The level of N₂-fixation of different genotypes of winter pea in comparison to spring pea in pure and mixed stands

Die Höhe der N₂-Fixierung verschiedener Wintererbsengenotypen im Vergleich zu Sommererbsen in Rein- und Gemengesaat

The main plot factor consisted of pea pure stands and two replacement designs with either 25% (Mix1) or 50% (Mix2) of the pea pure stand sowing density, complemented by either 75% or 50% of the cereal pure stand sowing density. Sowing density in pure stands (PS) was 80 germinable pea grains m^–2 and 300 (DFH) or 380 (WH) germinable cereal grains m^–2, respectively. As a partner for winter pea in mixture with cereal, winter rye (Secale cereale L.), cv. Amilo was selected in 2003/04 and cv. Danko from 2004/05 to 2006/07, while for spring pea oats (Avena sativa L.), cv. Aragon was the mixture partner in 2003/04 and barley (Hordeum vulgare L.), cv. Ria between 2004/05 and 2006/07.

Experimental plots were ploughed and prepared with a cultivator just before sowing the crop. Distance between rows was 21 cm at DFH and 15 cm at WH. Sowing depth was approx. 4 cm. Prior to sowing the spring crops, soil was prepared twice with a rotary cultivator and – if necessary, depending on weed cover – with a chisel plough beforehand.

The grain harvest was carried out with a Hege 140 plot combine (Wintersteiger AG, Egging am See, Germany) in Frankenhausen and with a Hege 160 plot combine in Waldhof. Winter crops were usually ready for harvest one week earlier than the spring crops (Table 1). In summer 2003/04, harvest of spring crops was delayed by the late ripening of oats, which subsequently was replaced by barley. Yield of above-ground residues without stubble was determined at DFH at grain maturity from 1.5 m² that were taken by cutting plants approx. 5 cm above the ground and put under a roof until dry enough for threshing. Crops were also threshed with the Hege 140 plot combine. At WH, the above-ground residues were sampled from the whole plot immediately after combine harvest.

Soil samples at biomass and mature harvest were taken at 0–30, 30–60, 60–90 cm soil depth at DFH, and only up to 60 cm soil depth at WH due to the pebbly subsoil being too dry for sampling (except in autumn).

The yield stability of above-ground biomass at flowering and at mature harvest of all treatments was calcu­lated with the coefficient of variation (CV) according to Francis and Kannenberg (1978):

CV (%) = 100 × standard deviation/overall mean

The amount of N₂ fixed (Nfix) was calculated for pure stands (Equation 1) according to Stülpnagel (1982) and for mixed stands (Equation 2) according to Karpenstein-Machan and Stülpnagel (2000) at flowering and mature harvest with the extended total-N-difference method. Weeds were harvested with the crops and taken into account.

(1) N-Fix_PS = N-uptake biomass_PPS – N-uptake biomass_CPS + N_min-N-soil_PPS – N_min-N-soil_CPS

(2) N-Fix_Mix = N-uptake biomass_Mix – N-uptake biomass_CPS + N_min-N-soil_Mix – N_min-N-soil_CPS

With PS = pure stand; PPS = pea pure stand; CPS = cereal pure stand; Mix = mixed stand

The extended total-N-difference method was selected for several reasons: Using this method makes estimation of N₂ fixation rather inexpensive and can be conducted with little technical equipment (Danso, 1995; Unkovich and Pate, 2000). Moreover, at a limited soil N supply as is generally the case in organic farming, there is a good correspondence between the simple total-N-difference method and the isotope-dilution method (Rennie, 1984; Danso, 1995). Taking into account inorganic N in soil increases the accuracy of estimation as compared to the simple total-N-difference method (Stülpnagel, 1982; Loges, 1998), because it allows consideration of the different N uptake from grain legumes and cereal.

The portion (%) of N derived from atmospheric fixation in total N uptake (NdfA) was determined for pure stands (equation 3). In mixed stands, calculation of NdfA had to be omitted because the aboveground residues of rye and winter pea could not be separated due to the pea’s high longitudinal growth up to three meters with several branches (Urbatzka, 2010).

(3) NdfA_PPS (%) = 100 × N-Fix_PPS/N-uptake biomass_PPS

The analysis of variance across years was confined to the growing seasons 2004/05 until 2006/07, due to a change of the winter and spring cereal mixture partner. Due to the mixed experimental designs used from 2004/05 until 2006/07, treatment effects were analyzed with dummies according to the method supplied by Piepho et al. (2006). Factors year and location were combined, yielding a fixed factor environment, to allow a common analysis of data from the two experimental sites. Further fixed factors were crop stands and genotype.

