Freshwater algae are ubiquitous and can be grown in media of varying quality, from sterile through to different types of wastewater. They require only water, light, nutrients, and CO₂. Environmental limitations such as growing temperature, pH, and culture densities depend on species. Algae are highly effective producers of organic material and do not compete with agricultural production for arable land. Multiple purposes for algae have been identified including producers of fuel, oil, gas or direct combustion, food and feed, pharmaceutical compounds, plastic, polymer compounds, and fertilizer. While most of these applications are feasible, the majority are not currently economically viable.

The production of algae biomass as a fertilizer has dual value; the use of algae for the purification of wastewater coupled with the application of obtained biomass as fertilizer. The underlying concept is circular economy fertilization and has been reviewed before (Solovchenko et al., 2016). Various methods have been used to recover nutrients from wastewater for agricultural purposes, including sewage sludge (directly or treated) (Duboc et al., 2017), precipitation of nutrients as struvite (Rahman et al., 2011), and algae cultivation. Besides approaches to mainly clean water with algae cultivation (e.g., Powell et al., 2008; Shilton et al., 2012) others were aimed to specifically produce algae biomass from simulated waste water (Gimondo et al., 2019) or real waste water (Mulbry et al., 2005) or different purchased species (Alobwede et al., 2019) to be used as a fertilizer on different plant species. The recycling of nutrients with algae can help to reduce organic contamination from sources such as bacteria and viruses, pollutants such as antibiotics, and nutrients that cause problems with wastewater (reviewed in Bloem et al., 2017). Other benefits, such as the specific accumulation of heavy metals shown in some species, remain to be investigated (Safi et al., 2014).

The ability of algae to accumulate nutrients, directly and specifically, make them a valuable vector for recycling nutrients in agricultural systems. Algae can store excess phosphorus (P) and nitrogen (N) inside their cells resulting in high amounts of P and N that can be stored in vacuoles or the cytosol (Ismagulova et al., 2018). Excess supply of phosphate, or similar P sources, can be stored as polyphosphate granules in storage bodies (Miyachi et al., 1964). Additional nitrogen is stored in form of crystalline guanine rich inclusions (Moudříková et al., 2017).

Phosphorus and N fertilization increase the risks for environmental pollution. However, they are the most essential macroelements for crop growth and therefore their acquisition and application are critical political and resource issues for global food and environment security (Conley et al., 2009). Although P in general is quite abundant in our earth crust (Shen et al., 2011), it limits agricultural yields because it is mainly not in the phosphate form; the form generally considered the only form taken up by plants (MacDonald et al., 2011). Phosphate is largely immobile in soils and is easily adsorbed or precipitated, or organically incorporated into other organisms (Shen et al., 2011). Plants need P in form of phosphate to produce membranes, RNA and DNA, as well as proteins, it is involved in energy metabolism (e.g., for ATP), and the regulation of enzyme activities. The chemical properties of phosphate make it an irreplaceable and important nutrient for plants, but also a scarce resource that introduces competition and investment of resources in the rhizosphere (Steidinger et al., 2015). N and P have been shown to influence each other (reviewed in Güsewell, 2004). N is, compared to phosphate, more mobile in soils and can become gaseous and evaporate. Crop plants such as wheat acquire N in the form of nitrate and ammonium organic forms, such as e.g., amino acids, can also be used by plants native to natural ecosystems (Turnbull et al., 1996). Nitrate uptake is more energy consumptive and ammonium can therefore be favored, but it is species (and pH) dependent (Falkengren-Grerup, 1995). The uptake of organic forms of P and N directly by plant roots has been rarely measured (Turnbull et al., 1996; Richardson et al., 2000; Paungfoo-Lonhienne et al., 2008).

