ON THE NATURE OF THINGS: ESSAYS New Ideas and Directions in Botany Crops for the future: on the way to reduce nitrogen pollution

Reactive nitrogen (Nr) is a limiting factor for plant growth in agriculture. Traditionally, manure and cover plants are used as an Nr source to support crop growth. Introduction of the Haber-Bosch process at the beginning of the last century greatly affected agriculture, offering relatively cheap access to Nr from N2, and its usage in mass-produced N fertilizers considerably increased crop production. Yet, crop plants take up less than 50% of this applied Nr (Cameron et al., 2013), and runoff nitrate, as the end product of nitrification in the soil, has become a key problem in many agricultural areas leading to contamination of groundwater. In the past decades, nitrate levels in groundwater often even exceeded the upper safe limit, which is currently set to 50 mg/L for short-term and 3 mg/L for long-term exposure (Ward et al., 2018). In addition, fertilizer surplus leads to increased greenhouse gas emission in form of ammonia and nitrogen oxides, contributing to climate change (Cameron et al., 2013; Erisman et al., 2013). Together, N fertilizer use has direct consequences on environmental and human health and the downsides are expected to soon outweigh the benefits of food production (Erisman et al., 2013). Hence, finding a more sustainable way to grow healthy plants that is compatible with high yield and good quality food production, urgently needs more attention. A possible approach is generating plants that are adapted to soils with minimal fertilization.

Reactive nitrogen (N r ) is a limiting factor for plant growth in agriculture. Traditionally, manure and cover plants are used as an N r source to support crop growth. Introduction of the Haber-Bosch process at the beginning of the last century greatly affected agriculture, offering relatively cheap access to N r from N 2 , and its usage in mass-produced N fertilizers considerably increased crop production. Yet, crop plants take up less than 50% of this applied N r (Cameron et al., 2013), and runoff nitrate, as the end product of nitrification in the soil, has become a key problem in many agricultural areas leading to contamination of groundwater. In the past decades, nitrate levels in groundwater often even exceeded the upper safe limit, which is currently set to 50 mg/L for short-term and 3 mg/L for long-term exposure (Ward et al., 2018). In addition, fertilizer surplus leads to increased greenhouse gas emission in form of ammonia and nitrogen oxides, contributing to climate change (Cameron et al., 2013;Erisman et al., 2013). Together, N fertilizer use has direct consequences on environmental and human health and the downsides are expected to soon outweigh the benefits of food production (Erisman et al., 2013). Hence, finding a more sustainable way to grow healthy plants that is compatible with high yield and good quality food production, urgently needs more attention. A possible approach is generating plants that are adapted to soils with minimal fertilization.

N FERTILIZATION AND PLANTS: FROM DESCRIBING TO UNDERSTANDING
Modern agriculture utilizes different mineral sources of N r , such as anhydrous ammonia, ammonium sulfate, ammonium nitrate and urea. All forms of N fertilizers are eventually transformed into nitrate at a high rate by nitrifying bacteria. Field experiments have shown that moderate N fertilization increases plant biomass (e.g., Lea and Morot-Gaudry, 2001). However, overapplication of N fertilizers can also cause a high N r stress ( Fig. 1) due to its effect on soil pH and salinity resulting in the well-known, so-called fertilizer burn. Fertilization with an excessive amount of ammonia and urea also causes toxicity to plants and eventually negatively affects their growth. Many studies point at an inadequacy of standard fertilization procedures commonly used (e.g., Albornoz, 2016). Often, an increase of N r , especially when applied at the wrong developmental stage negatively affects both yield and quality of agricultural products, such as their nutritional value or storage properties (Albornoz, 2016). How these traits are regulated at a molecular level is still not fully understood.
As the major source of N r in the soil, nitrate is taken up and distributed throughout the plant by a large number of different nitrate transporters (Krapp, 2015). Assimilation of nitrate into amino acids within a plant is highly regulated at a stage-, tissue-and even cell-specific level (Krapp, 2015;. Adding to this complexity, nitrate also acts as a signaling molecule regulating the expression of genes orchestrating N metabolism and transport, and developmental programs (Krapp, 2015;Fredes et al., 2019). N r availability affects crop yield not only by regulating vegetative biomass accumulation, but also influencing the onset of developmental programs inducing secondary growth, storage organ formation, and reproductive growth such as flowering and tuber induction Fernie et al., 2020). Despite positive effects on growth, a higher level of N r can also reduce yield by shifting the