Fig. 1. N-yield of above-ground biomass and inorganic N in soil (0–90 cm) as affected by crop stands and date of sampling (onset of flowering (BBCH 61) or grain maturity (BBCH 89)) (a) DFH 2003/04, (b) DFH 2004/05, (c) DFH 2005/06;
Different letters denote significant differences between pea genotypes (Tukey test, p < 0.05): capital letters for N_min-N, small letters for N uptake, n.s. = not significant, ¹ log-transformation, * no significant interaction of cultivar and crop stand in F-test, results for aboveground residues at grain maturity in DFH 2003/04 were deduced from other years at DFH via N harvest index; EF = EFB 33, Un = Unrra, Ni = Nischkes Riesengebirgs, GR = Griechische, Sa = Santana (spring pea), Ry = rye, Ba = barley.

Fig. 2. N-yield of above-ground biomass and inorganic N in soil (0–90 at DFH and 0–60 at WH) as affected by crop stand and date of sampling (onset of flowering (BBCH 61) or grain maturity (BBCH 89)) (a) DFH 2006/07, (b) WH 2005/06, (c) WH 2006/07;
See legend of Fig. 1; ² Square-root-transformation, # losses due to mice, results for aboveground residues at grain maturity in WH 2005/06 were deduced from other year at WH via N harvest index.

The cultivars Assas and Cheyenne were winter killed or greatly damaged over winter in three and four out of six environments, respectively (data not shown). Hence the analysis of variance was conducted with only six genotypes. At DFH, pure stands of winter peas were damaged severely by mice in 2006/07 and as a consequence mature harvest and calculation of N₂ fixation as well as NdfA had to be omitted.

Spring crops responded more strongly to the environmental conditions than winter crops (Table 5). The highest coefficient of variation for winter crops was determined for winter rye. N uptake of winter cereal pure stands varied most markedly (in descending order): DFH in 2005/06 and 2006/07 (94 kg N ha^–1) > DFH in 2003/04 and 2004/05 (60 kg N ha^–1) > WH (30 kg N ha^–1) (Fig. 1, 2).

Usually highest biomass yield stability was measured in mixture of winter peas (Table 5). Similar to harvest at flowering the coefficient of variation in winter crops was lower than in spring crops and in winter pea pure stands lower than in rye pure stands. Likewise, rye in pure stands yielded (in descending order, similar to flowering harvest): at DFH in 2005/06 and 2006/07 (178 kg N ha^–1) > at DFH in 2003/04 and 2004/05 (110 kg N ha^–1) > at WH (76 kg N ha^–1) (Fig. 1, 2).

During crop growth, not all treatments achieved an increase of N₂ fixation between flowering and mature harvest (Table 6, 7), which especially applied for winter pea pure stands in four out of five environments, but also for all spring pea treatments in one environment (Table 6d, 7d). At mature harvest N₂ fixation of spring peas in pure stands was in five out of six environments considerably higher than in mixtures (exception shown in Table 7d), but only in two out of five for winter peas (Table 7b, e). In mixture, winter pea variants showed considerably higher N₂ fixation than spring peas in five out of six environments (Table 7a–c, e, f), but only in two out of five for pure stands (Table 7a, e). Again at DFH in 2005/06 and 2006/07 in comparison to the earlier growing seasons higher N₂ fixation was usually measured for winter peas, but in contrast to harvest at flowering above all in mixtures with on average 76 kg ha^–1 lower (Table 6a–d, 7a–d).

At flowering harvest, the lowest NdfA of 45 and 55% was measured at DFH in 2005/06 and 2006/07, respectively, which was lower than in the other four envi­ronments (64–93%) (Table 8). Between flowering and mature harvest, NdfA decreased in three out of five environments and remained on the same level in the other two environments. At WH both at flowering and mature harvest the NdfA was higher than at DFH. Besides, no differences were found between pea genotypes: Ndfa at flowering harvest (66%) was slightly higher than at mature harvest (62%) (Table 9).

4 Discussion

The strong increase of inorganic N until mature harvest under winter peas in pure stands in comparison to spring pea pure stands was in accordance with the observations by Karpenstein-Machan and Stülpnagel (2000) and can not only be explained by N-sparing effects, but also by early mineralization of above ground debris, N-rhizodeposition (Urbatzka et al., 2009) and the subsequent N priming effects (Jenkinson et al., 1985). In contrast, the higher inorganic N under spring peas in pure stands – in comparison to mixed stands – was probably mainly caused by N-sparing effects (Urbatzka et al., 2009). Inorganic N was markedly lower in 2005/06 (Fig. 1c, 2b), probably as a result of the relatively dry weather conditions during June and July (Table 2, 3).

Moreover, the consistently lower availability of inorganic N under cereals in both pure and mixed stands compared with pea pure stands may be a consequence of the different root systems of the partners in mixture (Fig. 1, 2). It is widely-known that cereal plants have a higher root mass and length and display a deeper root soil penetration (e.g. Hamblin and Tennant, 1987; Schmidtke and Rauber, 2000).

