Next Article in Journal
Genotypic Variation in Nutrient Uptake Requirements of Rice Using the QUEFTS Model
Next Article in Special Issue
Floristic Composition Mediates Change in Forage Nutritive Quality of Atlantic Mountain Grasslands after Experimental Grazing Exclusion
Previous Article in Journal
Natural Resistances to Viruses in Cucurbits
Previous Article in Special Issue
Defoliation Dynamics in Kikuyugrass Pastures Subjected to Intensities of Defoliation
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agronomy
Volume 11
Issue 1
10.3390/agronomy11010024
agronomy-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Spatial Variability of Soil Phosphorus Indices under Two Contrasting Grassland Fields in Eastern Canada

by
Jeff D. Nze Memiaghe
1,2,
Athyna N. Cambouris
1,*,
Noura Ziadi
1,
Antoine Karam
2 and
Isabelle Perron
1
1
Agriculture and Agri-Food Canada, Quebec Research and Development Centre, 2560 Hochelaga Boulevard, Québec City, QC G1V 2J3, Canada
2
Soils and Agri-Food Engineering Department, Université Laval, 2425 rue de l’Agriculture, Québec City, QC G1V 0A6, Canada
*
Author to whom correspondence should be addressed.
Agronomy 2021, 11(1), 24; https://doi.org/10.3390/agronomy11010024
Submission received: 28 November 2020 / Accepted: 18 December 2020 / Published: 24 December 2020
(This article belongs to the Special Issue Sustainable Management and Utilization of Permanent Grassland)
Browse Figures
Versions Notes

Abstract

:
Phosphorus (P) is an essential nutrient for grassland production systems. However, continuous applications of P fertilizers result in soil P accumulations, increasing the risk of P losses in runoff and erosion. This study aims to investigate the field-scale variability of soil-test P (STP) in two contrasting grassland fields using descriptive statistics and geostatistics for accurate recommendations on soil sampling strategy and sustainable approaches to P management. A young grassland (YG; 2 years) and an old grassland (OG; 10 years under permanent pasture) were classified as humo-ferric podzol and received organic fertilizers. Soil samples were collected in 16-m by 16-m triangular grids at two depths (0–5 and 5–20 cm). They were analyzed for available P and other soil elements extracted using the Mehlich-3 method (M3). The agri-environmental P saturation index (P/Al)M3 was calculated. Phosphorus accumulation was observed in OG (0–5 cm) as a result of long-term manure applications. Repeated applications of organic fertilizers can impact the long-term buildup of soil P, thus decreasing soil P va-riability and spatial dependence in permanent grasslands. A soil sampling strategy focusing on the 0–5 cm layer should be retained in permanent grasslands for sustainable P recommendations in Eastern Canada.

1. Introduction

Phosphorus (P) is an essential nutrient for most crops, including associated plants of grassland communities. When it is applied in excess of crop requirements, P can accumulate at the soil surface, leading to P runoff via soil erosion and to P contamination of the environment, including eutrophication of surface waters such as lakes [1,2,3]. To prevent this P contamination in the environment, the degree of P saturation of soils (%) was calculated from the ratio of phosphorus extracted with ammonium oxalate (Pox) on the maximum retention capacity of the P (CRP) [4]. The critical P saturation value of 25% has been defined as an international benchmark intended to prevent the risk of P loss and water contamination increases [5].
Grasslands are defined as terrestrial ecosystems dominated by herbaceous and shrub vegetation [6,7] and are distributed according to landscape and site characteristics (soil, climate, water availability) and farm structures. Permanent grasslands include grazed pastures and grasslands harvested for hay and silage, or a combination of the two, and are located primarily in Western Europe, including the United Kingdom and Ireland [8]. Canadian grassland fields covered an area of 1.96 million ha in 2016 [9]. In the Eastern Canadian context, grassland cultivation represents an important economic opportunity for local farmers to increase their revenue. For instance, 50% of agricultural land was under intensive permanent grassland cultivation in the province of Quebec, producing 6.3 million tons of hay and generating a total economic profit of $19 M in 2016 [10]. Permanent grasslands and annual crops with reduced tillage represent crop systems with less risk of runoff and soil erosion, with respect to continuous annual crops under traditional systems with conventional tillage [11]. These two common crop systems were traditionally cultivated by grassland farmers, owing to their high economic revenue.
Moreover, P agricultural inputs were provided mainly from organic fertilizers (manure and slurry), mineral fertilizers, and other waste fertilizers in Eastern Canada [11]. Owing to their low economic costs, organic fertilizers represent 63% of total P sources, representing two-thirds of the total P applied by local farmers on agricultural lands [12], including grassland fields. Over a long period, agricultural farms produced more than 31 million tons of organic manures, which were applied in agricultural fields [11] and represented more than 95,000 tons of P2O5 [13]. Thus, grassland fields were intensively fertilized using organic manures with high P concentrations [11], increasing the environmental risk. Furthermore, in Eastern Canada, the soil sampling strategy for P fertilizer application is currently at 0–17.5 cm-depth in agricultural fields [11]. In 2012, total P loads in Eastern Canadian rivers were estimated at approximately 3800 tons of total P per year [14]. Thus, agri-environmental management of soil P in grassland fields should be based on scientific knowledge and understanding of soil-available P to develop better P fertilization strategies.
Soil P availability is defined as the total amount of soil P likely to end up in the soil solution as orthophosphate ions during a growing crop cycle [15]. In North America, soil-available P may be assessed for acidic soils using M3 extraction as an official method in Eastern Canada [16]. From this method, two main P indices were developed, specifically (1) P-Mehlich-3 (PM3), an agronomic indicator for acidic mineral soils in Eastern Canada and USA, and (2) (P/Al)M3, an important agri-environmental P indicator, which estimates soil P saturation to allow accurate P fertilizer recommendations and which integrates agronomical aspects and environmental risks. In the province of Quebec, the critical environmental threshold for coarse soil corresponds to the value of 15% [11]. These critical saturation indices were incorporated into the government’s environmental regulations related to agricultural lands [11,13]. In 2002, the national Agricultural Operations Regulation (REA) was adopted by the province of Quebec, establishing that the threshold level for P saturation (P/Al)3 corresponds to 13.1% in agricultural farms, including grassland fields [13]. These critical saturation indices were incorporated into the province of Quebec’s government environmental regulations related to agricultural lands [11,13].
Furthermore, the concept of balanced fertilization underpins the sustainable use and management of soil P under intensive cropping systems. To achieve this goal, it is important to have a better understanding of the spatial variability of soil P in grassland fields, as this can help in improving the profitability and sustainability of agricultural businesses through the economic and rational use of P fertilizers, reducing P losses by runoff, and protecting the environment. Understanding spatial variability of soil P in grassland fields may be mainly based on two strategies: (1) variable rate applications (VRA) and (2) the use of management zones (MZs) in the context of precision agriculture. The VRA approach manages the variability in a factor based on a prescription map obtained with georeferenced data and geostatistical analysis, while the MZs approach delineates smaller homogeneous units in which uniform management can be applied using variable management among the different units [17]. Those two approaches can help to reduce the losses of P in the environment by applying the P at the right place.
Many studies have been carried out on the spatial variability of soil P on a regional or territorial scale [3,18,19,20] and at the watershed scale [21,22,23]. Most of these studies focused on planning and mapping soil-available P with a view to preventing and assessing the risk of P-related environmental pollution. However, there are no studies conducted investigating spatial variability of soil-available P in grassland fields under intensive production in Eastern Canada. Thus, this study is, to our knowledge, the first one on the spatial variability of soil P in grassland fields in Eastern Canada. The main goal of this study was to evaluate the spatial variability of soil-available P in grassland fields managed under intensive farming systems. More specifically, the objectives were to: (1) investigate field-scale variability of soil P indices and other selected soil chemical properties (soil pHwater, total carbon, AlM3, CaM3, FeM3), and (2) study the relationships between soil P indices (referred to as PM3 and (P/Al)M3) and selected soil chemical properties for two contrasting grasslands.

