NPK in aerial biomass (BA) of A. pintoi and weeds

To estimate the production of the BA of weeds and A. pintoi sampling was conducted in August 2009 using a wooden frame 0.50 x 0.50 m (0.25 m²). Five repetitions per plot were performed in zig-zag, harvesting at ground level all the herbaceous vegetation (A. pintoi and weeds) within the frame.

In treatments with A. pintoi weed species were separated from this legume. BA was weighed fresh and placed in paper bags to dry in an oven for 72 h or till constant weight at 70 °C, weighed and pulverized. With dry matter (MS) data, calculated the BA average of weeds and A. pintoi; then these variables were extrapolated to kg MS ha^-1 for each treatment.

MS of weed and A. pintoi were ground to analyze the N, P and K content in each treatment. N was made by the Kjeldahl method, while P and K were solubilized by acid digestion (HNO₃-HClO₄ at 1:2) and quantified with ultraviolet-visible spectrophotometry and atomic absorption spectrophotometry techniques, respectively. The absorption of N, P and K by plants was calculated by multiplying the dry weight per plot with the percentage of NPK content in plants present on each treatment.

Experimental design and data analysis

Data reported are the average values of three replicates with standard error of the mean. The mean values of N, P and K absorbed were analyzed using multifactorial analysis of variance (ANOVA) with a randomized complete block design, with factorial arrangement 2 x 2 and test for significant interactions (p≤ 0.05) between main effects (type of coverage [main independent variable] and type of texture [independent variable conditional]) and between levels of main effects (^{Steel and Torrie, 1980}); formed by the factors: soil texture, loam and silty clay, and the levels were produced dry matter in treatment with just weeds and produced dry matter in treatment with A. pintoi. Mean comparison between treatments were assessed using Tukey test, calculated at p˂ 0.05. The amount of photosynthetically active radiation at the herbaceous layer was measured at the beginning of the experiment (μmolS^-1 m^-2) for each treatment and was considered as a covariate. The statistical procedures were performed using the Statistical Package STATGRAPHIC^® Centurion XV (StatPoint, Inc, 2005).

Soil properties

It was found that at 11 months of established treatments, soil condition was statistically equal in organic matter content (average), P Olsen and K (high). Total nitrogen (Nt) content was low in loam soil and high in silty-clay, showing no statistical differences between treatments. Inorganic nitrogen (Ni) content was average, with no statistical differences between treatments with silty-clay; and with loamy the treatments showed statistical differences, with bare soil was low, soil with weeds was average and high in soil with A. pintoi (Table 2). The amounts of NPK in the soil where treatments were established were ideal for the development of A. pintoi as soil cover; these allow to reflect the amounts of nutrients that can absorb weeds associated with plantain in southern Mexico.

Table 2 Nutrients content in soil at 11 months after the establishment of the treatments. 

TF= textura franca; TA= textura arcillo-limosa.

In turn another factor of importance to consider for competition, such as differences in emergence time and developmental stage of the species and differences in plant size within a community may reduce the effects and differences between effects (^{Andreasen et al., 2006}). However the 11 months of establishment of the treatments were not enough to record changes in soil characteristics regarding organic matter, CO, Nt, Ni, P and K. Despite being on soil conditions suitable for its establishment, A pintoi had a slow development (nine months), with 70% ground cover. A. pintoi grows well in poor, acid soils with high aluminum saturation, but grows better in soil of medium fertility, sandy loam with good organic matter content (^{Grof, 1985}; ^{Algiers and Villareal, 1998}; Rincón, 1999; Melendez and Vazquez, 2007).

Organic matter, P Olsen, K content showed no differences regarding the initial contents before establishing treatments; however an increase in Nt content passing from low to average in loamy treatment and average to high in silty clay was recorded. This increase can of Nt can be attributed to the decomposition of weed that developed in section without cover before establishing the treatments (^{Vargas, 1997}).

The results of the analysis from MO, CO, total N, inorganic N, P Olsen and K, were performed and interpreted according to the classes established in ^{NOM-021-RECNAT-2000} (^{SEMARNAT, 2002}).

N uptake in A. pintoi and weeds

There are no interactions or significant difference (p> 0.05) in the type of coverage x soil texture on the amount of N absorbed by A. pintoi and weeds (Figure 1). In Figure 1, can be appreciated an effect but due to the variability does not reach the level of significance (p= 0.07), showing that there is an identical response of weeds in amounts of N absorbed with loamy and silty-clay soil. Based on this lack of significant interaction, the main effects of the type of cover and texture are independent, so the main effects can be used separately to interpret the effects of both factors (^{Dowdy and Wearden, 1983}; ^{Kuehl, 2001}). Thus, the effect of the loamy soil produces an in increase N uptake in g/m² of A. pintoi of 42%, in weeds exhibit the same behavior, but in lower intensity (6%), going from silty-clay to loam texture.