The slightly negative values for N₂ fixation (Table 6) can be traced back to the fact that the roots of the plants were not taken into account by using the extended total-N-difference method. Recent research showed that below-ground N from roots and N-rhizodeposition represent a substantial part of legume N (Mayer et al., 2003; Wichern et al., 2007). Hence in accordance to Peoples et al. (2002) indirect measurements considerably underestimate N₂ fixation of legumes.

In contrast to spring pea, the apparent reduction in N₂ fixation of winter peas in pure stands in most environments was a consequence of the loss of biomass during vegetative growth (Table 6, 7). This is confirmed by the N₂ fixation of Mix2 of winter peas, which was found to be equally high as or higher than in pure stands in spite of the reduced seeding rates. The debris from above-ground biomass was certainly much more marked than the 2 kg ha^–1 mentioned by Kaul (2004) for spring peas, because of the winter pea’s great haulm length of up to 3 m and the relatively low number of pods at mature harvest (Urbatzka, 2010). Presumably, some of this N was mineralized until mature harvest due to low C/N-ratio, as it was shown for N-rhizodeposition of pea (Wichern et al., 2007). As a consequence, this fraction was only partially taken into account including inorganic N, when esti­mating N₂ fixation by the extended difference method. Hence N₂ fixation of winter peas in pure stands was most probably underestimated. In terms of the rate of its transformation and the amount of N in the debris of above-ground biomass, further research is certainly required.

In contrast to winter peas, a lower N₂ fixation was observed in spring crops mixture compared with pure stands, which was caused by the lower crop density in accordance with other studies (Jensen, 1996; Wichmann, 2003; Corre-Hellou et al., 2006). The calculated values in our study were comparatively low, since in other studies N₂ fixation of spring peas in pure stands ranged between 40 and 240 kg ha^–1 (e.g. Jensen, 1996; Wichmann, 2003; Corre-Hellou et al., 2006).

The higher N₂ fixation of winter peas in mixture as compared with spring peas was a consequence of earlier N₂ fixation and an indeterminate growth pattern resulting in higher above-ground biomass and probably higher below-ground biomass (Iwama and Yamaguchi, 1996). In pure stands, however, no consistent pattern could be observed, even though, based on the results on N₂ fixation at flowering harvest and at mature harvest (mixture), a higher N₂ fixation of winter peas would be expected. This, again, can probably be traced back to the loss of N (above all debris of above-ground biomass) not accounted for in our experiments.

The great variation of symbiotic N₂ fixation was a result of legume and cereal establishment and growth affected by the prevalent environmental conditions. For example the relative low N₂ fixation for spring pea pure stand at DFH in 2003/04 (Table 7a) was probably a consequence of severe weed infestation in the experimental plots (Urbatzka et al., 2011), which suppressed pea development. Also the lower N₂ fixation of winter peas in mixture at DFH in 2005/06 and 2006/07 (Table 7c, d) in comparison to the two earlier growing seasons can be explained by a relatively high portion of cereal grain yield (Urbatzka et al., 2011), which again was a result of the relatively high N availability resulting in a suppression of peas (Jensen, 1996). This was, analogous to harvest at flowering, also reflected by the relatively low portion of N derived from the atmosphere (NdfA) (Table 8) which was also observed by Voisin et al. (2002).

Results show that the earlier crop development of winter peas advances nodulation, and the generally higher above ground biomass and presumably root biomass leads to usually higher N₂ fixation of regular leaf winter peas at flowering and mature harvest (BBCH 61 and 89) as compared with spring peas. Moreover N₂ fixation mostly responds positively to increasing pea crop density. Another striking finding from the work was a similar N₂ fixation at mature harvest when grown in mixture compared with pure stands that could only be established for winter peas. The effect can most certainly be attributed to the better growth conditions in mixture, resulting in higher pea grain yields and N losses (debris) of pure stands that could at least partially be taken into account. Furthermore, N derived from atmospheric fixation of pure stands does not seem to respond to genotype, yet depends on the time of assessment, and is higher the lower soil fertility and/or N availability are.

N yield of aboveground biomass is usually higher for regular leaf winter peas than for spring pea because of earlier N uptake and an indeterminate growth pattern. Data on N uptake at mature harvest clearly show that the higher environmental stress resistance of mixtures is a major argument for cultivation of pea-cereal mixtures. In contrast this is not valid for spring peas probably due to their lower competitiveness concerning crop development and establishment in pure stands and mixture.

Overall, it can be concluded that cultivation of winter peas may be a valuable alternative to spring peas because of their higher level of N₂ fixation. Additionally, when grown in mixture, N availability after mature harvest may be rather low reducing the risk of N leaching.