Several studies have analyzed plant responses to algal biomass (Mulbry et al., 2005; Grzesik and Romanowska-Duda, 2014; Mukherjee et al., 2015, 2016; Kholssi et al., 2018; Schreiber et al., 2018). Application of cyanobacteria has a positive effect on maize under mild drought and slightly elevated temperature conditions (Piotrowski et al., 2016). Foliar application of both, cyanobacteria and green algae Chlorella, were shown to improve growth and metabolic activity in willow (Grzesik et al., 2017). Mulbry et al., 2005 found that cucumber and maize seedlings acquired N from algae that had been grown in diluted manure. Mukherjee et al., 2016 showed that rice seedlings had greater height, leaf width and biomass when supplied with different algae species. Polyphosphate was abundant in these algae. Plants receiving algae in unsterile soils, increased phosphate over time although at a rate twice as slow as that in plants provided with mineral phosphate. Evidence of slower P release from algae than mineral fertilizer in rice agrees with previous work of this group (Mukherjee et al., 2015). Grzesik and Romanowska-Duda, 2014 treated maize seeds with Chlorella sp., found positive healthy effects for 14 days in seedlings, and proposed a signaling mechanism rather than a nutrient mechanism. Kholssi et al., 2018 grew wheat for 2 weeks in solution with Chlorella sorokiniana and report that the filtered medium improved both growth and germination more than algae biomass itself. They propose hormonal signals but did not investigate which compounds were responsible for the observed response. Most recently, Schreiber et al. (2018) grew wheat for 55 days-after-sowing (DAS), longer than previous studies and far past the seedling stage when seed reserve is a source of nutrients to plants. Consistent with previous studies, they found that plants acquired P and N from algae biomass when growing in soil. Growth and elemental uptake varied depending on soil type and algal moisture content. In a nutrient-poor soil type, wheat shoot dry weight was 81 and 90% of mineral fertilized controls with application of wet and dry algae, respectively. In a sand, dried algae biomass was not as effective (87%) compared to wet algae (107%), which even exceeded the mineral fertilizer treatment. Interestingly length and density of the root hairs and root diameters were significantly different between plants treated with algae and mineral fertilizer in the nutrient-poor and sandy soils (Schreiber et al., 2018), suggesting nutrient responses typical of mineral P treatments (Kant et al., 2018) and/or other soil solution signals (Sasse et al., 2018a). Schreiber et al. (2018) suggest that algal biomass nutrients are more slowly acquired than those in mineral fertilizer and nutrient-rich soils. In four contrasting soil types the P fractions fluctuated mostly within the first 3 weeks after addition of either algal biomass or NPK fertilizer, while wheat plants took up qualitatively similar amounts of P from either source (Siebers et al., 2019).

Research to date shows that (1) algae biomass is a fertilizer that will release nutrients over time, and (2) algae biomass may be a source of other compounds that plants respond to in terms of growth. The increasing research into the use of algae as a biofertilizer has not resolved the timing of release and growth responses. Resolution of kinetics of nutrient release from algae biomass and the subsequent uptake by plants is expected to help reveal which forms of P are directly or indirectly accessed by the plant, and thereby improve our prediction about what plant stage and time in cropping seasons responses can be expected by a farmer.

Here we ask the question: what is the temporal dynamic of plant growth responses to algal biomass? To address this question, we combine the time resolved wheat leaf growth data from our group (Schreiber et al., 2018) with elemental analyses and other published assumptions to identify critical time points and processes of algal P uptake by wheat. The leaf growth dynamics of wheat plants upon treatment with dried or wet algal biomass or mineral nutrient sources allowed us to propose a model of the timing of P uptake by wheat. We hypothesized that phosphorus from algal biomass is directly available at seedling stage and released gradually over time through to 55 DAS.

Data were analyzed using Excel (version 2013, Microsoft) and R (Rstudio, version 4.0.2). The values for individual plants were either used directly or the arithmetic mean, and the corresponding error of the mean were calculated to represent the whole sample set. The growth function fitting to the data points was performed in Excel, using R² to identify the best fit. R² is the squared least square coefficient that will estimate what portion of the variance in the data is explained by the nutrient source factor in linear regressions. Some datasets yielded R² values <0.8 and are therefore a rather poor fit, but the best one to represent the data. The individual growth curves of plants were used, both exponential and linear phase, to analyze the differences in growth trajectories between different nutrient sources. Differences between growth trajectories were analyzed using a one-factorized ANOVA and HSD Tukey test for multiple comparison of the means (Lee and Lee, 2018) in R to compare the exponent in the exponential phase and the factor in the linear phase as a measure for the different growth trajectories. Differences between individual time-points were analyzed using the same approach for individual time-points. The growth trajectories and time-points were complemented by the calculation-based relative growth rate - calculated using the leaf area at any given time point, the increase over a defined time span, depending on the measurement points and the occurred increase (Paine et al., 2012). They were compared using a one-factorized ANOVA and HSD Tukey test to compare the RGR between different treatments.

The positive relationship between P in the plant and its biomass that is reported here (Figure 1) is known for wheat and other plant species. For example (Rose et al., 2007) similarly, show for pot experiments with spring wheat that P content and biomass increase up until 60 DAS, and that after a phase of exponential increase, the highest relative P content and dry matter is reached at pre-anthesis.

Similarly, the high and low tissue P contents in Figure 1 match previous data on plant growth under P limitation, and have been attributed to the metabolic costs associated with limited P by e.g., exchanging lipids in membranes (reviewed by Nakamura, 2013). However, the high and low P contents, as related to algal moisture content were unexpected. Dried algae biomass led to approximately twice the total P per plant (mean: 24.2 mg shoot P) compared to wet algae (mean: 11.3 mg shoot P) at 55 DAS. We suggest that the dried algae were putatively dead while the wet contained some living cells, and these nutrient sources led to different shoot growth, and greater dead algal material led to more forms of P available to the plant in the soil. The wet algae biomass living cells may have divided and competed for nutrients with the plant roots, which would change the availability of nutrients in general, as shown for other microorganisms (Kuzyakov and Xu, 2013).