Figure 1 Average values of amounts of N (g m^-2) absorbed in two types of covers (IC; Tukey) on plantain plantations soil. Different types of lines represent the two types of comparative texture. There are no intensity and direction changes of the response. The sample size for each observation is 15 for each type of cover. 

There are highly significant differences in the type of cover factor (p˂ 0.001) on N absorption (g m^-2). Weeds (0.42 g N m^-2 ± 0.13) showed 38% larger N uptake than A. pintoi (0.26 g N m^-2 ± 0.14).

Regarding the texture factor, significantly affected N absorption g m^-2 in cover (p< 0.05). Being 21.5% higher the amount of N absorbed in soil with loamy than silty clay soil.

P uptake in A. pintoi and weeds

Our results showed a great effect on the main independent variable (cover), independent variable conditioning (soil texture) and interaction associated to green plantain plantations (p˂ 0.05), indicating that the simple effects of a factor differ between the levels of another factor. Figure 2, shows the presence of the interaction between the two factors, here the response lines are not parallel. The difference in the magnitude of response represents the interaction between the factors, these do not act independently and the interpretations will be based on the contrasts of simple effects. The effects of the interaction of the two levels of texture and type of cover in the amount of P absorbed by each plant species are shown in Figure 2.

Figure 2 Average values of the amount of P (g m^-2) absorbed in two types of texture (IC; Tukey) on plantain plantations soil. Different types of lines represent the two types of compared vegetation. Do not present changes of intensity and direction of the response. The sample size for each observation is 15 for each type of texture. 

The texture factor also significantly affect P absorption (p< 0.05). P uptake in loamy soil is 26.8% higher than in silty clay soil texture. Treatment with weeds at two levels of texture resulted in higher amounts of P absorbed than A. pintoi in silty clay soil (± 1.49 g P m^-2); however, in loamy texture showed a decrease of 56% in the amount of P absorbed. The silty clay texture has a strong influence, a positive effect indicating that the response varies in the same direction as the factor. A. pintoi showed higher P uptake (0.68 g P m^-2) in silty clay soil texture, with an insignificant decrease of 7.5% in loamy soil, which is reflected in the statistically equal amount absorbed in both textures (Figure 2).

The amount of P absorbed by vegetation type showed highly significant differences (p< 0.001). The highest amount of P absorbed was made by weeds (1.2 g P m^-2 ± 0.56) (Figure 3).

Figure 3 Average values of P absorption (g m^-2) (IC; Tukey) at each vegetation level. Uneven letters indicate statistically significant differences with different N concentration in dry matter (p˂ 0.05). The sample size for each treatment is n= 30. 

K uptake in A. pintoi and weeds

The amounts of K absorbed are affected by the type of cover and interaction, producing significant differences in the amounts absorbed by the species here evaluated (p> 0.05), this interaction is represented in Figure 4 as a change of intensity and direction of the response variable, which is in function of the effect of the type of cover this is greater than the interaction.

Figure 4 Average values of the amount of K (g m^-2) absorbed in two types of texture (IC; Tukey) on green plantain plantations soil. The different types of lines represent the two types of comparative vegetation. No effect on the type of cover and interaction. The sample size for each observation is 15 for each type of texture. 

As the two slopes are equal (positive) the effect of cover causes an increase in the amount of K absorbed passing from A. pintoi to weeds, since under silty clay texture the effect is much greater than in loamy. A. pintoi increased 21% the amount of K absorbed in loamy soil, in turn, weeds showed a decrease of 68% in this soil. However, the amount of K absorbed by weeds was higher than A. pintoi in silty clay soil texture with 75% and loamy soil with 53%.

The factorial analysis of variance indicates that there are highly significant differences in the type of vegetation on K (g m^-2) absorbed by A. pintoi and weeds associated with green plantain plantations (p˂ 0.001). The weeds had higher K absorption (0.5 g N m^-2 ± 0.25) than that of A. pintoi (0.17 g K m-2 ± 0.11) (Figure 5).

Figure 5 Average values of K absorption (g m^-2) (IC; Tukey) at each level of vegetation. Uneven letters indicate statistically significant differences with different concentration of N in dry matter (p˂ 0.05). The sample size for each treatment is n= 30. 

NPK uptake by type of cover

A. pintoi absorbed higher amount of P (6.65 g m^-2 ± 3.3) and K (5.08 g m^-2 ± 2.5) than N (2.6 g m^-2 ± 1.4) in their aerial biomass. This behavior differs from that reported by ^{Vargas (1997)}, who found greater accumulation of N and K than P, at nine months after establishing the association of this legume with bananas in Costa Rica, however, A. pintoi had six years established in the area evaluated. The large amount of P absorbed by the legume in this experiment can be explained by the importance of this element in the formation of nodules (^{Ulrich, 1997}) during the establishment of the legume; as cover crop in citrus A. pintoi at 10 months of establishment accumulates 54% more N (80.89 kg N ha^-1) than weeds (37.60 kg N ha^-1) (^{Dalcolmo et al., 1997}).

