Microbial release of apatite- and goethite-bound phosphate in acidic forest soils

Highlights

• Phosphate release from hydroxyapatite (HAP) was higher than from goethite.

• Organic acid release by microbes controlled phosphate release from HAP.

• Organic acids caused the acidification of the soil solution.

• Phosphate release from P-loaded goethite was not associated with acidification.

• Microbial phosphate release from HAP and goethite was strongly carbon-limited.

Abstract

Phosphorus (P) is an element crucial for plant nutrition. P can be bound in primary minerals such as apatites or to secondary minerals, such as metal(hydr-)oxides. Microorganisms are capable of releasing mineral-bound P, and thus, transforming it into plant-available forms. This study examined the potential of native microbial communities of five beech forest soils to release P either from hydroxyapatite or P-loaded goethite. Incubation experiments with soil extracts, either with or without glucose, were conducted. Desorption of phosphate from goethite was less effective than the solubilization of phosphate from hydroxyapatite. We found that the net P solubilization from hydroxyapatite was driven by the microbial production of low molecular weight organic acids (LMWOAs), such as D-gluconic and 2-keto-D-gluconic acids, and differed among the microbial communities extracted from the five forest soils. In contrast, the net P desorption rates from goethite did not vary significantly among the microbial communities extracted from the different soils. Microbial acidification of the solution increased the adsorption of phosphate to goethite, whereas less acidic conditions promoted progressive desorption of phosphate. Microbial communities in soil extracts incubated with P-loaded goethite downregulated the release of organic acids, which reduced acidification. The net P solubilization rates from hydroxyapatite and the net P desorption rates from goethite were strongly increased by the addition of glucose, suggesting that microbial P mobilization from minerals is strongly carbon limited. In conclusion, the study shows that microbial communities from acidic beech forest soils are much more efficient at releasing phosphate from hydroxyapatite than from goethite.

Introduction

Phosphorus (P) is an essential and non-substitutable component for all living organisms (Filippelli, 2002, Ruttenberg, 2003). Soil bacteria and mycorrhizal fungi can increase availability of P by mobilizing it from organic and inorganic sources (Jacoby et al., 2017, Hallama et al., 2019). However, the specific processes underlying the release of P from mineral sources, and their contribution to forest P nutrition, are far from being well understood. Traditionally, in temperate and boreal regions, forest productivity is considered to be mostly limited by nitrogen (N), whereas tropical forests are primarily limited by P (Turner et al., 2007, Menge et al., 2012, Darcy et al., 2018). Yet, a growing number of studies have recognized that increases in atmospheric N deposition impact the biogeochemical cycle of P (Hedwall et al., 2017, Dirnböck et al., 2017, Remy et al., 2017, Heuck et al., 2018). Therefore, it is expected that temperate forests may shift from N to NP co-limitation or even P limitation (Peñuelas et al., 2013, Jonard et al., 2014, Heuck et al., 2018).

Although total soil P concentrations can be high (~400–1200 g kg⁻¹; Rodriguez and Fraga, 1999, Borggaard et al., 2005, Rouached et al., 2010), the portion of dissolved P readily available to plants and microorganisms is often less than 0.1% of total P (Zhou et al., 1992, Zhu et al., 2011). The reason for the low availability is that phosphate either exists within poorly soluble primary minerals or becomes increasingly bound to reactive secondary phases, such as aluminium (Al) and iron (Fe) hydrous oxides, with progressing soil development (Walker and Syers, 1976). In alkaline soils, phosphate ions tend to precipitate mainly with calcium (Ca) cations (Hinsinger, 2001). Apatites, including hydroxyapatite (HAP; Ca₅(PO₄)₃OH), are the dominant P-containing primary minerals in most rocks (Nezat et al., 2008). In acidic soils, phosphate ions tend to adsorb strongly to Al and especially Fe hydroxides (Hoberg et al., 2005, Osorio and Habte, 2013). Goethite (α-FeO(OH)) is overall the most common secondary Fe oxyhydroxide in natural environments and is effective in forming strong monodentate- or bidentate-complexes with phosphate ions (Geelhoed et al., 1998, Cornell and Schwertmann, 2003, Antelo et al., 2005). In addition, due to its prevalence in the soil environment, goethite is commonly used in model experiments addressing basic mechanistic features of hydrous Fe oxides.