NEWS & VIEWS
plant's developmental program (Albornoz, 2016) (Fig. 1). The correct timing of the floral transition is one of the main determinants for productivity and, as such, is tightly controlled by endogenous and exogenous cues such as plant age, hormone status, sucrose availability, photoperiod, and ambient temperature (Srikanth and Schmid, 2011;Wahl et al., 2013). The transition to flowering is also affected by N r availability. In the last decade substantial advances have been made in understanding the role of N as a cue regulating the onset of flowering at the molecular level Teng et al., 2019;reviewed by Fredes et al., 2019). Some of the findings are seemingly contradictory, but can be explained by the various N r sources used, the way they are applied, and general growth conditions such as substrate, day length, light intensity, and temperature regimes ( which are prone to causing plant stress and masking N effects (Fig. 1). Therefore, some of the recent studies have used more natural systems, which are more likely to result in a deeper understanding of the regulation of downstream effects of N signaling, mimicking field conditions and a range of N r concentrations that are nonstressful and to which plants can adapt (e.g., Melino et al., 2015;van Dingenen et al., 2019). In the future, knowledge of these molecular mechanisms will make it possible to develop crops with better performance in limited N r soils, stable or even increased high yield, and little requirement for N fertilization (Fig. 1).

FUTURE DIRECTIONS OF N RESEARCH
Currently, many breeding programs aim at developing crop varieties with high yield on soil with limited N fertilization by focusing on increasing plant nitrogen use efficiency (NUE) via improving N r uptake, its transport within a plant, and N metabolism (Fernie et al., 2020). Knowledge of genes regulating these processes allows accelerating the breeding by using more targeted approaches, e.g., crossing varieties with favorable versions of particular genes associated with NUE. Our current understanding of NUE-associated gene functions, however, mainly originates from research in model plants. Studies of genetic modification of the homologous genes in crops have rarely resulted in improved NUE (Fernie et al., 2020), highlighting the need to acquire a deeper understanding of species-specific N signaling mechanisms, regulating NUE and developmental processes directly, and the importance of standardized growth regimes and the fundamental research basis for each crop species on the way to the development of future crops   (Fig. 1). Unlike modern crops, some ancestral wild species and old varieties were found to better adapt to limited N r conditions due to their more efficient usage of the available N r (Hawkesford, 2017;van Dingenen et al., 2019). Many traits valuable for plant adaptation to limited N r have likely been lost during domestication of ancestral crop species because the artificial selection of the modern high yield crops has been performed on high N r soils where NUE and the tolerance to N r limitation were not considered. Using the approach described by Swarbreck and colleagues (2019) to investigate varieties based on N r responsiveness in addition to NUE as traits will likely increase our basic understanding of the rich source of adaptation strategies of crop wild relatives and might identify target genes for genetic modification in modern crops. We consider this strategy very promising in the endeavor to generate "crops for the future" with minimal N r requirement (Fig. 1).
The demand for the development of sustainable agronomic practices also requires the more efficient use of N fertilizers. Modern knowledge-based methods of N fertilization have a potential to reduce environmental pollution. Among those are more frequent split application of fertilizers at optimal doses necessary for plant growth, use of new fertilizer forms with a slow N r release, soil treatment with nitrification inhibitors that significantly reduce N r losses and increases crop yield (Xia et al., 2017). Therefore, modern N fertilizing management techniques should become mandatory to minimize N r losses.
Lastly, future food production must take into consideration the detrimental consequences of environmental pollution. As such the global N challenges are manifold, including the removal of damage already done. We also strongly believe that it is also important to FIGURE 1. Despite an obvious positive correlation between the level of N r in the soil and plant growth, when applying N fertilizers, balancing between positive and negative effects should be considered. Each crop adapts to N r levels within a species-specific range, beyond which plant development is hampered due to N stress. Moderate soil fertilization avoids "low N r stress", increases crop yield, while achieving high quality with minimal environmental impact. Aiming at high yield, current intensive fertilization practices operate with an excess of N r sometimes even exceeding the upper limit of the "crop-specific adaptation" range, causing "high N r stress". Most importantly, surplus N r has significant detrimental consequences for the environment. Toward decreasing the anthropogenic N r pollution, research should now focus on developing "crops for the future" with better adaptation to lower levels of N r , through increased NUE and N responsiveness.

NEWS & VIEWS
improve relevant communication between different research fields and dissemination of new achievements to farmers and the public. Attempts have been made to reduce agricultural N r pollution by setting strict restrictions and penalties for farmers. Although this policy has been effective in some countries (Swarbreck et al., 2019), farmers often oppose these restrictions, due to the common belief that decreasing fertilizer application reduces profit. As opposed to a penalty-driven system, better training methods for farmers should be developed to foster implementation of advanced crop management strategies in the field (e.g., Eanes et al., 2017).