2. Materials and Methods

2.1. Study Site and Soil Sampling

The study was conducted in the Bras d’Henri watershed (Chaudière-Appalaches region) located 30 km from Quebec City, QC, Canada. Two commercial fields were selected to represent different grassland management options: a young grassland field (46°48′N; 71°22′W, 2.35 ha), referred to as the YG field, and an old grassland field (46°48′N; 71°21′W, 2.47 ha), referred to as the OG field (Figure 1). The YG field was managed for corn (Zea mays L.; 2006; 2010) and soybean (Glycine max; 2007; 2011) rotation with 2 years of grassland (i.e., 2004–2005, 2008–2009 and 2012–2013) for 10 years. The YG field was planted in early May 2012. The OG field was used as a permanent pasture for 10 years (2004–2013). The YG field was fertilized with cattle manure and was harvested three times per year (mid-June, mid-August and late September). The OG field was fertilized with organic manure (e.g., calf slurry, pig manure and cattle excreta) and was harvested twice per year (mid-June and mid-August). Crop management and grassland fertilization practices were based on local recommendations for the province of Quebec [11].
Both soils were classified as humo-ferric podzols [24] and attributed to the Beaurivage soil series [24]. Soil surface texture for the YG and OG fields varied from sandy loam to loamy sand. They were classified to the same G3 soil texture group (i.e., texture with clay ≤ 20%) utilized in the province of Quebec recommendations [11]. Both fields were moderately well-drained. The annual winter mean temperature is −9.4 °C, and the mean growing season temperature is 16.6 °C [25]. Mean annual precipitation is 1120 mm, of which 320 mm is snow [24].
The main grass species consisted of Ladino white clover (Trifolium repens) cultivated in association with timothy (Phleum pratense L.) in the YG field and tall fescue (Festuca arundinacea Schreb.) in the OG field. In 2004–2005, 2008–2009 and 2012–2013, cattle manure was applied and incorporated twice a year in the first 5 cm using a low-ramps spreading system in the YG field at the rate of 18,700 L ha−1, equivalent approximately to 20 kg P ha−1 year−1. Cattle manure was applied in spring and later after the first cut (mid-June).
In the OG field, calf and pig manures were applied at the rate of 30 Mg ha−1 and 20 Mg ha−1, respectively, giving approximately 67 kg P ha−1 year−1 over 10 years based on reference values of P published in CRAAQ [26]. At the beginning of May and after the first cut (mid-June), organic manures were surface applied in the OG field using a low-ramps spreading system.
An intensive soil sampling using a grid of 16 m by 16 m, including three cross-shaped designs with several inter-points at about 4 m apart, was performed in spring 2013, providing 151 and 149 georeferenced sampling points for the YG and OG fields, respectively (Figure 1). A composite soil sample of four cores was taken within a radius of 1 m around each sampling point at the two soil depths (0–5 cm and 5–20 cm) using a 0.05 m-diameter Dutch auger (Eijkelkamp company, Giesbeek, Netherlands). Thus, a total of 302 and 298 soil samples were collected from the YG and OG fields, respectively.
Soil samples were air-dried, ground and sieved through a 2 mm sieve for characterization. Soil pH (1:1 water) was measured, according to Hendershot et al. [27]. Soils were extracted with a soil solution ratio of 1:10 using Mehlich-3 solution [28], and the concentrations (mg kg−1) of PM3, iron (FeM3), calcium (CaM3) and aluminum (AlM3) in the extract were determined by inductively coupled plasma optical emission spectroscopy (ICP–OES; Model, 4300DV, PerkinElmer, Shelton, CT, USA). The soil P agri-environmental indicator (P/Al)M3 was then calculated. Total carbon (TC, g kg−1) content was measured using an Elementar Vario MAX CN analyzer (Elementar Analysensysteme GmbH, Hanau, Germany).

2.2. Statistical and Geostatistical Analyses

Descriptive statistics were carried out using SAS Software version 9.4 [29] to investigate soil pHwater, PM3, AlM3, (P/Al)M3, FeM3, CaM3 and TC and their spatial variability. The minimum, maximum, mean, coefficient of variation (CV) and standard deviation of the mean (SDM) were determined for each of the soil layers from each field. The CV of the different soil chemical properties was classified based on the approach of Nolin and Caillier [30] using five classes: (1) low (CV < 15%); (2) moderate (15% < CV < 35%); (3) high (35% < CV < 50%); (4) very high (50% < CV < 100%); and (5) extremely high (CV > 100%).
Pearson’s correlation coefficients (r) between soil P indices (PM3, (P/Al)M3) and other soil chemical properties (AlM3, CaM3, FeM3, TC, and pHwater) were also calculated for each soil layer using SAS. Data were analyzed using the CORR and univariate procedures. Multiple regression equations were generated for the soil P indices for each soil depth.
Geostatistical analyses were performed for all soil parameters. Data were achieved for each of the soil layers using the GS+ version 9 software [31] geostatistical computations and model validations. The spatial structure of the different soil chemical properties was evaluated via isotropic and anisotropic semivariograms. Experimental semivariograms, the main component of kriging, are an effective tool for assessing spatial variability [32]. Semivariogram parameters for each theoretical model (spherical, exponential and Gaussian) were generated. The corresponding nugget (C0), partial sill (C), sill (C0 + C), and range values of the best-fitting theoretical model were calculated. The partial sill ratio (C/(C0 + C)), expressed as the percent of total semivariance, was used to define the spatial dependency or structure of the soil variables for the two fields. Semivariograms with a partial sill ratio of < 25%, 25% to 40%, 40% to 60%, 60% to 75%, or > 75% were considered to have a low, low–moderate, moderate, strong–moderate, or strong spatial dependency, respectively [33]. After the suitable theoretical model for each data set and the corresponding semivariogram parameters were selected, spatial variability maps were generated using 1 m × 1 m block kriging interpolation. Kriging map reliability was evaluated using cross-validation analysis (R2CV) [34].

3. Results

3.1. Descriptive Statistics of Soil Phosphorus Indices and Other Chemical Properties

For the YG field, the mean values of soil P indices and other soil chemical properties were similar in both soil layers (0–5 cm and 5–20 cm), with the exception of CaM3 (Table 1). Overall, the CVs ranged from 5% to 81%, with the highest obtained for the (P/Al)M3, and were classified as follows: low for soil pHwater and AlM3, moderate for FeM3, high for CaM3 and TC, and very high for the two soil P indices.
For the OG field, soil P indices and other soil chemical properties were higher in the 0–5 cm layer than in the 5–20 cm layer, with the exception of AlM3, for which the mean concentrations were similar (Table 1). On average, the CVs ranged from 6% to 64% in both soil layers, with the highest value obtained for the (P/Al)M3, and were classified as follows: low for soil pHwater and AlM3, moderate for FeM3 and TC, and high for CaM3. The CVs of the soil P indices varied from high to very high in both soil layers. The intensity of variability of the soil P indices was higher in the YG compared to OG for both soil layers.

3.2. Geostatistical Parameters of Soil Phosphorus Indices and Other Chemical Properties

In the YG field, soil P indices and the other soil chemical properties (except for CaM3) were fitted with geostatistical spherical models, and pure nugget models (PNs) were used for CaM3 in both soil layers. Spatial ranges varied from 28 to 81 m. Spatial ranges of soil chemical properties extended beyond the 16 m × 16 m sampling grid, indicating that the sampling grid used is appropriate. Spatial dependence ratios of soil P indices and other soil chemical properties ranged from 58% to 100% for the two soil layers and were classified according to Whelan and McBratney [33] as follows: FeM3 ranged from moderate to strong–moderate, pHwater and soil P indices were classified as strong–moderate, and TC and AlM3 were classified as strong (Table 2).
In the OG field, soil P indices and the other soil chemical properties (except for AlM3 and CaM3) were fitted with geostatistical spherical models, and PNs were used for CaM3 and for AlM3 in the 0–5 cm soil layer (Table 2). Spatial ranges varied from 27 to 76 m and extended beyond the 16 m × 16 m sampling grid, indicating that the sampling grid used is appropriate. Spatial dependence ratios of soil P indices and the other soil chemical properties ranged from 26% to 69% in both soil layers. According to Whelan and McBratney [33], the spatial dependence of AlM3 was low–moderate in the 5–20 cm layer. Spatial dependence values were moderate for soil pHwater and CaM3, strong–moderate to moderate for FeM3, low–moderate to moderate for soil P indices, and strong–moderate for soil TC in both soil layers (Table 2). Spatial dependency of soil P was lower under OG compared to YG for both soil layers.
In the YG field, cross-validation coefficients (R2CV) ranged from 0.31 to 0.54 for soil chemical properties and from 0.31 to 0.40 for soil P indices. In the OG field, R2CV were generally lower. The R2CV values ranged from 0.20 to 0.36 for the soil P indices. High R2CV values (R2CV > 0.6, [35]) indicate a good fit for kriged map reliability. Consequently, the R2CV values obtained for the YG field may indicate relative mean fits for map reliabilities for most soil chemical properties, including soil P indices, while the fits were weaker for map reliabilities of soil chemical properties in the OG field.

