1. Introduction
In dry Mediterranean areas as well as in many arid and semi-arid areas in the world, irrigation is the only land management practice that ensures biomass production. Frequently, this is only possible after a careful irrigation strategy and design, since lack of water or low quality soil or water in these areas prevents sustainable and efficient use of water.
A higher irrigation efficiency has been promoted in recent years in the Ebro Valley (NE Iberian Peninsula) by government policies and water district authorities. The Ebro Valley is an area with more than 5 x 105 ha being irrigated with water of variable quality from the Ebro River and its tributaries as well as subsurface water coming from reservoirs in the Pyrenees. Many of the soils are saline or sodic and have low organic matter (OM) contents. In addition, the predictions of climate change in the area indicate marked increases in evapotranspiration that imply greater irrigation needs. Fruit production is expanding in the Mediterranean Basin, sometimes at the expense of marginal land transformation. As a result of less available water and greater irrigation needs, irrigation management practices need to be optimised such as by using subsurface drip irrigation.
Mulching techniques have arisen to address these limitations. Kumar et al. (1985) and more recent studies (e.g. Kar and Kumar, 2007; Diana et al., 2008; Mubarak et al., 2009; Nwite et al., 2011) reported positive effects of organic by-products amendments on soil physical properties (e.g., soil water retention increase and soil evaporation decrease).
With the objective of improving irrigation efficiency in fruit orchards in the Ebro valley, Lordán et al. (2013) carried out an experiment using different methods of applying irrigation water, including, amongst others, deficit irrigation, the application of pressurized air and a rice husk amendment to subsurface irrigation. The results of tree growth after 2 years indicated a substantial effect of the rice husk amendment on soil physical properties (infiltration and hydraulic conductivity), which resulted in a higher canopy growth and trunk diameter. It seemed thus that soil amendment with rice husk substrate improved soil structure. (Lordán et al., 2013).
A goal of this study is to understand the mechanisms that resulted in the improvements measured by this experiment, by a detailed analysis of the porosity around the pipes using micromorphological and micromorphometrical approaches.
2. Materials and methods
2.1. Experimental design and treatments
The experiments was carried out in a commercial peach orchard (Prunus persica L.) located in Villanueva de Sigena (Huesca, Spain) (Figure 1).
The soil is a Xeric Torriorthent (Soil Survey Staff, 1999), developed from marls and with a silty-loam texture (USDA; 40% sand, 40% silt and 20% clay). The soil is saline (ECe = 6.0 dSm-1 at 25 ºC), has a pH of 8.7 and an organic matter content of 10.0 g kg−1. The trees were planted in 2010 at 5 × 2.5 m, in a ridge planting system (1.5 m bottom and 0.8 m high), and irrigated through a subsurface irrigation system. The dripperline was buried at a depth of 0.25-0.3 m below the surface and was placed at a distance of 0.15-0.2 m from each tree trunk. Four different treatments of subsurface irrigation were tested:
(i). Full (100%) ridge subsurface drip irrigation (RSDI 100).
(ii). Deficit ridge subsurface drip irrigation consisting of a 30% reduction (70% irrigation) in crop evapotranspiration through the season (RSDI 70).
(iii). Soil was mixed with rice husks and gypsum at respective proportions of 12:12:1 and placed around the irrigation pipes, prior to planting, at an application rate of 5.0 Mg ha-1 of rice husk (MR SDI 70). Irrigation was the same as described in item (ii).
(iv). Pressurized air injection and deficit subsurface irrigation as in (ii) (RSDI 70 air).
A randomized complete block design with three replications was set up. Air injection was performed with a Venturi air injector (Mazzei Air Injector A3, Toro Ag, USA). The dripper line was buried between 0.25 m and 0.3 m below the surface and was at a distance of between 0.15 m and 0.2 m from each tree trunk. The complete soil and experiment design can be found in Lordán et al. (2013).
2.2. Analysis of soil properties
In order to evaluate changes in soil physical properties, one undisturbed block (0.3 m each side) was taken from each treatment, and two vertical thin sections, 5 × 13 cm, including the subsurface pipe, were made following the methods of Benyarku and Stoops (2005). A fluorescent dye (Uvitex®) was added to the polyester resin to obtain images of the porosity visible in UV light.
