Plant and soil tests to optimize phosphorus fertilization management of grasslands

Highlights

• Positive relation between relative forage yield and the P nutrition index (PNI).

• Critical PNI value of 92% separates P-limited and non-P limited grasslands.

• Stronger relationship of PNI with Olsen P than with other soil tests.

• Critical Olsen P stock for a target PNI of 92%: 12.9 kg P ha⁻¹.

Abstract

Developing more sustainable forage systems requires efficient decision support tools for fertilization management. Soil phosphorus (P) tests have long been used as decision support tools for fertilization management but, more recently, plant nutrition indices using the P concentration of shoot biomass were developed to assess the P nutrition status of grasslands. The objectives of this study were to (i) evaluate the relationship between the phosphorus nutrition index (PNI) and the yield response to P fertilization (ii) analyze relationships between PNI and soil plant-available P (SPAP) indicators, and (iii) evaluate PNI assets for P diagnosis in forage system. Five long-term (≥ 9 years) grassland P fertilization experiments under different soil and climate environments in Canada, Switzerland, France and Romania were used. Three SPAP indicators were tested: C_P, the soil solution orthophosphate ions (oPion) concentration (mg P L⁻¹), Olsen P (mg P kg⁻¹), and, a process-based assessment (Qw + Pr) from the sum of oPion in the soil solution (Qw, mg P kg⁻¹) and diffusive oPion with time and C_P (Pr, mg P kg⁻¹). PNI was calculated as sward P concentration divided by the critical P concentration.

The cumulative effect of P fertilization resulted in a wide range of SPAP values. Overall, C_P varied from 0.03 to 3.6 mg P L⁻¹, (Qw + Pr60 min) from 6−52 mg kg⁻¹, and Olsen P from 4−40 mg kg⁻¹. The PNI varied from 48–94% in plots with no applied P, and from 83–121% in P-fertilized plots. A generally positive relationship between relative forage dry matter yield and PNI was established, with a critical PNI value of 92% that distinguishes P-limited and non-P-limited grassland nutrition. Positive relationships between PNI and the three SPAP indicators confirmed that the soil P status influenced the grassland P nutrition status. Critical values on a stock basis for a target PNI value of 92% were similar for Olsen P (12.9 kg P ha⁻¹) and (Qw + Pr60 min) (13.5 kg P ha⁻¹). This study opens perspectives for P diagnosis improvement in forage systems.

Introduction

Grasslands provide an essential source of cattle feed, and the development of more sustainable agricultural systems tends to increase their contribution to livestock production systems (Carrère et al., 2020). In a given environment, forage production is determined largely by the amounts of nutrients supplied by soil reserves and/or organic or mineral fertilizers applied by farmers. The aim of phosphorus (P) fertilization is to meet crop requirements, which are determined largely by the nitrogen (N) supply (Bélanger et al., 1989; Schellberg et al., 1999; Griffin et al., 2002; Valkama et al., 2016). Environmental and economic concerns require developing more sustainable forage systems. Excess P threatens the integrity of terrestrial and aqueous ecosystems (Sharpley and Menzel, 1987; Janssens et al., 1998. Ceulemans et al., 2011), while readily available and high-quality reserves of phosphate rocks will be exhausted in the medium term (Cordell and White, 2011).

Developing more sustainable forage systems requires effective decision support tools for fertilization management. To this end, plant- and soil-based indicators with threshold values for assessing nutrient status and managing fertilizers have been developed. Nutrient concentration ratios in plant tissues were developed as assessment tools for alfalfa (Medicago sativa L.; Walworth et al., 1986) and perennial ryegrass (Lolium perenne L.) (Bailey et al., 1997, 2000). These ratios are also used for natural ecosystems to determine whether biomass production in terrestrial plant communities is N- or P-limited or co-limited by both nutrients (Güsewell, 2004). Soil P tests inform on soil plant available P (SPAP) and provide response thresholds useful in decision support tools for fertilization management (Schulte and Herlihy, 2007; Reinjeveld et al., 2010). However, both plant and soil tests remain of limited general value. Ratios that indicate of N or P limitation vary according to the type of ecosystems. For instance, they are much lower in upland grasslands than in wetlands (Craine et al., 2008; Mamolos et al., 2005). Similarly, there is no agreement on a universal soil P test likely to provide, for a given crop or grassland sward, a single threshold value regardless of the soil type (Schulte and Herlihy, 2007; Jordan-Meille et al., 2012).

More recently, significant advances have been made with the development of innovative assessment tools. The approach of nutrition indices based on the nutrient concentration of shoot biomass allows grassland P nutrition status to be assessed during growth (Duru and Ducrocq, 1997). For P, this approach is more reliable than those based on a single critical concentration since it considers changes in nutrient concentration as a function of sward biomass accumulation and concentration of other nutrients (Duru and Ducrocq, 1997). The P nutrition index (PNI) is an effective tool for P fertilization management in grasslands since it assesses the sward P nutrition status well during growth and can verify and validate fertilization practices a posteriori (Thélier-Huché et al., 1999). The PNI is adequate for the interpreting of the effect of plant P nutrition status on plant growth in grasslands managed at different intensities (Liebish et al., 2013); as well, PNI provides appropriate plant nutrients status evaluation at the interface between agricultural land and saline wetlands in protected saline habitats (Luna et al., 2019). At an ecosystem level, the N nutrition index (NNI) and PNI provide appropriate evaluation of the functional response of species and communities to fertility gradients induced by practices (Garnier et al., 2007; Lavorel et al., 2009).

At the same time, Morel (2002) developed a mechanistic model based on the assumption that the SPAP pool represents the sum of the amount of orthophosphate ions (oPion) in the soil solution (Qw) and the amount of soil P that can diffuse from the soil to the solution over time (Pr). This model assumes that (i) diffusion of oPion at the solid-to-solution interface of soils is quantitatively the dominant process in plant P nutrition and (ii) depletion of oPion concentration at the root surface, due to absorption, creates a gradient of oPion concentration between the root surface, the soil solution, and the soil solid phase. This gradient is the driving force behind the flux of diffusive oPion from the soil solid phase to the soil solution. A general model, based on a Freundlich kinetic equation, was developed to calculate Pr as a function of the oPion concentration in soil solution (C_P) and time (Morel, 2002). This model correctly simulated the changes in SPAP in the 0−5 cm soil horizon oven seven years from a long-term grassland experiment with contrasting P fertilization regimes (Stroia et al., 2007).In the present study, we examined the abilities of C_P (Morel et al., 2000), Olsen P (Olsen et al., 1954), and the sum (Qw + Pr) to assess SPAP in grasslands. Both C_P and Olsen P are used around the world for this purpose (Jordan-Meille et al., 2012; Ziadi et al., 2013).

The objectives of this study were to: (i) evaluate the relationship between PNI and the yield response to P fertilization over a range of soil types and climate conditions; (ii) analyze relationships between PNI and three SPAP indicators in order to compare their abilities to assess the soil P status and (iii) evaluate PNI assets for P diagnosis in forage system. The study relied on five long-term experiments that measured the response of forage yield to P fertilization under contrasting environments representative of grassland ecosystems in Canada, Switzerland, France and Romania. The four sites offered the opportunity to explore large gradients in soil P fertility caused by cumulative effects of P fertilization over several years.