3.3. Spatial Distribution of Soil Phosphorus Indices and Other Chemical Properties

The PM3 and (P/Al)M3 spatial distribution maps showed visual similarities for each soil layer studied (Figure 2a,b,g,h and Figure 3a,b,g,h). In the YG field, spatial distribution values of soil P indices, TC and FeM3 were relatively similar in both soil layers (Figure 2a,b,d–g,i,j). Spatial distribution values of AlM3 were lower in the 0–5 cm soil layer compared to the 5–20 cm soil layer (Figure 2c,i).
However, in the OG field, spatial distribution values of soil PM3, (P/Al)M3, TC, and FeM3 were higher in the 0–5 cm soil layer than in the 5–20 cm soil layer (Figure 3a,b,d,e,g,h,j,k). For instance, spatial distribution values of these soil chemical properties varied from 46 to 232 mg kg−1, 2.8% to 13%, from 20 to 55 g kg−1 and from 135 to 324 mg kg−1 against 15–139 mg kg−1, 1.1% to 8%, 20–40 g kg−1, 135–297 mg kg−1 in 0–5 cm and 5–20 cm OG soil layers, respectively. In the OG field, no visual similarities were noted between AlM3 0–5 cm and 5–20 cm or between CaM3 0–5 cm and 5–20 cm. This was not surprising, as the PN models were obtained with the semivariogram analysis.
In general, spatial distribution maps also revealed visual associations between soil P indices and other soil elements, indicating the influence of these different soil chemical properties on the soil-available P. For instance, in the YG field, the AlM3 and TC concentrations from the soil layers were inversely related to the soil P indices (Figure 2a–d,g–j).
In the OG field, the AlM3 concentrations in the OG soil layer (5–20 cm) were inversely related to the soil P indices in OG (5–20 cm) (Figure 3g–i). The high CaM3 concentrations correspond to soil P accumulation regions in the OG soil layer (5–20 cm), confirming PM3-CaM3 accumulation (Figure 3g,h,l). These results showed that AlM3 and TC content might influence soil P retention capacity in both YG soil layers, whereas AlM3, FeM3 and CaM3 may affect soil P in both OG soil layers.

3.4. Relationships Between Soil Phosphorus Indices and Other Chemical Properties

In the YG field, Pearson’s correlation coefficients (r) with the highest significance were observed between PM3 and AlM3, followed by TC in both soil layers (Table 3). For (P/Al)M3, the highest r values were obtained with TC in both soil layers. In the OG field, significant relationships were observed between PM3 and AlM3, CaM3 and soil pHwater in both soil layers. However, the highest r was observed between (P/Al)M3 and CaM3 in each soil layer (Table 3). Thus, soil PM3 in the YG field may be influenced significantly by Al oxides and hydroxides (soil Al), soil organic matter (TC) and soil solution (soil pHwater), whereas in the OG field, Al oxides/hydroxides (soil Al), calcium carbonate (soil Ca), soil solution (soil pHwater) and iron oxides/hydroxides (soil Fe) may have a significant influence on soil P accumulation.
Multiple regression equations were generated for the soil P indices in the YG and OG soil layers to investigate the influence of these soil chemical properties (Table 4). In the YG field, the regression coefficients (R2) were very significant, ranging from 0.789 to 0.893, which means that 79% to 89% of soil-available P variations may be explained by AlM3, TC and soil pHwater in both soil layers. Similarly, in the OG field, the R2 varied between 0.532 and 0.659, indicating that 53% to 66% of soil-available P accumulations in both soil layers may be affected significantly by AlM3, soil pHwater, CaM3 and FeM3. Consequently, permanent grassland may accumulate soil P while significantly affecting the effect of these soil chemical properties.

4. Discussion

4.1. Soil Phosphorus Stratification and Its Environmental Implications

In this study, soil P accumulation was observed in the 0–5 cm OG soil layer, mainly due to multiple applications of organic fertilizers (calf slurries, pig manure, and cattle excreta) over a long time period. Indeed, soil P accumulation in the 0–5 cm OG top layer may also be due to the elimination of soil disturbance (no tillage, compared to the conventional tillage) under permanent grasslands.
Many studies [3,19,36,37] have shown that soil P accumulation in the surface layer from intensively cultivated lands is linked to increased applications of P fertilizers, organic inputs, soil land-use conversion and management. For instance, Sun et al. [3] reported that significant increases in soil-available P (P Olsen) were attributed to intensive local applications of swine manure over a 15-year period (1982–1997) in the Rugao region (China). Other studies [19,38] found that soil P indices (P-Bray I, P-ammonium-acetate-EDTA (PAAE): referred to as the Swiss reference method) varied significantly in different land uses and crop systems, including permanent grassland fields.
This study revealed that repeated application of organic fertilizers raised soil P accumulation in the first five soil centimeters, increasing the potential agri-environmental risk for P pollution in permanent grasslands. The mean (P/Al)M3 indicator raised up to twice (7%) in the 0–5 cm OG soil layer, relative to the 0–5 cm YG soil layer (3%). These results have several environmental implications for the sustainable management of soil P in grassland fields. Indeed, the (P/Al)M3 mean values obtained from the OG site were close to the national critical threshold P saturation (15%) elaborated by the Province of Quebec for coarse-textured soils, meaning that (P/Al)M3 values higher than 15% increased up to water pollution or groundwater contamination in this Eastern Canadian region.
Moreover, these results also revealed potential agri-environmental impacts for the management of permanent grasslands in the Eastern Canadian region. According to the national Agricultural Operations Regulation (REA) adopted by the province of Quebec in 2002, these obtained results were close to the threshold level (13%) for P saturation (P/Al)M3 for agricultural lands, including grassland fields. Soil P stratification or buildup is an important agri-environmental concern because the excess surface P may be lost in runoff [39]. At the same time, however, lower concentrations at depth in the rooting zone may reduce crop yields [40].

4.2. Variability of Soil Phosphorus Indices and Other Soil Chemical Properties

The highest CVs were obtained for soil P indices and the lowest CVs for soil pH measurements. These results were similar to other studies [20,41,42,43,44]. A low CV for soil pHwater may be due to the logarithmic scale of pH measurement [45] or may be attributable to management practices, including regular soil liming activities in intensive agriculture [46]. West et al. [47] reported that a high CV (47%) for soil-available P could be explained by chemical forms of measured P that are highly labile.
Many studies [20,48,49,50] have likewise reported very high CVs for soil-available P (CV > 50%), particularly on Irish soils under intensive grassland farming. McCormick et al. [50] reported a CV of 57% (n = 334) for a 50-ha parcel of land encompassing 15 small grassland fields ranging in size from 0.94 ha to 5.65 ha in Northern Ireland. Fu et al. [49] obtained a CV of 68% (n = 537) for a 52-ha grassland field located in southeastern Ireland.
This study shows that a reduction in the CV of soil P was observed in permanent grasslands. For example, Jia et al. [18] reported a large decrease in the CV of soil P from 151% to 55% after 25 years of intensive farming practices in China. However, some authors [18,19,30,48] found that long-term fertilization practices or intensive land-use changes, including permanent grasslands, may influence CV values of soil P indices. Roger et al. [19] reported extremely high CV values for soil-available P (PAAE) at permanent grasslands sites (CV = 107%) compared to mountain pasture sites (CV = 87%) and cropland sites (CV = 70%). From their study, the CVs of soil P saturation [(Pox/(Al+ Fe)ox) × 100] ranged from 35% to 38%. In general, CV values are good indicators of the intensity of soil variability, but not of the nature of this variability (structured or randomized variability), nor do they provide an accurate model of soil variability [42].