The description and study of the thin sections was done according to the guidelines of Stoops (2003). The thin sections were scanned in transmitted light. For the quantification of structure and porosity, three fields from each section, 13.23 cm2each (3.15 cm x 4.20 cm) were photographed in a light box to obtain transmitted-light images through the thin sections. Three different images were obtained for each of the 12 fields: (i) transmitted polarized light, (ii) transmitted light under crossed polarizers (using 2 perpendicularly oriented polarizing sheets, above and below the thin section), and (iii) incident UV light. The latter image was binarized using ImageJ® (Rasband, 2012). The other two images were used as visual controls of the binarization quality of the porosity.
In all cases, the background noise (features below 1 pixel) was removed, which means that only the porosity larger than 15 µm was visualized.
The pore parameters were analysed using the ImageJ® software. The field parameters (a single value for each image) were total pore area and total pore perimeter, and were obtained from pore images including pore edges. The morphological parameters of each pore (excluding those pores cut by frame edges) were circularity (Equation 1) and solidity (Equation 2). Equation 1 A circularity value of 1.0 indicates a perfect circle, while values close to 0.0 indicate elongated or irregular pores. Equation 2
A circularity value of 1.0 indicates a perfect circle, while values close to 0.0 indicate elongated or irregular pores.
The convex area is the area included within the curve surrounding the pore (without concave segments). High values correspond to smooth pores, and low values to rough contours.
2.3. Data analysis
The pore size distribution was calculated using Quantim4 (Vogel, 2008). The Quantim4 program applied an opening algorithm to the pore space, overlaying circles of increasing diameters and discarding any pores that did not fill them. The images obtained were associated with different apparent pore diameter intervals. The percentage of (decreasing) porosity of each image was quantified. These apparent pore diameters are more meaningful for measuring soil porosity than equivalent pore diameters (diameter of a circle with the same area as the pore), as they can be assessed in terms of root penetration or water retention at a given matrix potential. Porosity ranges were set up according to different apparent pore diameter intervals: 15-30 µm, 30-60 µm, 60-100 µm, 100-200 µm, 200-400 µm and over 400 µm. Total porosity values accounted for all pores with equivalent diameter >15 µm.
A multivariate analysis of variance (MANOVA) model was used for analysis of soil average pore size, pore area fraction, average pore perimeter and specific perimeter, assuming colinearity between these variables. Later, links between significant variables and treatments were studied by multivariate discriminant analysis. The effects of treatments on tree growth and yield were determined using a linear mixed model (Type III sum of squares). The Tukey Honestly Significant Difference (THSD) post hoctest was used for comparing treatments and crop response data. All data analysis was performed using JMP-SAS software (JMP, Version 12 Pro. SAS Institute Inc., Cary, NC, 1989-2014).
3. Results and discussion
3.1.Micromorphology
The groundmass of all thin sections is composed of a coarse fraction made of fine and medium sand of calcite and quartz, gravels of micritic limestone, and a micromass of clay, fine silt and micrite. The coarse/fine (c/f) ratio is approximately 1/2 (c/f limit 20 µm) and the b-fabric is crystallitic micritic. Organic components are mainly root sections, and rice husks in the MR SDI 70 treatment.
3.2. Structure and porosity
Noticeable changes in the pore types of the four treatments are evident in the scanned thin sections (Figure 2). The soils of the full and restricted irrigation treatments (RSDI 70 and 100) are characterised by a weakly developed, subangular blocky structure, with intra-aggregate vesicular pores, which are indicative of a poor structure. Some randomly oriented fragments of groundmass with an internal laminar structure are present. These fragments are probably originated during the disturbance and packing of the soil materials when burying the pipe.
The restricted irrigation with pressurized air treatment (RSDI 70 air) produced, in its upper part, a soil structure similar to that of RSDI 70 and 100. In the surroundings and below the pipe, where the effect of pressurized air was higher, the structure is apedal, with vesicles (bubbles) as main pores, with diameters of 100-400 µm (Figures 3).
The soil of restricted irrigation with rice husk treatment (MR SDI 70) has the highest porosity. Its micromorphology is characterized by the presence of many oval organic excrements, with sizes from 50 to 100 µm, associated with mildy decomposed rice husk tissues (Figure 4), in a very open fabric. The shape, size, organic nature and distribution of the excrements indicate that they are most probably produced by Oribatid mites (Zaiets and Poch, 2016, 2018). Mixed with the excrements it is possible to observe a mesh of siliceous threads that resemble silica skeletons of a freshwater sponge (Figures 4.3, 4.4, 4.5, Chad L. Yost, pers. comm.).
