.; Atkins, C. and Hamblin, J. (Eds.) - Lupines as crop plants. Biology, Production and Utilization. New York, CAB International, p. 149-180.         [ Links ]

Jensen, E.S. (2004) Session B. Grain legumes in crop rotations. Grain legume functions in farming systems. In: Abstracts of the International Workshop on the Methodology for Environmental Assessment of Grain Legumes. Swiss Federal Research Station, European Association for Agriculture and Nature Grain Legumes Research. Zurich, Agroscope FAL Reckenholz, p. 20-21.         [ Links ]

Jensen, E.S. and Hauggaard-Nielsen, H. (2003) How can increased use of biological N2 fixation in agriculture benefit the environment?Plant and Soil, 252: 177-186.         [ Links ]

Jones, H.E.; Madeira, M.; Herraez, L.; Dighton, J.; Fabião, A.; González-Rio, F.; Fernandez-Marcos, M.; Gomez, C.; Tomé, M.; Feith, H.; Magalhães, M.C. and Howson, G. (1999) The effect of organic matter management on the productivity of Eucalyptus globulus stands in Spain and Portugal: tree growth and harvest residue decomposition in relation to site and treatment. Forest Ecology and Management, 122:73–86        [ Links ]

Lecoeur, J. and Sinclair, T.R. (2001) Nitrogen accumulation, partitioning, and nitrogen harvest index increase during seed fill of field pea.Field Crops Research, 71: 87-99.         [ Links ]

Madeira, A.C.; Madeira, M.; Fabião, A.; Carneiro, M. and Marques, P. (2010) Effects of harvest residues, N-fixing plants or fertilization management on the growth and nutritional status of young Eucalyptus globulus plantations. European Journal of Forest Research, 129, 49: 591-601.         [ Links ]

Malik, R.K.; Green, T.H.; Brown, G.F.; Beryl, C.A.; Sistani, K.R. and Maes, D.A. (2001) Biomass production of short-rotation bioenergy hardwood plantations affected by cover crops. Biomass Bioenergy, 21:21–33        [ Links ]

Mathieu, S.; Fustec, J.; Faure, M-L.; Corre-Hellou, G. and Crozat, Y. (2007) Comparison of two 15 labelling methods for assessing nitrogen rhizodeposition of pea. Plant and Soil, 295: 193-205.         [ Links ]

Mayer, J.; Buegger, F.; Jensen, E.S.; Schloter, M. and He, J. (2004) Session C. Nitrogen ycle of grain legume crops. The role of N rhizodeposition of grain legumes in the N cycle of crop rotations. In: Abstracts of the International Workshop on the Methodology for Environmental Assessment of Grain Legumes. Swiss Federal Research Station, European Association for Agriculture and Nature Grain Legumes Research. Zurich, Agroscope FAL Reckenholz, p. 40-41.         [ Links ]

McNeill, A.M. and Fillery, I.R.P. (2008) Field measurement of lupin belowground nitrogen accumulation and recovery in the subsequent cereal-soil system in a semi-arid Mediterranean-type climate. Plant and Soil, 302: 297-316.         [ Links ]

Mendham, D.S.; Kumarawamy, S.; Balasundaran, M.; Sankaran, K.V.; Corbeels, M.; Grove, T.S.; O´Connell, A.M. and Rance, S.J. (2004) Legume cover cropping effects on early growth and soil nitrogen supply in eucalypt plantations in southwestern India. Biol Fertil Soils, 39:375–382        [ Links ]

Menino, M.R.; Carranca, C. and De Varennes, A. (2007) Distribution and remobilization of nitrogen in young non-bearing orange trees grown under Mediterranean conditions. Journal of Plant Nutrition, 30: 1083-1096.         [ Links ]

Merbach, W. and Schilling, G. (1980) Effectiveness of symbiotic N2-fixation in leguminous plants, as affected by inoculation with rhizobia, by substrate, N-fertilizing, and 14C-sucrose application. Zentralbl Bakteriol Naturwiss, 135: 99-118.         [ Links ]

Neto, C.; Carranca, C.; Clemente, J. and De Varennes, A. (2008) Nitrogen distribution, remobilization and re-cycling in young non-bearing ‘Rocha’ pear trees. Scientia Horticulturae, 118: 299–307.         [ Links ]

Pálmason, F.; Danso, S.K.A. and Hardarson, G. (1992) Nitrogen accumulation in sole and mixed stands of sweet-blue lupin (Lupinus angustifolius L.), ryegrass and oats.Plant and Soil, 142: 135-142.         [ Links ]

Reis, R.M. and Gonçalves, M.Z. (1981) – Caracterização climática da região agrícola Ribatejo e Oeste. Lisboa, Instituto Nacional de Meteorologia e Geofísica. (O Clima de Portugal, fasc. Vol. XXXII).         [ Links ]

Rocha, M.E.N. and Santos, C.L. (2007) O uso comercial e popular do eucalipto. Eucalyptus globulus Labyll-Myrtaceae. Saúde & Ambiente em Revista, 2: 23-34.         [ Links ]

Sasikumar, K.; Vijayalakshmi, C. and Parthiban, K.T. (2001) Allelophatic effects of four eucalytptus species on redgram (Cajanus cajan L.). Journal of Tropical Agriculture, 39: 134-138.         [ Links ]

Shah, Z.; Mohammad, W.; Shah, S.M. and Iqbal, M.M. (2004) Yields and N2 fixation of lentil and chickpea as influenced by tillage and P levels under rainfed conditions. Sarhad Journal of Agriculture, 20: 243-249.         [ Links ]

Yasmin, K.; Cadish, G. and Baggs, E.M. (2006) Comparing 15-labelling techniques for enriching aboveand below-ground components of the plant-soil system. Soil Biology & Biochemistry, 38: 397-400.         [ Links ]


Recebido/Received: 2012.06.28

Aceitação/Accepted: 2012.10.22

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License