4.3. Spatial Dependence of Soil Phosphorus Indices and Other Soil Chemical Properties

Spatial dependence (R = (C/C0 + C)) describes the spatial structure of a soil property [51]. Some authors [20,52,53] have indicated that soil chemical properties with strong spatial dependence (R > 75%) are mainly influenced by soil intrinsic factors (soil formation factors, parental material, soil types or series), those with low spatial dependence (R < 25%) are strongly influenced by extrinsic factors (agronomic/cultural or fertilization practices, intensive use of organic fertilizers or mineral amendments), and those with moderate spatial dependence (25% < R < 75%) result from the interaction between intrinsic and extrinsic soil factors, which codetermine the soil property analyzed. The PNs models are characterized by a spatial structure smaller than the sampling grid because of an inappropriate sampling scale [44].
The spatial dependence of soil P indices in both YG and OG soil layers was moderate (31% to 71%) as a result of the interaction between soil intrinsic and extrinsic factors, such as intensive fertilization or farming practices. Many authors [18,20,33,54,55,56,57,58] have found moderate spatial dependence for soil P, while others [3,19,42,50] have reported strong spatial dependence for soil P (R > 75%) in their respective studies conducted on different scales. The strong spatial dependence of soil-available P observed in these studies may be attributable to the strong influence of intrinsic factors, such as different soil types or soil series, on the spatial distribution of soil-available P at different agricultural, local, or regional scales.
Spatial dependence values for soil pHwater and FeM3 in the YG and OG soil layers were moderate (41% to 74%), owing to the interaction between soil intrinsic and extrinsic factors. Arrouays et al. [44], Cambouris et al. [42], and Di Virgilio et al. [41] obtained similar results for soil pHwater. Spatial dependence values for soil AlM3 and TC ranged from strong to moderate in the YG and OG soil layers, respectively. Spatial dependence values obtained for AlM3 in YG soil layers were lower than those reported by Cambouris et al. [42]. For soil TC, Kokulan et al. [59] and Arrouays et al. [44] reported similar results for YG and OG soil layers (0–5 cm and 5–20 cm). Consequently, AlM3 and TC in both YG soil layers may result from the strong influence of soil intrinsic factors, whereas AlM3 and TC in both OG soil layers result from the interaction between soil intrinsic and extrinsic factors.
This study also revealed that a combination of intensive fertilization practices over a longer period reduced the spatial dependence of soil P in both OG soil layers. This may be explained by the increasing influence of soil extrinsic factors, such as P-rich fertilization, which modified the long-term spatial structure of soil P in the OG field. Other authors [3,18,48] reported a reduction of spatial dependence for soil P on Chinese and Irish soils at different regional and agricultural scales. These studies demonstrated that a weak spatial dependence of P distribution could be explained by various intensive agricultural farming practices (i.e., increasing fertilization, long-term P-rich mixed fertilization, and strong intensification of sheep grazing pasture activities).
From an agronomic perspective, the reduced spatial structure of soil P may affect the use of delineated MZs for sustainable management of soil P in grassland fields. A strong spatial structure can have a positive effect in terms of delineating MZs, while a weak spatial structure makes it difficult to use MZs, owing to the lack of spatial dependence.
The use of MZs is an agricultural approach based on the existence of within-field spatially structured soil and crop variability [42,45]. Subdividing a field into MZs is an effective way of understanding or controlling spatial variability of soil and crop properties in order to optimize soil nutrients and crop management [35]. Delineation of MZs has been tested successfully for various soil chemical properties [17,60] and crop fields [42,45], including old grassland fields [43], as well as different field sizes [35,45]. However, there are few examples of the use of this approach with the aim of optimizing the management of P variability in grassland fields.
The present study on soil P variability was carried out in 2.5-ha fields. Consequently, it would be mitigated to delineate MZs for these fields due to (1) the reduced spatial structure of soil P, which is caused by an increase in extrinsic soil factors, and (2) the size of the fields with respect to the crop studied. Further research on larger fields is needed to better define P spatial structure in grasslands.

4.4. Spatial Distribution of Soil Phosphorus Indices and Other Soil Chemical Properties

The R2CV values provide an indication of the reliability of spatial distribution maps. Weak R2CV values were obtained for soil P indices and other chemical properties in the OG field compared to the YG field. Use of R2CV values in grassland fields aims to evaluate the reliability of kriged spatial distribution maps of soil P indices for developing better agri-environmental strategies for sustainable P management in the context of precision agriculture. Thus, R2CV should also be considered as an important reliability factor for mapping soil P indicators in grassland fields to prevent P pollution. Some authors [58] reported lower R2CV values (0.0–0.20), while others [20,35] found higher R2CV values (0.59–0.62) for soil-available P. In this study, the lowest R2CV values obtained may be explained by the weak spatial structure of the soil P obtained, as suggested by Kravchenko [61].
Despite their low reliability (R2CV < 0.4), spatial distribution maps of soil P indices (Figure 3a,b) confirmed that the soil P accumulation observed in the OG soil layer (0–5 cm) is attributable to the repeated application of organic fertilizers. Furthermore, in this study, high P concentration areas were found in the YG (81–129 mg kg−1; 5% to 8%) and OG (139–232 mg kg−1, 8% to 13%) fields, probably due to uniform applications of P-rich organic fertilizers. The YG field was fertilized with cattle manure under corn and soybean rotation and included 2 years of grassland over a period of 10 years, whereas the OG field was fertilized for 10 years with multiple applications of organic fertilizers (e.g., calf slurries, pig manure, and cattle excreta).
Spatial distribution maps for soil AlM3 in both sites also confirmed the strong influence of high available Al content on soil P adsorption capacity in grassland soils in Eastern Canada. In a previous study, Tran et al. [62] highlighted this possibility in relation to Quebec agricultural soils. Taking into account soil AlM3 concentrations, these authors classified Quebec soils into three groups according to their capacity for soil P adsorption: (1) low capacity: [Al]M3 < 1100 mg kg−1; (2) average capacity: 1100 mg kg−1 < [Al]M3 < 1600 mg kg−1; and (3) high capacity [Al]M3 > 1600 mg kg−1. Furthermore, this study revealed that grassland fields in Eastern Canada have an average–high capacity for soil P adsorption. From these results, the spatial distribution of soil AlM3 could be evaluated as a potential indicator for soil P adsorption in grassland fields. This approach would aim to implement sustainable management of P as mineral or organic fertilizer, based on VRA of P fertilizer for preventing pollution. For instance, other studies [51] found that AlM3 could be considered as an important indicator of maximum P adsorption capacity for a VRA of P fertilizer in the agricultural fields from Eastern Canada.
The spatial distribution maps for soil P indices and other soil elements (AlM3, TC, FeM3 and CaM3) in YG and OG soil layers developed in this study confirm the observed correlation, particularly between soil P indices and soil CaM3 and FeM3 in the 5–20 cm OG soil layer. The southern area of the OG field (5–20 cm) may reflect soil P accumulation linked to Fe oxides or hydroxides and neoformation of Ca phosphates as a result of long-term fertilization (Figure 3k,l).

4.5. Relationships Between Soil Phosphorus Indices and Other Soil Chemical Properties

The results showed differences between YG and OG. Soil P may accumulate or increase in OG, and there may be significant changes of different soil parameters (AlM3, TC, CaM3, FeM3) on soil P indices, which are probably due to a decrease in soil pH.
Decreasing soil pH is common in intensively farmed soils due to heavy applications of inorganic N fertilizers [36,63,64]. It is well known that soil pH plays an important role in P solubility and adsorption processes, such as P adsorption onto soil oxides, minerals and soil organic matter [65,66]. For instance, Fu et al. [48] mentioned a significant relationship between soil-available P and soil pH (r = 0.69, n = 396, p < 0.01) in a grassland field used for grazing for 16 years. In this study, we found significant relationships between soil P indices and soil pH (r = 0.43 for PM3, p < 0.001; r = 0.46 for (P/Al)M3, p < 0.001) in the 0–5 cm OG soil layer.
Results also show that soil P accumulation in the OG soil layers (0–5 cm and 5–20 cm) may be related to soil-available concentrations of Al, Ca and Fe resulting from a decrease of soil pH. Many studies [3,36,67,68,69] reported that various agricultural and farming practices affect soil P accumulation. Other studies [70,71] observed relationships between soil P buildup and Fe/Al/-P, attributed to the intensive use of organic and inorganic P fertilizers in cultivated acid soils.
Positive relationships observed between soil PM3 buildup and soil CaM3 in the OG field may be indicative of soil P accumulation as Ca-bound P or neoformation of Ca phosphates after long-term P fertilization. This possibility has previously been advanced by several authors [36,72]. These earlier studies showed that the P-Ca relationship also depends on the relationship between soil pH and Ca. In the present study, we reported a significant relationship between soil pH and soil CaM3 in the 0–5 cm (r = 0.65; p < 0.001) and 5–20 cm (r = 0.57; p < 0.001) OG soil layers.