The biomass produced by legume species returns and incorporates nutrients to the soil through fallen leaves, varying between species (^{Teasdale et al., 2007}), as recorded by ^{Zwart et al. (2005)} with an accumulation of K (3.2%) and N (2.9%) and P (0.3%). The weeds have reduced release rate of N accumulated in aerial biomass (in 234 days), which quantifies as immobilization in their tissues for several months of this element; thus their availability for the crop is reduced, but not like this for A. pintoi presenting larger constants of decomposition, with a release rate of N (44 days), P (32 days), K (8 days) with an opposite behavior on weeds (^{Thomas, 1995}; ^{Espindola et al., 2006a}). Although weeds do not increase a lot their productivity, their N absorption in excess reduces the amount of N available to the crop and thereby reduces crop competitive ability (^{Andreasen et al., 2006}).

The group of weed species evaluated in this experiment absorbed larger amount of P (12.17 g m^-2 ± 5.6) followed by N (4.2 g m^-2 ± 1.3) and the least amount of K (1.69 g m^-2 ± 1.1). However it is important to note that some weed species absorb more N or P than cultivated plants, although they produce less biomass, besides, that some species absorb more N and P when grown alone (^{Andreasen et al., 2006}). These larger amounts of NPK by weeds compared to A. pintoi, can be given in a certain extent because perennial weeds are often best competitors, and are more difficult to control with cover crops than perennial weeds, due to greater nutritional reserves and the fastest establishment rates (multiplication by seeds and rhizomes) (^{Teasdale et al., 2007}).

Accumulations of NPK in aboveground biomass of A. pintoi for the agro-ecological conditions into which treatments were established differ from results obtained by other authors, with fertilization practices and inoculation of symbiotic microorganisms, ^{Perin et al. (1998)}, 29.99 kg N ha^-1, 1.36 kg P ha^-1, 16.73 kg N ha^-1 after five months of its establishment (mde); ^{Perin et al. (2000)} 109.1 kg N ha^-1, 6.67 kg P ha^-1, 51.08 kg K ha^-1 at 8 mde; ^{Perin et al. (2003)} 572 kg N ha^-1, 37 kg P ha^-1 and 247 kg K ha^-1 at 24 mde; ^{Puertas et al. (2008)}, 130.58 kg N ha^-1, 7.61 kg P ha^-1, 83.20 kg K ha^-1, at 12 months.

NPK uptake by type of texture

The type of soil texture plays an important role on the amounts of nutrients that A. pintoi and weeds may absorb. ^{Baruch and Fisher (1996)} when comparing the establishment of A. pintoi in two types of soil texture concluded that this legume that grows best in sandy clay soils can play an important role in the availability of water and nutrients. N concentrations (in clay loam texture: 1.92 kg N ha^-1, loamy: 3.37 kg N ha^-1) found in A. pintoi associated as cover in green plantain at 11 months after its establishment may not reflect the fixation the comes from bacteria-legume symbiotic association, because as it might existed a good nodulation due to adequate P concentrations in the study site. In addition, it is likely that there is a negative influence by the occurrence of a drought period (four months) that started four months after sowing A. pintoi. ^{Valentim et al. (2003)} mentions the negative influence that some factors may have, limiting growth of this legume, such as diseases or pests, drought, overgrazing, reduced leaf area by the cuttings, which is related to decreased N fixation of the legume in the coming weeks.

Our findings suggest that weeds reflect the amounts of NPK that can absorb and immobilize during their life cycle, also competing with green plantain. The amounts absorbed by weeds are: in silty clay texture 4.1 kg N ha^-1, 14.94 kg P ha^-1 and 6.06 kg K ha^-1; loamy texture: 4.31 kg N ha^-1, 9.39 kg P ha ^-1 and 4.09 kg K ha^-1, finding influence from the type of soil texture, as mentioned by ^{Andreasen et al. (2006)} reports that the variation of nutrition available in the field for plant is influenced by variation in soil properties. ^{Qasem (1992)} analyzed nutrients absorption by weeds, beans and tomatoes finding that the concentrations of N, P, K and Mg were higher in weed species than in cultivated species. For example, the high population of weeds can increase the pressure of weeds in subsequent years, as are annual, spreading their seeds and perennial weeds have a good chance to increase their root or rhizome biomass (^{Kristoffersen et al., 2008}) with a strong competition for nutrients with the crop.