The linearity of shoot P and shoot dry weight that we observed in the plants grown with low nutrient levels and wet algae, was not as strong for the plants grown with high nutritional levels (Figure 1). This may have been attributed to their larger size (greater nutrient demand and/or phenology), or limitation of other nutrients such as nitrogen.

Our analyses of leaf area increase (Figure 2) and relative growth rates (Figure 3) over time followed two-tiered trajectories with an initial exponential phase, followed by a linear phase, similar to that described by Rose et al., 2007. The only exception were plants grown in soil 2 + 0, but they were severely starved for P and also had accelerated maturation (BBCH 51–59 at DAS 55). In the early, exponential growth phase no significant differences between the nutrient sources were found, meaning that similar amounts of nutrients were available and taken up during this phase by plants provided with different nutrient sources. The soil 2 + 1x dried algae biomass and soil 2 + 1x wet algae-fertilized plants did access some nutrients from the soil, as did the plants grown in soil 2 + 0, but these were very limited as reflected in their limited leaf area (Figure 2, see Figure 4 for soil 2 + 1x dried algae biomass or soil 2 + 1x wet algae biomass or soil 2 + 1x mineral fertilizer effect on plants). Literature states that after 2 weeks most of the nutrients from the wheat seed have been consumed (White and Veneklaas, 2012), but growth here after 14 DAS did not differ much between the nutrient sources (Figures 2, 3). After 30 DAS, the relative growth rate changed for all plants, indicating that the plants changed their overall growth mode or maybe even entered a different developmental phase, which should be assessed in further experiments. The P values used for our calculations were literature-based and will need experimental verification in the future, because different varieties and growth conditions of mother plants influence the amount and quality of the seed storage in wheat (Kisko et al., 2018).

We assume that root growth was the most relevant growth-limiting parameter for nutrient uptake during the exponential phase, while nutrient release from the added nutrient sources became more dominating once the pot volume had been explored. Root length density (RLD), however, was maximally 0.11 m/cm³ for all three nutrient sources. The phosphate uptake rate will, as published by Newman and Andrews, 1973, not yet be limited at RLDs of 0.4 m/cm³ by direct root competition. In this context it is worth mentioning that Schreiber et al., 2018 identified differences in root morphology, especially in root hair length, which slightly increased when algae, both wet and dried, were the nutrient sources compared to the mineral fertilizer, allowing the former to access a greater volume of soil given the same RLD. Algal biomass will mainly release phosphate close to the root, because microbial and enzymatic activity will be higher there than in the bulk soil, even though phosphatase activity derived from wheat roots has been shown to be comparably wide spread (Ma et al., 2018), but will also depend on the soil type and moisture content. There could be additional enzyme activity from the biomass itself, because algal cells were shown to express phosphatases themselves (Solovchenko et al., 2019). The RLD will, however, not restrict the uptake of more mobile nutrients, which will probably be accessed independent of the RLD at 30 DAS (Newman and Andrews, 1973). The increment in root biomass (Supplementary Figure S2) between soil 2 + 1x mineral fertilizers and the algal treatments (soil 2 + 1x dried algae biomass or soil 2 + 1x wet algae biomass) is not reflected in an increased leaf area. One possible explanation for this might be a bigger need for soil exploitation in the soil 2 + 1x mineral fertilizer plants due to the differences in P availability. This will be elaborated in detail in the next paragraphs.

Our results implicate that, in addition to seed P, an initial amount of P was available early on from all nutrient sources. We assume that residual algal growth media that might still have been attached after a rinse with tap water (Schreiber et al., 2018) was not responsible for the initial growth boost since algae biomass, both dried and wet, led to different growth trends by 30 DAS (Figure 2). Even the effects of additional metabolically active compounds can be ruled out due to this measure (Kholssi et al., 2018). There is some evidence that plants might also be able to take up larger organic molecules and digest them within the roots (Paungfoo-Lonhienne et al., 2010), but the main pathway for P uptake is still considered to be in the form of phosphate using transporters within the outer root membranes; especially root hairs (Gahoonia and Nielsen, 1998). While plants use phytate to store excess P in their cells, algae use polyphosphate to store excess phosphate in granules within storage vacuoles or the cytosol (Aoki and Miyachi, 1964; Ismagulova et al., 2018). Polyphosphate can be the largest portion of P within the algal cells, if they took up excessive amounts of it (Miyachi and Tamiya, 1961b; Solovchenko et al., 2019). Under sufficient growth conditions it might be: 27% phosphate, 57.6% organic compounds and 15.3% polyphosphate (Feng et al., 2016). It has been shown that wheat plants can survive on different organic P sources, e.g., ATP, RNA, and Glucose-1-phosphate, under sterile conditions, but this is limited (Richardson et al., 2000). They can release enzymes and organic acids that can facilitate phosphate degradation of organic compounds and the desorption of phosphate from inorganic compounds or organic matter (Lyu et al., 2016; Yang et al., 2019). Exudation of different components into the rhizosphere will increase the amount available within the rhizosphere by: release of adsorbed phosphate, such as by competition for adsorption sides by the release of organic acids (reviewed in Wang and Lambers, 2019); introduction of enzymes that will speed up the break-down of P containing compounds; or by supplying photosynthates to the rhizosphere microbial population including the mycorrhizal fungi, as carbon is limited in soils (Wang et al., 2016; Sasse et al., 2018b).