4.6. Agronomic and Environmental Implications

Repeated applications of organic fertilizers and other agronomical practices such as no-till soil may increase soil P accumulation in the first 5 cm of soil in agricultural fields [58,73,74], including grasslands [43]. This points to the importance of developing an effective soil sampling optimization strategy to provide better guidance for practical grassland management. Fertility management in grasslands is related to sampling depth. Although some studies [75,76,77] demonstrated that the 0–5 cm layer might provide the majority of N and P in grassland fields, a deeper soil profile may be important to sustaining healthy grasslands.
Soil sampling strategy in grassland fields represents a balancing act between fertility management and minimizing environmental impacts (i.e., P pollution, water quality). The current 0–17.5 cm sampling strategy may be more suitable for yield optimization, while the 0–5 cm sampling strategy would be better for more sustainable P recommendations in grassland fields to reduce P pollution in the environment. According to the results of this study, we recommend the use of a soil sampling strategy focusing on the 0–5 cm layer in permanent grassland fields, given the high risk of soil P stratification associated with P-rich fertilization, instead of the soil sampling strategy (0–17.5 cm) currently used in Eastern Canada, in order to derive more sustainable P recommendations for grasslands. Furthermore, other studies [1,50,76,78] recommended soil sampling strategies in grassland fields, focusing on the 0–5 cm, 0–7.5 cm or 0–10 cm layers in other countries.
In this study, reduction of soil P variability under permanent grasslands was observed, inversely to other previous studies achieved within grassland fields [8,47]. Therefore, sustainable management of soil P in grassland fields should aim to better control within-field soil P variability. Spatial distribution maps of soil P indices should be used to guide variable-rate P fertilizer applications in order to prevent under- and over-fertilization with P in grassland fields.
Furthermore, new technology advances (reflectance and evaluation of P) and the democratization of prices for VRA will make possible the potential use of this technique in the grasslands in the near future. Indeed, the implementation of local or environmental policies about agri-environmental and sustainable management of P will encourage grassland farmers to turn to these new technologies in the context of PA. Hence, although it is true that the additional economic costs can be high, the VRA approach would be an advantage or benefit for Eastern Canadian grassland farmers in reducing P losses in the environment.
Delineating MZs in grassland fields represents another option that can be used in the future to optimize the management of soil P variability in grassland fields. However, this agronomical practice has some limitations in adopting such an approach. These constraints are (1) a more reduced spatial structure of soil P, caused by a strong influence of extrinsic factors comparatively to the intrinsic factors, and (2) the field size with respect to the crop studied, which may influence the spatial distribution of MZs and their reliability. To overcome these constraints, MZs should be based on stronger soil P spatial structures obtained from reliable semivariograms, which are needed to calculate sampling intervals and develop an accurate site-specific application scheme for soil P in grasslands. In this context, field size is an important soil-related factor that needs to be taken into consideration. Further research on larger fields is needed to better define P spatial structure in grasslands.

5. Conclusions

This study represents, to our knowledge, the first one on the spatial variability of soil P in grassland fields in Eastern Canada. The main goal of this study was to evaluate the spatial variability of soil-available P in grassland fields managed under intensive farming systems, using descriptive statistics and geostatistical tools for accurate recommendations on soil sampling strategy, and developing potential future and sustainable approaches of P management in grassland fields.
In this study, repeated applications of organic fertilizers under grassland production with minimum soil disturbance increased soil-available P buildup [PM3, (P/Al)M3] at the 0–5 cm OG soil layer. This soil P accumulation increased the potential agri-environmental risk for P pollution in permanent grasslands. Under permanent grasslands, reducing variability and spatial dependence of soil P indices were observed in both soil layers (0–5 cm and 5–20 cm), as a result of the increasing influence of extrinsic soil factors, such as repeated applications of organic fertilizers over a long time. From an agronomic perspective, the reduced spatial structure of soil P may affect the use of delineated MZs for sustainable management of soil P in grassland fields. Despite their low reliability, spatial distribution maps of soil P indices should be used to guide variable-rate P fertilizer applications in order to prevent under- and over-fertilization with P in grassland fields.
In keeping with the results of this study, we recommend a soil sampling strategy focusing on the 0–5 cm layer in permanent grassland fields, given the high risk of soil P stratification caused by P-rich fertilization, as opposed to the soil sampling strategy (0–17.5 cm) currently used in Eastern Canada, in order to generate more sustainable P recommendations for grasslands. Moreover, this primary essay on the spatial variability of soil P was only restricted to two selected grassland fields. Further studies in other and larger grassland fields are required for a better understanding of the spatial variability of P to improve agri-environmental P fertilization for grassland fields.

Author Contributions

Conceptualization, J.D.N.M. and A.N.C.; methodology, J.D.N.M., I.P., N.Z. and A.N.C.; statistical analysis, J.D.N.M. and I.P.; writing—original draft preparation, J.D.N.M.; writing—review and editing, A.N.C., N.Z. and A.K.; supervision, A.N.C., N.Z. and A.K. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by Agriculture and Agri-food Canada (AAC), grant number J-002238.

Acknowledgments

We thank Claude Lévesque, Sylvie Michaud, and Marc Duchemin for their field assistance and technical support.

Conflicts of Interest

The authors declare no conflict of interests.