3.3. Micromorphometry of the porosity
Table 1 shows the results of the image analyses of the porosity. The pore area fraction confirms the visual estimations of the total pore area of the treatments. RSDI 100, RSDI 70, and RSDI 70 air have no significant differences with each other. MR SDI 70 has twice the pore area fraction as the other three. The equivalent pore size (in this case the diameter of a circular pore having the same area) follows the same trend.
MR SDI 70 | RSDI 70 | RSDI 70 air | RSDI 100 | |
---|---|---|---|---|
Pore area fraction (%) | 42.39 a | 19.76 b | 25.32 b | 16.45 b |
Equivalent pore size (mm) | 0.28 a | 0.14 b | 0.14 b | 0.11 |
Average pore perimeter (mm) | 1.49 a | 1.24 ab | 1.31 ab | 1.05 a |
Specific perimeter (mm-1) | 2.33 a | 1.70 b | 2.46 a | 1.67 b |
Circularity (pore average)* | 0.764 | 0.768 | 0.757 | 0.775 |
Solidity (pore average)* | 0.832 b | 0.832 b | 0.849 a | 0.853 a |
Number of replications | 9 | 6 | 8 | 6 |
The specific perimeter indicates the length of the interphase solid-void in the section, and corresponds to the surface of the same interphase related to the soil volume. It is higher in the treatments with air and with rice husks, but due to different processes: vesicles due to bubbling and biological activity. The physical behaviour is therefore different, since vesicles are non-connected pores and faunal activity creates highly connected pores. This high irregularity of the pores in the rice husk treatment (MR SDI70) is reflected in the lower solidity compared with the other treatments. No differences in pore circularity are observed.
The relation between these variables can be better seen with a MANOVA of the first four variables (Figure 5). The model is highly significant (Wilks' Lambda test P < 0.0001*).
Figure 5 shows the centroid graph of the first two canonical components. They indicate that the porosity fraction is the main variable explaining differences between treatments, followed by the pore size.
The pore size distribution also differs depending on the treatment (Figure 6), with the MR SDI 70 (rice husk treatment) having higher porosities in all fractions, typically twice that of the other treatments (Table 2).
Diameter (μm) | MR SDI 70 | RSDI 70 | RSDI 70 air | RSDI 100 |
---|---|---|---|---|
0-15 | 0.14 a | 0.05 b | 0.06 b | 0.07 b |
15-30 | 0.67 a | 0.21 b | 0.24 b | 0.24 b |
30-60 | 3.20 a | 0.95 b | 1.10 b | 0.89 b |
60-100 | 7.52 a | 2.28 b | 3.07 b | 2.17 b |
100-200 | 15.73 a | 5.78 bc | 8.82 b | 5.17 c |
200-400 | 12.54 a | 6.41 b | 6.97 b | 4.74 b |
Number of replications | 6 | 6 | 7 | 6 |
The most different fraction is 30-60 µm (porosity fraction responsible for water retention), where the treatment with rice husk (MR SDI 70) triples the porosity compared to the other treatments. These porosity changes, due to faunal activity, explain the positive effects in the water retention capacity and also in the saturated hydraulic conductivity observed by Lordán et al. (2013), in comparison to pressurized air that only increases vesicular (non-connected) porosity.
Oribatid richness and diversity have been related to soil structure (Ducarme et al., 2004; Jorrín and González-Fernández, 2016) and to soil organic matter (Noti et al., 2003) among other factors, but without establishing any clear cause-effect relationship. Our micromorphological study shows that rice husk packing, together with the mite activity feeding on them, are responsible for the high packing porosity.
4. Conclusions
The improvement of physical properties of a soil amended with rice husk, translated to a higher canopy cover and tree growth in peach orchards with subsurface irrigation in the Ebro valley, is caused by the activity of mites (probably Oribatid) feeding on these organic residues. This treatment, even with restricted irrigation, modifies the soil structure by doubling the soil porosity and reducing pore smoothness compared to the full irrigation, restricted irrigation and pressurized air with restricted irrigation treatments. These porosity changes imply higher water retention and a higher efficiency conveying water to the roots. In all the treatments, the nature of the material (intentionally altered soil around the pipes) is responsible for the absence of pedofeatures, other than those with faunal origin. Our results stress the importance of promoting the biological activity of micro and mesofauna in all management practices tending to increase the irrigation efficiency in poorly structured soils.