We developed a time and space resolved concept of P uptake (Figure 5) based on the information that is available on P forms in Chlorella sp. (Miyachi and Tamiya, 1961a; Feng et al., 2016) and the ability of wheat plants to take it up. We hypothetically displayed the temporal and spatial differences that will occur between a nutrient source with multiple P forms, such as algae biomass, and a mineral source. Our model on the subsequent release of P from algae biomass and its uptake by plants is based on the assumption that plants will usually take up the cheapest P form, phosphate, first (Turner, 2008), because the more laborious the uptake is, the more energy consuming it becomes. We therefore take into account the enzymatic steps that will be needed to release phosphate to be accessed by the plant (Figure 5), as described before (Turner, 2008; Steidinger et al., 2015).

The mineral fertilizer, especially Hoagland solution, will contain free phosphates and it will, upon application, be available at once. Free phosphates on the other hand can easily become adsorbed, immobilized or leached out of reach, in a time dependent manner, and the longer it is in solution in high amounts the larger the portion that is unavailable for the plant (Shen et al., 2011; Deraoui et al., 2015; Demiraj et al., 2017). Living algae contain a metabolic active P supply, in form of phosphate contained in the cytosol (Solovchenko et al., 2016), while other phosphate groups are bound organically, in order of decreasing amounts: in RNA, phospholipids, DNA, as well as proteins or are stored in inorganic form as polyphosphates (Miyachi and Miyachi, 1961). Chlorella vulgaris cell walls are especially robust, and the fast availability of their nutrients is therefore rather surprising (Němcová and Kalina, 2000; Schreiber et al., 2018). The free, inorganic phosphate is, once the algal cell wall is ruptured, assumed to be directly accessible (Figure 5) and similar to the inorganic P from mineral fertilizer also prone to immobilization and leakage. The subsequent release of organic P and polyphosphates into the soil will result in breakdown by either plant or microbial enzymatic activity and the resulting release of phosphate will prolong the P supply of the plant and change the immobilization and adsorption pattern in the soil. The time frame in which free phosphate is abundant will determine the amount of incorporation by microbial activity and adsorption onto soil particles, a subsequent release will therefore reduce the vulnerable time and increase the overall availability. Phospholipids have been shown to have a half-life of almost one day in soils (Zhang et al., 2019), which is similar to what has been described for other compounds. The algal biomass will not have released all its P yet, because as described by Mukherjee et al., 2015, even after 30 days, not all phosphate was released when they incubated algal biomass in soil. They could also show that the involvement of the microbial community is important for the availability of P Mukherjee et al., 2016, which is also an indicator for active facilitation by e.g., enzymes. In addition, polyphosphates can be stable up to 28 days in soil if the pH is between 5.8 and 6.4 (McBeath et al., 2007), unless enzymatic breakdown occurs. If accessed with either phosphatases or phytases, polyphosphate will be degraded within 10 h (Shand and Smith, 1997). The easier accessible compounds such as phosphorylated enzymes and other phosphoproteins, as well as DNA and RNA, might have lost their phosphate groups due to degrading effects prior to polyphosphates. Free phosphates are vulnerable to adsorption onto organic compounds, precipitation, or consumption by microbes, which would transform the available forms into temporarily unavailable ones. The gradual release of phosphate from the larger organic compounds within the algae biomass might prolong these processes and allow a continuous and even increasing supply of nutrients over time. It will be released over time in a specific pattern and our results indicate that algae biomass still supplies available P after 30 DAS up until 55 DAS.

The model presented in Figure 5 and discussed in the paragraph above assumes temporal steps of P breakdown and availability to plants from algae, to explain the quick and continuous growth of wheat compared to mineral, nutrient fertilizer forms. It assumes that plants prefer phosphate to organic forms of P due to the energetics of acquisition. The presented analyses of different P forms and their release as well as their temporally resolved movement from the algal biomass into the plant will be investigated in future experiments.