References

  1. Stroia, C.; Morel, C.; Jouany, C. Dynamics of diffusive soil phosphorus in two grassland experiments determined both in field and laboratory conditions. Agric. Ecosys. Environ. 2007, 119, 60–74. [Google Scholar] [CrossRef]
  2. Jordan, C.; McGuckin, S.; Smith, R. Increased predicted losses of phosphorus to surface waters from soils with high Olsen-P concentrations. Soil Use Manag. 2000, 16, 27–35. [Google Scholar] [CrossRef]
  3. Sun, W.-X.; Huang, B.; Qu, M.-K.; Tian, K.; Yao, L.-P.; Fu, M.-M.; Yin, L.-P. Effect of Farming Practices on the Variability of Phosphorus Status in Intensively Managed Soils. Pedosphere 2015, 25, 438–449. [Google Scholar] [CrossRef]
  4. Maguire, R.O.; Sims, J.T. Measuring agronomic and environmental soil phosphorus saturation and predicting phosphorus leaching with Mehlich 3. Soil Sci. Soc. Am. J. 2002, 66, 2033–2039. [Google Scholar] [CrossRef]
  5. Breeuwsma, A.; Silva, S. Phosphorus Fertilisation and Environmental Effects in The Netherlands and the Po Region (Italy); DLO The Winand Staring Centre: Wageningen, The Netherlands, 1992. [Google Scholar]
  6. Obermeier, W.; Lehnert, L.; Kammann, C.; Müller, C.; Grünhage, L.; Luterbacher, J.; Erbs, M.; Moser, G.; Seibert, R.; Yuan, N. Reduced CO2 fertilization effect in temperate C3 grasslands under more extreme weather conditions. Nat. Clim. Chang. 2017, 7, 137–141. [Google Scholar] [CrossRef]
  7. Zhou, Q.; Daryanto, S.; Xin, Z.; Liu, Z.; Liu, M.; Cui, X.; Wang, L. Soil phosphorus budget in global grasslands and implications for management. J. Arid Environ. 2017, 144, 224–235. [Google Scholar] [CrossRef] [Green Version]
  8. Higgins, S.; Schellberg, J.; Bailey, J. A review of Precision Agriculture as an aid to Nutrient Management in Intensive Grassland Areas in North West Europe. Adv. Anim. Biosci. 2017, 8, 782–786. [Google Scholar] [CrossRef]
  9. Statistics Canada. Farm and Farm Operator Data; Statistics Canada: Ottawa, ON, Canada, 2016. [Google Scholar]
  10. MAPAQ. Portrait-Diagnostic Sectoriel de L’industrie des Plantes Fourragères du Québec; Ministère de l’Agriculture de l’Alimentation et des Pêcheries du Québec: Québec, QC, Canada, 2018; p. 27. [Google Scholar]
  11. CRAAQ. Guide de Référence en Fertilisation, 2nd ed.; Centre de Référence en Agriculture et Agroalimentaire du Québec: Québec, QC, Canada, 2010; p. 473. [Google Scholar]
  12. Hébert, M.; Busset, G.; Groeneveld, E. Bilan 2007 de la Valorisation des Matières Résiduelles Fertilisantes; Ministère du Développement durable, de l’Environnement et des Parcs: Québec, QC, Canada, 2008; p. 14.
  13. MDDELCC. Guide de Référence du Règlement sur les Exploitations Agricoles; Ministère du Développement durable, de l’Environnement et de la Lutte contre les changements climatiques: Québec, QC, Canada, 2017; p. 185.
  14. Patoine, M.; Hébert, S.; Simoneau, M.; d’Auteuil-Potvin, F. Charges de Phosphore, D’azote et de Matières en Suspension à L’embouchure des Rivières du Québec, 2009 à 2012; Ministère du Développement durable, de l’Environnement et de la Lutte contre les changements climatiques, Direction générale du suivi de l’état de l’environnement: Québec, QC, Canada, 2017; p. 25.
  15. Frossard, E.; Julien, P.; Neyroud, J.-A.; Sinaj, S. Le Phosphore Dans les sols: État de la Situation en Suisse: Le Phosphore Dans les Sols, les Engrais, les Cultures et L’environnement; Office fédéral de l’environnement, des forêts et du paysage OFEFP: Berne, Switzerland, 2004; p. 180. [Google Scholar]
  16. Mehlich, A. Mehlich 3 soil test extractant: A modification of Mehlich 2 extractant. Commun. Soil Sci. Plant Anal. 1984, 15, 1409–1416. [Google Scholar] [CrossRef]
  17. Cambouris, A.N.; Zebarth, B.J.; Ziadi, N.; Perron, I. Precision agriculture in potato production. Potato Res. 2014, 57, 249–262. [Google Scholar] [CrossRef]
  18. Jia, S.; Zhou, D.; Xu, D. The temporal and spatial variability of soil properties in an agricultural system as affected by farming practices in the past 25 years. J. Food Agric. Environ. 2011, 9, 669–676. [Google Scholar]
  19. Roger, A.; Libohova, Z.; Rossier, N.; Joost, S.; Maltas, A.; Frossard, E.; Sinaj, S. Spatial variability of soil phosphorus in the Fribourg canton, Switzerland. Geoderma 2014, 217–218, 26–36. [Google Scholar] [CrossRef]
  20. Vasu, D.; Singh, S.; Sahu, N.; Tiwary, P.; Chandran, P.; Duraisami, V.; Ramamurthy, V.; Lalitha, M.; Kalaiselvi, B. Assessment of spatial variability of soil properties using geospatial techniques for farm level nutrient management. Soil Till. Res. 2017, 169, 25–34. [Google Scholar] [CrossRef]
  21. Nolan, S.; Little, J.; Casson, J.; Hecker, F.; Olson, B. Field-scale variation of soil phosphorus within small Alberta watersheds. J. Soil Water Conserv. 2007, 62, 414–422. [Google Scholar]
  22. Yuan, Y.; Locke, M.A.; Gaston, L.A. Tillage effects on soil properties and spatial variability in two Mississippi Delta watersheds. Soil Sci. 2009, 174, 385–394. [Google Scholar] [CrossRef] [Green Version]
  23. Wilson, H.F.; Satchithanantham, S.; Moulin, A.P.; Glenn, A.J. Soil phosphorus spatial variability due to landform, tillage, and input management: A case study of small watersheds in southwestern Manitoba. Geoderma 2016, 280, 14–21. [Google Scholar] [CrossRef]
  24. Lamontagne, L.; Martin, A.; Nolin, M.C. Étude Pédologique du Bassin Versant du Bras d’Henri (Québec); Laboratoires de pédologie et d’agriculture de précision, Centre de recherche et de développement sur les sols et les grandes cultures, Service national d’information sur les terres et les eaux, Direction générale de la recherche, Agriculture et Agroalimentaire Canada: Québec, QC, Canada, 2010; p. 188. [Google Scholar]
  25. Environnement Canada Canadian Climate Normals or Averages 1971–2000. Available online: http://www.climat.meteo.ec.gc.ca/climate_normals/index_e.html (accessed on 27 November 2016).
  26. CRAAQ. Guide de Référence en Fertilisation, 1st ed.; Centre de référence en agriculture et agroalimentaire du Québec: Québec, QC, Canada, 2003; p. 294. [Google Scholar]
  27. Hendershot, W.H.; Lalande, H.; Duquette, M. Soil reaction and exchangeable acidity. In Soil Sampling and Methods of Analysis; Carter, R., Gregorich, E.G., Eds.; Taylor & Francis: Boca Raton, FL, USA, 2008; pp. 173–178. [Google Scholar]
  28. Ziadi, N.; Tran, T. Mehlich 3-Extractable Elements. In Soil Sampling and Methods of Analysis; Carter, R., Gregorich, E.G., Eds.; Taylor & Francis: Boca Raton, FL, USA, 2008; pp. 81–88. [Google Scholar]
  29. SAS Institute. SAS User’s Guide. Statistics. Version 9.3.; SAS Institute: Cary, NC, USA, 2010. [Google Scholar]
  30. Nolin, M.; Caillier, M. La variabilité des sols. II-Quantification et amplitude. Agrosol 1992, 5, 21–32. [Google Scholar]
  31. Webster, R.; Oliver, M.A. Geostatistics for Environmental Scientists, 2nd ed.; John Wiley & Sons: Chichester, UK, 2007. [Google Scholar]
  32. Wu, C.; Wu, J.; Luo, Y.; Zhang, L.; DeGloria, S.D. Spatial prediction of soil organic matter content using cokriging with remotely sensed data. Soil Sci. Soc. Am. J. 2009, 73, 1202–1208. [Google Scholar] [CrossRef]
  33. Whelan, B.; McBratney, A. The “null hypothesis” of precision agriculture management. Precis. Agric. 2000, 2, 265–279. [Google Scholar] [CrossRef]
  34. Kravchenko, A.; Bollero, G.A.; Omonode, R.; Bullock, D. Quantitative mapping of soil drainage classes using topographical data and soil electrical conductivity. Soil Sci. Soc. Am. J. 2002, 66, 235–243. [Google Scholar] [CrossRef]
  35. Perron, I.; Cambouris, A.N.; Chokmani, K.; Vargas Gutierrez, M.F.; Zebarth, B.J.; Moreau, G.; Biswas, A.; Adamchuk, V. Delineating soil management zones using a proximal soil sensing system in two commercial potato fields in New Brunswick, Canada. Can. J. Soil. Sci. 2018, 98, 724–737. [Google Scholar] [CrossRef]
  36. Darilek, J.; Huang, B.; De-Cheng, L.; Zhi-Gang, W.; Yong-Cun, Z.; Wei-Xia, S.; Xue-Zheng, S.J.P. Effect of land use conversion from rice paddies to vegetable fields on soil phosphorus fractions. Pedosphere 2010, 20, 137–145. [Google Scholar] [CrossRef]
  37. Fu, B.; Chen, L.; Ma, K.; Zhou, H.; Wang, J. The relationships between land use and soil conditions in the hilly area of the loess plateau in northern Shaanxi, China. Catena 2000, 39, 69–78. [Google Scholar] [CrossRef]