Precipitation-dissolution equilibria and adsorption-desorption onto variably charged compounds determine to which extent phosphate exists dissolved in solution (Lindsay, 1979, Pierzynski et al., 2005). Soil microorganisms can solubilize mineral-bound P by either dissolving the mineral phases or desorbing sorbed P species. The dissolution of P-bearing minerals and desorption of phosphate from minerals is driven by the release of (i) low-molecular-weight organic acids (LMWOAs), (ii) protons (H⁺), (iii) siderophores and exopolysaccharides (Yi et al., 2008, Frey et al., 2010, Ordoñez et al., 2016). These substances cause release of phosphate either via complexation of metal cations (i, iii) or by acidifying the soil solution (i, ii). The effect of LMWOAs released by roots and microorganisms is partly due to the number of carboxylic groups they bear, which determines their capability to desorb phosphate from metal(hydr-)oxides (Beebout and Loeppert, 2006). Depending on their acidity and the pH of the soil solution, carboxylic groups can dissociate, thus affecting the release of phosphate by the amount of protons released into the solution (Fox and Comerford, 1990, Richardson and Simpson, 2011). Similarly, siderophores have strong affinities for divalent and trivalent metals, especially Fe. Their exudation by plant roots and microbes can cause increased solubilization of Fe oxyhydroxides and subsequently release of sorbed phosphate (Marschner et al., 2011).

Despite the increasing awareness of the role of microorganisms in plant nutrition (Chung et al., 2005, Jones and Oburger, 2011, Beneduzi et al., 2013, Sharma et al., 2013, Panhwar et al., 2014), the potential of microbial communities to release phosphate from mineral phases has not been explored with detail, so far. Laboratory incubation experiments addressing potential effects of microorganisms have mainly been performed with (a) cultured microorganisms (Hoberg et al., 2005, Schneider et al., 2010), (b) high doses of LMWOAs addition (Welch et al., 2002), and (c) initial solutions with adjusted and buffered pH values (Raulund-Rasmussen et al., 1998, He and Zhu, 1998, Welch et al., 2002). Solubilization experiments have demonstrated that microbial communities are effective in releasing phosphate from Ca-P minerals (Hinsinger, 2001). Other authors showed that fungi and bacteria have only a limited capacity to desorb sorbed phosphate from goethite (Hoberg et al., 2005) and, therefore, microbial-mediated desorption of phosphate from Fe hydrous oxides is likely less effective than the solubilization of phosphate from Ca-P minerals.

The objective of this study was to determine the potential of microbial communities in soil extracts from five temperate forest soils along a gradient of P availability to release phosphate from one typical primary mineral source, namely hydroxyapatite, and desorb phosphate bound to secondary minerals, represented by goethite (hereafter referred to as P-loaded goethite). Two soil depth increments were chosen to evaluate whether the microbial communities of different soil depths affect the solubilization of phosphate. Incubation experiments were performed with and without addition of glucose to test for possible carbon (C) and energy limitations. We hypothesized that the net microbial P solubilization from hydroxyapatite results from acidification of the soil solution due to the production of LMWOAs by microbes from different soil depths (hypothesis 1). Further, we postulated that desorption of phosphate from goethite is not promoted by acidification since increasing the positive charge on the mineral causes stronger adsorption of phosphate (hypothesis 2). In addition, we hypothesized that differences in dissolved phosphate availability affect the microbial solubilization of mineral-bound P (hypothesis 3). Finally, we assumed that amendment with glucose stimulates microbial production of LMWOAs and metal-complexing compounds, which cause increased phosphate release from the minerals (hypothesis 4).