  38. Bennett, E.M.; Carpenter, S.R.; Clayton, M.K. Soil phosphorus variability: Scale-dependence in an urbanizing agricultural landscape. Landsc. Ecol. 2005, 20, 389–400. [Google Scholar] [CrossRef]
  39. Sharpley, A.N.; Smith, S. Wheat tillage and water quality in the Southern Plains. Soil Till. Res. 1994, 30, 33–48. [Google Scholar] [CrossRef]
  40. Lupwayi, N.; Clayton, G.; O’Donovan, J.; Harker, K.; Turkington, T.; Soon, Y. Soil nutrient stratification and uptake by wheat after seven years of conventional and zero tillage in the Northern Grain belt of Canada. Can. J. Soil. Sci. 2006, 86, 767–778. [Google Scholar] [CrossRef]
  41. Di Virgilio, N.; Monti, A.; Venturi, G. Spatial variability of switchgrass (Panicum virgatum L.) yield as related to soil parameters in a small field. Field Crops Res. 2007, 101, 232–239. [Google Scholar] [CrossRef] [Green Version]
  42. Cambouris, A.; Nolin, M.; Zebarth, B.; Laverdière, M. Soil management zones delineated by electrical conductivity to characterize spatial and temporal variations in potato yield and in soil properties. Am. J. Potato Res. 2006, 83, 381–395. [Google Scholar] [CrossRef]
  43. Bélanger, G.; Cambouris, A.N.; Parent, G.; Mongrain, D.; Ziadi, N.; Perron, I. Biomass yield from an old grass field as affected by sources of nitrogen fertilization and management zones in northern areas. Can. J. Plant Sci. 2016, 97, 53–64. [Google Scholar] [CrossRef]
  44. Arrouays, D.; Vion, I.; Jolivet, C.; Guyon, D.; Couturier, A.; Wilbert, J. Variabilité intraparcellaire de quelques propriétés des sols sableux des Landes de Gascogne (France): Conséquences sur la stratégie d’echantillonnage agronomique. EGS 1997, 4, 5–16. [Google Scholar]
  45. Farooque, A.A.; Zaman, Q.U.; Schumann, A.W.; Madani, A.; Percival, D.C. Delineating management zones for site specific fertilization in wild blueberry fields. App. Eng. Agric. 2012, 28, 57–70. [Google Scholar] [CrossRef]
  46. Nolin, M.; Caillier, M.; Wang, C. Variabilité des sols et stratégie d’échantillonnage dans les études pédologiques détaillées de la plaine de Montréal. Can. J. Soil. Sci. 1991, 71, 439–451. [Google Scholar] [CrossRef] [Green Version]
  47. West, C.; Mallarino, A.; Wedin, W.; Marx, D. Spatial variability of soil chemical properties in grazed pastures. Soil Sci. Soc. Am. J. 1989, 53, 784–789. [Google Scholar] [CrossRef]
  48. Fu, W.; Zhao, K.; Jiang, P.; Ye, Z.; Tunney, H.; Zhang, C. Field-scale variability of soil test phosphorus and other nutrients in grasslands under long-term agricultural managements. Soil Res. 2013, 51, 503–512. [Google Scholar] [CrossRef]
  49. Fu, W.; Zhao, K.; Tunney, H.; Zhang, C. Using GIS and geostatistics to optimize soil phosphorus and magnesium sampling in temperate grassland. Soil Sci. 2013, 178, 240–247. [Google Scholar] [CrossRef]
  50. McCormick, S.; Jordan, C.; Bailey, J. Within and between-field spatial variation in soil phosphorus in permanent grassland. Precis. Agric. 2009, 10, 262–276. [Google Scholar] [CrossRef]
  51. Quenum, M.; Nolin, M.; Bernier, M. Cartographie numérique de la capacité maximale de sorption du phosphore des sols à l’échelle de la parcelle agricole à l’aide de variables auxiliaires. Can. J. Soil. Sci. 2012, 92, 733–750. [Google Scholar] [CrossRef]
  52. Cambardella, C.; Moorman, T.; Parkin, T.; Karlen, D.; Novak, J.; Turco, R.; Konopka, A. Field-scale variability of soil properties in central Iowa soils. Soil Sci. Soc. Am. J. 1994, 58, 1501–1511. [Google Scholar] [CrossRef]
  53. Goovaerts, P. Geostatistical tools for characterizing the spatial variability of microbiological and physico-chemical soil properties. Biol. Fertil. Soils 1998, 27, 315–334. [Google Scholar] [CrossRef] [Green Version]
  54. Chien, Y.-J.; Lee, D.-Y.; Guo, H.-Y.; Houng, K.-H. Geostatistical analysis of soil properties of mid-west Taiwan soils. Soil Sci. 1997, 162, 291–298. [Google Scholar] [CrossRef]
  55. Chang, J.; Clay, D.E.; Carlson, C.G.; Malo, D.; Clay, S.A.; Lee, J.; Ellsbury, M. Precision farming protocols: Part 1. Grid distance and soil nutrient impact on the reproducibility of spatial variability measurements. Precis. Agric. 1999, 1, 277–289. [Google Scholar] [CrossRef]
  56. Shi, Z.; Wang, K.; Bailey, J.; Jordan, C.; Higgins, A. Sampling strategies for mapping soil phosphorus and soil potassium distributions in cool temperate grassland. Precis. Agric. 2000, 2, 347–357. [Google Scholar] [CrossRef]
  57. Guo-Shun, L.; Xin-Zhong, W.; Zheng-Yang, Z.; Chun-Hua, Z. Spatial variability of soil properties in a tobacco field of central China. Soil Sci. 2008, 173, 659–667. [Google Scholar] [CrossRef]
  58. Cambouris, A.; Messiga, A.; Ziadi, N.; Perron, I.; Morel, C. Decimetric-Scale Two-Dimensional Distribution of Soil Phosphorus after 20 Years of Tillage Management and Maintenance Phosphorus Fertilization. Soil Sci. Soc. Am. J. 2017, 81, 1606–1614. [Google Scholar] [CrossRef] [Green Version]
  59. Kokulan, V.; Akinremi, O.; Moulin, A.P.; Kumaragamage, D. Importance of terrain attributes in relation to the spatial distribution of soil properties at the micro scale: A case study. Can. J. Soil. Sci. 2018, 98, 292–305. [Google Scholar] [CrossRef] [Green Version]
  60. Zebarth, B.; Drury, C.; Tremblay, N.; Cambouris, A. Opportunities for improved fertilizer nitrogen management in production of arable crops in eastern Canada: A review. Can. J. Soil. Sci. 2009, 89, 113–132. [Google Scholar] [CrossRef]
  61. Kravchenko, A. Influence of spatial structure on accuracy of interpolation methods. Soil Sci. Soc. Am. J. 2003, 67, 1564–1571. [Google Scholar] [CrossRef]
  62. Tran, T.; Giroux, M.; Guilbeault, J.; Audesse, P. Evaluation of Mehlich-III extractant to estimate the available P in Quebec soils. Commun. Soil Sci. Plant Anal. 1990, 21, 1–28. [Google Scholar] [CrossRef]
  63. Wang, H.-J.; Huang, B.; Shi, X.-Z.; Darilek, J.L.; Yu, D.-S.; Sun, W.-X.; Zhao, Y.-C.; Chang, Q.; Öborn, I. Major nutrient balances in small-scale vegetable farming systems in peri-urban areas in China. Nutr. Cycl. Agroecosyst. 2008, 81, 203–218. [Google Scholar] [CrossRef]
  64. Darilek, J.L.; Huang, B.; Wang, Z.; Qi, Y.; Zhao, Y.; Sun, W.; Gu, Z.; Shi, X.J.A. Changes in soil fertility parameters and the environmental effects in a rapidly developing region of China. Agric. Ecosys. Environ. 2009, 129, 286–292. [Google Scholar] [CrossRef]
  65. Havlin, J.L.; Beaton, J.D.; Tisdale, S.L.; Nelson, W. Soil Fertility and Fertilizers: An Introduction to Nutrient Management; Pearson Prentice Hall: Upper Saddle River, NJ, USA, 2005. [Google Scholar]
  66. Devau, N.; Le Cadre, E.; Hinsinger, P.; Gérard, F. A mechanistic model for understanding root-induced chemical changes controlling phosphorus availability. Ann. Bot. 2010, 105, 1183–1197. [Google Scholar] [CrossRef] [Green Version]
  67. Sharpley, A.; Smith, S. Fractionation of Inorganic and Organic Phosphorus in Virgin and Cultivated Soils. Soil Sci. Soc. Am. J. 1985, 49, 127–130. [Google Scholar] [CrossRef]
  68. Guo, F.; Yost, R.; Hue, N.; Evensen, C.; Silva, J. Changes in phosphorus fractions in soils under intensive plant growth. Soil Sci. Soc. Am. J. 2000, 64, 1681–1689. [Google Scholar] [CrossRef]
  69. Motavalli, P.; Miles, R. Soil phosphorus fractions after 111 years of animal manure and fertilizer applications. Biol. Fertil. Soils 2002, 36, 35–42. [Google Scholar]
  70. Kuo, S.; Huang, B.; Bembenek, R. Effects of long-term phosphorus fertilization and winter cover cropping on soil phosphorus transformations in less weathered soils. Biol. Fertil. Soils 2005, 41, 116–123. [Google Scholar] [CrossRef]
  71. Devau, N.; Hinsinger, P.; Le Cadre, E.; Gérard, F. Root-induced processes controlling phosphate availability in soils with contrasted P-fertilized treatments. Plant Soil 2011, 348, 203–218. [Google Scholar] [CrossRef]
  72. Eriksson, A.K.; Hesterberg, D.; Klysubun, W.; Gustafsson, J.P. Phosphorus dynamics in Swedish agricultural soils as influenced by fertilization and mineralogical properties: Insights gained from batch experiments and XANES spectroscopy. Sci. Total. Environ. 2016, 566, 1410–1419. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  73. Cade-Menun, B.J.; Carter, M.R.; James, D.C.; Liu, C.W. Phosphorus forms and chemistry in the soil profile under long-term conservation tillage: A phosphorus-31 nuclear magnetic resonance study. J. Environ. Qual. 2010, 39, 1647–1656. [Google Scholar] [CrossRef]
  74. Abdi, D.; Cade-Menun, B.J.; Ziadi, N.; Parent, L.E. Long-term impact of tillage practices and phosphorus fertilization on soil phosphorus forms as determined by 31 P nuclear magnetic resonance spectroscopy. J. Environ. Qual. 2014, 43, 1431–1441. [Google Scholar] [CrossRef]
  75. Haygarth, P.; Hepworth, L.; Jarvis, S. Forms of phosphorus transfer in hydrological pathways from soil under grazed grassland. Eur. J Soil Sci. 1998, 49, 65–72. [Google Scholar] [CrossRef]
  76. Watson, C.; Matthews, D. A 10-year study of phosphorus balances and the impact of grazed grassland on total P redistribution within the soil profile. Eur. J Soil Sci. 2008, 59, 1171–1176. [Google Scholar] [CrossRef]
  77. Watson, C.; Jordan, C.; Kilpatrick, D.; McCarney, B.; Stewart, R. Impact of grazed grassland management on total N accumulation in soil receiving different levels of N inputs. Soil Use Manag. 2007, 23, 121–128. [Google Scholar] [CrossRef]
  78. Stroia, C.; Morel, C.; Jouany, C. Nitrogen fertilization effects on grassland soil acidification: Consequences on diffusive phosphorus ions. Soil Sci. Soc. Am. J. 2011, 75, 112–120. [Google Scholar] [CrossRef]
Agronomy 11 00024 g001 550
Figure 1. Location and sampling strategies design of the studied grassland sites (Google Maps. Accessed 15 June 2020).
Figure 1. Location and sampling strategies design of the studied grassland sites (Google Maps. Accessed 15 June 2020).
Agronomy 11 00024 g001
Agronomy 11 00024 g002 550
Figure 2. Spatial distribution maps of PM3 (a,g), (P/Al)M3 (b,h), AlM3 (c,i), TC (d,j), FeM3 (e,k), and CaM3 (f,l) for the 0–5 cm and 5–20 cm soil layers, respectively, from the young grassland field in 2013.
Figure 2. Spatial distribution maps of PM3 (a,g), (P/Al)M3 (b,h), AlM3 (c,i), TC (d,j), FeM3 (e,k), and CaM3 (f,l) for the 0–5 cm and 5–20 cm soil layers, respectively, from the young grassland field in 2013.
Agronomy 11 00024 g002
Agronomy 11 00024 g003 550
Figure 3. Spatial distribution maps of PM3 (a,g), (P/Al)M3 (b,h), AlM3 (c,i), TC (d,j), FeM3 (e,k) and CaM3 (f,l) for the 0–5 cm and 5–20 cm soil layers, respectively, from the old grassland field in 2013.
Figure 3. Spatial distribution maps of PM3 (a,g), (P/Al)M3 (b,h), AlM3 (c,i), TC (d,j), FeM3 (e,k) and CaM3 (f,l) for the 0–5 cm and 5–20 cm soil layers, respectively, from the old grassland field in 2013.
Agronomy 11 00024 g003
Table
Table 1. Descriptive statistics of the soil chemical properties for two soil layers (0–5 cm and 5–20 cm) in the young and old grassland fields.
Table 1. Descriptive statistics of the soil chemical properties for two soil layers (0–5 cm and 5–20 cm) in the young and old grassland fields.
0–5 cm5–20 cm
UnitnMeanMinMaxSTDCV1 (%)nMeanMinMaxSTDCV(%)
Young grassland
Soil pHwater 1516.75.87.40.451516.85.97.50.46
Total carbong kg−11511452853815114528640
PM3mg kg−11515210161336315155102033665
AlM3mg kg−11511875115322521961015119711360238121511
FeM3mg kg−1151104601962423151111692262725
CaM3mg kg−115111803903156497421511532483339560539
(P/Al)M3 2(%)1513.00.5130.02791513.00.5150.0281
Old grassland
Soil pHwater 1495.54.76.40.46.51495.34.76.10.36
Total carbong kg−1149411161923149301047620
PM3mg kg−11491252932758461497582264560
AlM3mg kg−114918311494222515891491891154021861216
FeM3mg kg−11491871073553217149169943603621
CaM3mg kg−11491101325245741337149583155147226345
(P/Al)M3(%)1497.01.4210.04521494.00.4130.0364
1 CV: Coefficient of variation, 2 (P/Al)M3 = (PM3/AlM3) × 100. This soil property is a ratio, determined by soil phosphorus and aluminum concentrations, measured using Mehlich-3 solution.
Table
Table 2. Geostatistical parameters of the soil chemical properties for two soil layers (0–5 cm) and (5–20 cm) from the young and old grassland fields.
Table 2. Geostatistical parameters of the soil chemical properties for two soil layers (0–5 cm) and (5–20 cm) from the young and old grassland fields.
0–5 cm5–20 cm
Model 1Sill Ratio 2 (%)Range 3 (m)R2CV 4ModelSill Ratio (%)Range (m)R2CV
Young grassland
Soil pHwaterSph65810.54Sph74620.54
Total carbonSph100290.43Sph98280.34
PM3Sph63640.31Sph70670.40
AlM3Sph76400.32Sph86400.38
FeM3Sph58640.31Sph69390.39
CaM3PN---PN---
(P/Al)M3Sph69620.32Sph71660.35
Old grassland
Soil pHwaterSph41370.15Sph42390.22
Total carbonSph68310.07Sph62620.23
PM3Sph45510.36Sph51670.21
AlM3PN---Sph26320.10
FeM3Sph69270.32Sph43280.26
CaM3PN---Sph48200.09
(P/Al)M3Sph31470.32Sph52760.20
1 Semivariogram model: Sph: spherical, PN: pure nugget; 2 Sill ratio (%) = (C/(C0+C)) x100; this ratio measures spatial dependence or structure according to Whelan and McBratney (2000); 3 Distance at which a semivariance becomes constant; 4 Coefficient of determination of cross-validation.
Table
Table 3. Pearson’s correlation coefficients (r) of the soil-available phosphorus (P) indices (PM3, and (P/Al)M3 in relationship with soil chemical properties for two soil layers (0–5 cm) and (5–20 cm) from the young and old grassland fields.
Table 3. Pearson’s correlation coefficients (r) of the soil-available phosphorus (P) indices (PM3, and (P/Al)M3 in relationship with soil chemical properties for two soil layers (0–5 cm) and (5–20 cm) from the young and old grassland fields.
0–5 cm5–20 cm
PM3(P/Al)M3PM3(P/Al)M3
Young grassland
Soil pHwater0.38***0.37***0.48***0.44***
Total carbon−0.67***−0.64***−0.67***−0.63***
AlM3−0.86***na −0.87***na
FeM30.04ns0.10ns−0.01ns0.02ns
CaM3−0.29**-0.28**−0.13ns−0.13ns
Old grassland
Soil pHwater0.43***0.46***0.26**0.28**
Total carbon0.02ns0.06ns−0.08ns−0.09ns
AlM3−0.68***na −0.64***na
FeM30.26**0.23**0.39***0.41***
CaM30.40***0.43***0.44***0.46***
Significance of correlation indicated by **, ***, and ns are equivalent to p-value < 0.05, <0.01, <0.001, and non-significant, respectively; na: non-available.
Table
Table 4. Multiple regression equations calculated for soil P indices for two soil layers (0–5 cm and 5–20 cm) for the young and old grassland fields.
Table 4. Multiple regression equations calculated for soil P indices for two soil layers (0–5 cm and 5–20 cm) for the young and old grassland fields.
0–5 cm5–20 cm
Young grassland
PM3 = −0.135 1 AlM3 − 0.52 2 TC + 9.513 3 soil pHwater +248.6 (R2 = 0.818 ***)PM3 = −0.123AlM3 − 1.049 TC + 0.773 soil pHwater + 307.5(R2 = 0.789 ***)
(P/Al)M3 =−2.573 x 10−4AlM3 − 0.003 TC + 0.022 soil pHwater + 0.8 (R2 = 0.893 ***)(P/Al)M3= −2.156x10−4AlM3 − 0.0025 TC + 0.011 pHwater + 0.96(R2 = 0.868 ***)
Old grassland
PM3 = −0.227 AlM3 + 0.377 4FeM3 − 0.015 5CaM3 + 16.792 soil pHwater + 391.9 (R2 = 0.532 ***)PM3 = − 0.235 AlM3 +0.23 FeM3 + 0.011 CaM3 − 4.662 soil pHwater+499.97 (R2 = 0.547 ***)
(P/Al)M3 = −3.2 x10−4 AlM3 + 4.1 x 10−5 FeM3– 1.3 x 10−5 CaM3 + 0.012 soil pHwater + 0.99 (R2 = 0.659 ***)(P/Al)M3 = −4.3 × 10−4 AlM3 − 3.3 × 10−4 FeM3 + 2 ×10−5 CaM3 − 0.008 soil pHwater + 1.2 (R2 = 0.629 ***)
Significance of regression indicated by *** is equivalent to p-value p < 0.001. 1 AlM3: aluminum extracted using Mehlich-3 solution; 2 TC: total carbon was analyzed with an Elementar Vario MAX CN analyzer; 3 soil pHwater (1:1 water) was analyzed according to the method of Hendershot et al. [27]; 4 FeM3: iron extracted using Mehlich-3 solution; 5 CaM3: calcium extracted using Mehlich-3 solution.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Nze Memiaghe, J.D.; Cambouris, A.N.; Ziadi, N.; Karam, A.; Perron, I. Spatial Variability of Soil Phosphorus Indices under Two Contrasting Grassland Fields in Eastern Canada. Agronomy 2021, 11, 24. https://doi.org/10.3390/agronomy11010024

AMA Style

Nze Memiaghe JD, Cambouris AN, Ziadi N, Karam A, Perron I. Spatial Variability of Soil Phosphorus Indices under Two Contrasting Grassland Fields in Eastern Canada. Agronomy. 2021; 11(1):24. https://doi.org/10.3390/agronomy11010024

Chicago/Turabian Style

Nze Memiaghe, Jeff D., Athyna N. Cambouris, Noura Ziadi, Antoine Karam, and Isabelle Perron. 2021. "Spatial Variability of Soil Phosphorus Indices under Two Contrasting Grassland Fields in Eastern Canada" Agronomy 11, no. 1: 24. https://doi.org/10.3390/agronomy11010024

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop