The regulation of plant growth and the development and alleviation of the negative effects of environmental stresses during ontogenesis, are important factors determining the productivity of cultivated plants. While it is well recognized that biotic and abiotic stress prevents essentially all crop systems from achieving their yield potential, current understanding of the mechanisms involved, and the strategies to mitigate these effects are limited. Abiotic stresses may be prevented by optimizing plant growth conditions and through provision of water and nutrients and plant growth regulators (PGRs—auxins, cytokinins, gibberellins, strigolactones, brassinosteroids). In addition to these traditional approaches, biostimulants are increasingly being integrated into production systems with the goal of modifying physiological processes in plants to optimize productivity. Plant biostimulants based on natural materials have received considerable attention by both the scientific community and commercial enterprises especially in the last two and a half decades (Crouch and van Staden, 1993a; Herve, 1994; Zhang and Schmidt, 1999; Maini, 2006; Khan et al., 2009; Apone et al., 2010; Craigie, 2011; Sharma et al., 2014; Brown and Saa, 2015; Du Jardin, 2015; Yakhin et al., 2016a). Biostimulants offer a potentially novel approach for the regulation/modification of physiological processes in plants to stimulate growth, to mitigate stress-induced limitations, and to increase yield. In the following review, we do not attempt to discern if the effects of biostimulants on plant productivity is a direct response of plants or soils to the biostimulant application or an indirect response of the biostimulant on the soil and plant microbiome with subsequent effects on plant productivity. Ultimately discerning if biostimulant effects are direct or microbially mediated will be critical to the development of this technology. The general goals of the current review are to provide a comprehensive analysis of the current situation in the field of biostimulants and to develop a science-based theoretical foundation for the conceptualization, classification, and practical application of these materials. A focus of this review is to understand and define the appropriate place of biostimulants among other agricultural products such as plant protection compounds and fertilizers, and to consider the unique attributes of complex, multi-component biostimulants. The structure of the review is based on the consideration of biostimulants in terms of their action on different regulatory and functional systems of plants (signaling, metabolism, uptake, and transport mechanisms, etc.) using both conceptual and methodological approaches. The overarching objective of the work is to highlight innovative concepts and to establish a scientific framework for future development of biostimulant science.

To understand the development of biostimulant science, several seminal publications warrant discussion. To our knowledge, the first discussion of “biogenic stimulant” theory can be attributed to Prof. V.P. Filatov and was started in 1933 in the USSR (Filatov, 1944, 1951a,b; Gordon, 1947; Sukhoverkhov, 1967). Filatov proposed that biological materials derived from various organisms, including plants, that have been exposed to stressors could affect metabolic and energetic processes in humans, animals, and plants (Table 1). Blagoveshchensky (1945, 1955, 1956) further developed these ideas with specific reference to their application for plants, considering biogenic stimulants as “organic acids with stimulating effects due to their dibasic properties which can enhance the enzymatic activity in plants.” Filatov's concept (1951b), was, however, not limited to these compounds alone (Filatov, 1951b). Herve's (1994) pioneering review provides the first real conceptual approach to biostimulants. Herve suggests the development of novel “bio-rational products” should proceed on the basis of a systemic approach founded in chemical synthesis, biochemistry, and biotechnology as applied to real plant physiological, agricultural, and ecological constraints. He suggests these products should function at low doses, be ecologically benign and have reproducible benefits in agricultural plant cultivation. Zhang and Schmidt (1999) emphasized the need for comprehensive and empirical analysis of these products with particular emphasis on hormonal and antioxidant systems as the basis for many important benefits of biostimulants. They discuss the concept of biostimulants as “pre-stress conditioners,” their effects being manifested in improved photosynthetic efficiency, reduction of spread and intensity of some diseases and in better yields. Basak (2008) initiated the systematic discussion on biostimulants and created the conceptual preconditions for the formation of present biostimulant science while Du Jardin (2012, 2015) provided the first in-depth analysis of plant biostimulant science with an emphasis on biostimulant systematization and categorization on the basis of biochemical and physiological function and mode of action and origin. Du Jardin's (2015) analysis and categorization was influential in informing the development of subsequent legislation and regulation in the European Union.

The study and development of biostimulants has been approached utilizing a wide range of methodological approaches including chemical and non-chemical characterization of composition (Crouch and van Staden, 1993b; Yakhin et al., 2005; Parrado et al., 2008; Sharma et al., 2012a,b; Ertani et al., 2013a,b; Aremu et al., 2015a,b), plant growth and yield studies (Khan et al., 2009; Kunicki et al., 2010; Parađiković et al., 2011; Zodape et al., 2011; Yakhin et al., 2012, 2016b; Chbani et al., 2013; Kurepin et al., 2014; Colla et al., 2015; Saa et al., 2015; Tandon and Dubey, 2015; Tian et al., 2015), application of the so-called -omics strategies with variations, including microarray and physiological analysis (Jannin et al., 2012, 2013), transcriptome (Wilson et al., 2015; Goñi et al., 2016), genomic (Santaniello et al., 2013), phenomic and molecular (Petrozza et al., 2014), proteomic (Martínez-Esteso et al., 2016), chemical and metabolomic (Ertani et al., 2014). Ultimately, the integrative synthesis of results from multiple methodologies, particularly when integrated with the most relevant—omic technology, “agronomics,” will be required if the science and legitimacy of plant biostimulants is to advance.

Several significant scientific meetings in the field of biostimulants have been held over the past ten years and have contributed greatly to our understanding of conceptual and methodological development of the biostimulant theory: “Biostimolanti in agrocoltura” (Italy, 2006), “Biostimulators in Modern Agriculture” (Poland, 2008), “Biostimulants and Plant Growth” (Belgium, 2014), among others. Of particular significance were the first (France, 2012) and the second (Italy, 2015) World Congresses on the “Use of Biostimulants in Agriculture” which were valuable in highlighting the development of novel concepts and methodology as applied to biostimulants. While many of the following papers are not published in a peer-reviewed format, they do represent important advances in this field. Dumas et al. (2012), for example, proposed a multi-part approach to study biostimulants based on large-scale genomic approaches and high-throughput screening tests with genetically-modified reporter plants. Others suggested that biostimulant mode of action can be best determined using molecular microarray analysis to identify gene changes in transcript levels (Gates et al., 2012). This approach has the potential to reveal biostimulant activated signaling pathways involved in the stimulation of plant response. Microarray analysis is not, however, adequate and must be supplemented with carefully conducted field testing or high throughput plant phenotyping (Summerer et al., 2013). The complexity of known biostimulant response, the dependency of crop environment and the diversity of biostimulant products demands the application of novel statistical approaches not commonly used in agronomic research (Sleighter et al., 2015). The principle espoused by Sleighter et al. (2015) is based on the identification of a subset of molecular markers that represent the active ingredients in complex biostimulants and then to correlate these markers with observations of plant response. Chemical genomics that utilizes small molecules to perturb target protein function is a useful strategy for biostimulant discovery as it overcomes constraints imposed by traditional molecular approaches that often fail due to gene redundancy and loss-of-function lethality. Botta et al. (2015) proposed probing the function of biostimulants using an enantiomeric analysis of active compounds in the biostimulant coupled with a proteomic profiling approach. In contrast, Conan et al. (2015) proposed identification of the bioactive compounds responsible for the plant growth response by means of a metabolomic profiling of biostimulant products and analysis of their physiological effects through transcriptomic and metabolomic strategies. Such methodology allows the determination of metabolite pathways affected by biostimulants as well as providing insight into gene regulation. To integrate the diversity of methodologies available Santaniello et al. (2015) emphasizes the need to use bioinformatics strategies to analyse similarities and differences in procedures of ingredient extraction and biostimulant formulation in terms of molecular plant responses. This integrative concept can be used to derive new technologies and novel biostimulant products through the identification of new target genes, enzymes and metabolites.

While the development of robust, multi-faceted approach to the analysis of biostimulant composition and function will greatly aid in the development of this field, all advances must ultimately be interpreted in the context of plant response. The complexity of plant response to the environment is daunting and was elegantly highlighted by Krouk (2015) who demonstrated that root response to nitrogen in the environment is mediated by combinations of signaling molecules and nitrogen sources in a manner that cannot be predicted by exposure to single compounds provided individually (Krouk, 2016; Krouk et al., 2009, 2010, 2011). Inevitably, as our understanding of the molecular networks that control plant growth improves our ability to predict plant response to biostimulants under specific environmental conditions, will improve. Only through a combination of methodologies will progress in biostimulant research be possible.

The development of plant biostimulant science, as well as the principles governing its legislation in the context of the existing legal frameworks of plant protection products and fertilizers, requires the development of a clear definition of term “biostimulant.” Currently, the term “biostimulant” is poorly defined and includes many products that have variously been described as biogenic stimulants, metabolic enhancers, plant strengtheners, positive plant growth regulators, elicitors, allelopathic preparation, plant conditioners, phytostimulators, biofertilisers, or biofertiliser/biostimulant (Table 1). One area of significant challenge is evoked in the question “are biostimulants PGRs?” Historically, biostimulants have been considered as a subgroup of growth regulators (Herve, 1994), as plant growth regulators (Huang, 2007), and as subgroup of bioregulators (Basak, 2008). “From a legal point of view, biostimulants can contain traces of natural plant hormones, but their biological action should not be ascribed to them, otherwise they should be registered as plant growth regulators” (Bulgari et al., 2015). Likewise, biostimulants cannot by definition be pesticides or fertilizers (Russo and Berlyn, 1991; Karnok, 2000; Hamza and Suggars, 2001; Banks and Percival, 2012; Du Jardin, 2012; Torre et al., 2013, 2016).

A concise and biologically meaningful definition of biostimulants has eluded researchers and regulators for many years. Table 1 presents a chronological evolution of concept of the term biostimulant. While several of biostimulant definitions presented are useful in their breadth, many of them have significant limitations and are overly generic, while several do not exclude possible effects of nutrients contained within any putative biostimulant product. In practice, biostimulants may deliberately include nutrients for regulatory approval as fertilizers and on occasions the included nutrients or hormones may be responsible for the perceived agronomic benefit. Given the state of public mistrust of many “biostimulant” products, it is necessary to provide a definition of biostimulants that explicitly denies the use of this term for products that do not have biological efficacy or have efficacy only by virtue of the inclusion of known plant hormones or nutrients.

While the adoption of a definition of biostimulants for regulatory purposes is important, any definition of biostimulant should also be based on scientific principles. Several concepts have been proposed to define plant biostimulants. Basak (2008), proposed that biostimulants could be classified depending on the mode of action and the origin of the active ingredient while Bulgari et al. (2015), proposed that “biostimulants should be classified on the basis of their action in the plants or, on the physiological plant responses rather than on their composition.” Du Jardin (2015), however, has emphasized the importance of the final impact on plant productivity when he suggests that “any definition of biostimulants should focus on the agricultural functions of biostimulants, not on the nature of their constituents nor on their modes of actions.” The term “plant productivity” is used here to describe any improvement in plant yield or quality or increased efficiency of production. These concepts reflect important differences in approaches to providing a definition of biostimulants as a discrete category of agricultural products. Thus, biostimulants could be defined by their demonstrated mode of action and origin, or solely by their demonstrated beneficial impact on plant productivity. The challenges in developing a definition are also complicated by the multi-component and largely undefined composition of many biostimulant products and the possibility that the activity of a biostimulant may not be explained by the presence of any individual constituent, but is a result of the interaction of many constituents in the product.

On this basis two approaches to the definition of complex biostimulants emerge. The first is based on the possibility that the biostimulant contains within it, previously unrecognized molecules that are the sole and discrete cause of the observed improvement in plant productivity. This concept emphasizes both the need for clear demonstration of plant productivity benefits and the unknown nature of the mode of action. Thus, a biostimulant could be defined as “a formulated product that improves plant productivity by a mechanism of action that is not the sole consequence of the presence of known essential plant nutrients, plant hormones, plant growth regulators or plant protective compounds.” By this definition, once the primary biological mechanism of biostimulant function has been identified it should henceforth, be subject to classification on the basis of that functional component.

The majority of biostimulants in use today are complex mixtures of chemicals derived from a biological process or extraction of biological materials. The complexity of these mixtures is often considered to be essential to the performance of the biostimulant, and biostimulants may have properties of the whole, that cannot be fully elucidated by knowing the characteristics of the separate components or their combinations. This theory of complexity or “emergence” was described by Mayr (1982), who argued that in many biological systems “the properties of the whole cannot be fully elucidated by knowing the characteristics of the separate components or their combinations.” “The term emergence describes the onset of novel properties that arise when a certain level of structural complexity is formed from components of lower complexity. In the last few decades, emergence has been discussed in a number of different research fields, such as cybernetics, theory of complexity, artificial intelligence, non-linear dynamics, information theory, and social systems organization” (Luisi, 2002). “Emergence” and “emergent properties” are thus closely related with the notion of the “systems biology” (Luisi, 2002; Johnson, 2006; Korosov, 2012; Lüttge, 2012; Bertolli et al., 2014). Emergence was described by Johnson (2006) as “unexpected behaviors that stem from interaction between the components of an application and their environment,” “there is, however, considerable disagreement about the nature of ‘emergent properties.’ Some include almost any unexpected properties exhibited by a complex system. Others refer to emergent properties when an application exhibits behaviors that cannot be identified through functional decomposition. In other words, the system is more than the sum of its component parts” (Johnson, 2006).

Thus, a biostimulant could also be defined as “a formulated product of biological origin that improves plant productivity as a consequence of the emergent properties of its constituents.”

To our knowledge, however, there have been no clear demonstrations that any biostimulant exhibits truly emergent properties. This is not however a unique challenge and all “biological systems are extremely complex and have emergent properties that cannot explained, or even predicted, by studying their individual parts” (Van Regenmortel, 2004). Emergent properties have been demonstrated in the networks of biological signaling pathways (Bhalla and Iyengar, 1999); in system-level study of traditional Chinese medicine (Chen et al., 2014), and in microbial communities (Wintermute and Silver, 2010; Chiu et al., 2014). To adequately explain the biological complexity present in plants and their interactions with the environment, Lüttge (2012) and Bertolli et al. (2014) emphasize that classic reductionist biology/chemistry is indeed insufficient.

While the two theoretical definitions provided in this section share a requirement that the mode of action is unknown, they differ in the core assumption that biostimulant function is a consequence of the discrete components in the biostimulant or as a consequence of the “emergent” properties of the biostimulant as a whole. Each of these definitions is also incomplete in that it is certainly possible that a biostimulant may contain several molecules that act synergistically while not being truly “emergent,” and it is indeed possible and indeed likely, that even if a biostimulant is demonstrated to have emergent properties, that not all components of that biostimulant are required for that property to be expressed.

We propose, therefore, a definition of a biostimulant that integrates these two concepts. Thus, a biostimulant is defined here as:

“a formulated product of biological origin that improves plant productivity as a consequence of the novel, or emergent properties of the complex of constituents, and not as a sole consequence of the presence of known essential plant nutrients, plant growth regulators, or plant protective compounds.”

Consistent with this definition, the ultimate identification of a novel molecule within a biostimulant that is found to be wholly responsible for the biological function of that biostimulant, would necessitate the classification of the biostimulant according to the discovered function.

A review of the history of biostimulants and related products provides insight into the diversity of these products and the development of this field of study. The evolution of biostimulant classifications as described by various authors is presented in the Table 2. To the best of our knowledge, one of the first attempts to categorize biostimulants was provided by Filatov (1951b) when 4 groupings of biogenic stimulants were suggested. Karnok (2000) compiled a list of 59 materials presenting in 15 biostimulants; Ikrina and Kolbin (2004) systematized patent literature and specified 9 categories of natural raw materials used to derive biostimulants; Basak (2008) suggested that biostimulants could be grouped on the basis of single or multicomponent formulations and classified on the origin of the active ingredient, and the mode of action of the active ingredient. Du Jardin (2012) developed a scientific rationale of classification considering 8 categories of biostimulants and subsequently reduced this list to 7 categories (Du Jardin, 2015). Du Jardin (2012) was explicit in his exclusion of microorganisms from his categorization primarily to avoid conflict with existing categorization of microorganisms as biopesticides and sources of plant hormones. Later Bulgari et al. (2015) proposed a biostimulant classification on the basis of their mode of action rather than on their composition.

Many biostimulant products have been classified into completely divergent groups and categories of function, use, and type of activity (Tables 3, 4). For example, humate-based products are often described as soil health amendments while plant growth promoting rhizobacteria (PGPRs) could be categorized as biofertilizers, phytostimulators, and biopesticides (Martínez-Viveros et al., 2010; Bhattacharyya and Jha, 2012). Du Jardin (2015) has proposed that biofertilisers are a subcategory of biostimulants. Seaweed extracts have been considered as biofertilizers (Zodape, 2001) and microorganisms have also been described as biofertilizers (Vessey, 2003; Fuentes-Ramirez and Caballero-Mellado, 2006; Roy et al., 2006; Malusá et al., 2012; Bhardwaj et al., 2014; Malusá and Vassilev, 2014). Some inorganic elements or small molecules that are not known to be essential may also be classified as biostimulants if evidence of plant growth promotion is available (Michalski, 2008; Kleiber and Markiewicz, 2013; Radkowski and Radkowska, 2013). Thao and Yamakawa (2009), for example, consider phosphites to be biostimulants since plant response to phosphites frequently cannot be explained as a consequence of the known anti-fungal function of these molecules. While the categorization of biostimulants by their origin does not, a priori, provide information on their mode of action this categorization may still be a useful tool to aid in the process of discovery and facilitate comparison between similar products.

Registration of products used in agriculture is crucial to ensure their practical, safe and legitimate application. In the absence of a sound definition of biostimulants as a discrete group of products (Basak, 2008), the registration procedure and subsequent classification regime is untenable and this inevitably creates a barrier to trade and development. Various countries, states, and administrative regions have developed different categories for registration of potential biostimulants including terminology such as plant conditioners, “other fertilizers,” supplements, soil improvers, plant strengtheners, fitofortificants, etc. (Basak, 2008; Torre et al., 2013; Traon et al., 2014). In many jurisdictions regulatory practices require an itemized description and identification of substances in all commercial product classifications while in others the registration of non-fully identified substances is allowed if those products are considered of complex composition. There is even a proposal for complex biostimulants to not specify the chemical name (IUPAC) and note as “None” with the definition that “this product is a complex mixture of chemical substances” (Traon et al., 2014). If we accept the concept that a biostimulant is a product of clear benefit but unknown mode of action, then it can only be regulated by its safety and proof of efficacy. For example, in pharmacology it has been suggested that “the demand to demonstrate the mode of action of each single component in a phytopharmaceutical may not be obligatory any more” (Ulrich-Merzenich et al., 2009).

The complex multicomponent nature of many biostimulants clearly complicates discovery of their modes/mechanisms of action, production, registration and use. What is clearly needed however, is a regulatory mechanism to ensure that the products are “generally recognized as safe,” have “a positive benefit on crop productivity” and are discrete from exisiting categories of products. The task of identifying function and agronomic utility can then be pursued independently and will be driven by the marketplace imperative for product quality and consistency. Coordinating national legislation within this framework will become critical for the optimization of biostimulants and trade between different countries. The possible place of biostimulants in the regulatory system of pesticides and agrochemicals is illustrated in Figure 1.

We have conducted an exhaustive analysis of the literature and categorized the majority of the reported biostimulants by origin (Table 4). Microorganisms are widely used for the production of biostimulants and may be derived from bacteria, yeasts, and fungi. These preparations may include living and/or non-living microorganisms and their metabolites. The concept of microorganism-based preparations as biostimulants is described by Xavier and Boyetchko (2002), Sofo et al. (2014), Colla et al. (2015), Matyjaszczyk (2015), and Ravensberg (2015). Different species of algae, mostly seaweeds, are also commonly used for producing biostimulants. Seaweed-based preparations as biostimulants are described in reviews by Crouch and van Staden (1993a), Khan et al. (2009), Craigie (2011), Sharma et al. (2014); and experimental papers by Goatley and Schmidt (1991), Jannin et al. (2013), Billard et al. (2014), Aremu et al. (2015b). Raw materials for biostimulants are also commonly based on higher plant parts including seeds, leaves, and roots and exudates from families Amaryllidaceae, Brassicacae, Ericaceae, Fabaceae, Fagaceae, Moringaceae, Plantaginaceae, Poaceae, Rosaceae, Solanaceae, Theaceae, Vitaceae, among others (Naumov et al., 1993; Yakhin et al., 1998, 2011a, 2012, 2014; Pretorius, 2007, 2013; Parrado et al., 2008; Apone et al., 2010; Ertani et al., 2011a, 2013a, 2014; Colla et al., 2014; Yasmeen et al., 2014; Lucini et al., 2015; Ugolini et al., 2015). Biostimulants may also be based on protein hydrolysates and amino acids of animal origin including wastes and by-products (Mladenova et al., 1998; Maini, 2006; Kolomazník et al., 2012; Ertani et al., 2013b; Rodríguez-Morgado et al., 2014), and insect derived chitin and chitosan derivatives (Sharp, 2013). Humate-based raw materials are widely used to derive biostimulants and have been reviewed by Sanders et al. (1990), Kelting et al. (1998), Ertani et al. (2011b), and Jannin et al. (2012). A final category of biostimulants includes those derived from extracts of food waste or industrial waste streams, composts and compost extracts, manures, vermicompost, aquaculture residues and waste streams, and sewage treatments among others. Because of the diversity of source materials and extraction technologies, the mode of action of these products is not easily determined.

The technologies used in the production and preparation of biostimulants are highly diverse and include cultivation, extraction, fermentation, processing and purification, hydrolysis, and high-pressure cell rupture treatment (Table 4). In some instances, a biostimulant product may also contain mixes of components derived from different sources and production methods. Frequently the rationale for utilizing extracts rather than raw biomass is a consequence of the need for a standardized manufacturing process to produce a uniform commercial product (Michalak and Chojnacka, 2014). For many products, the production processes are driven by process and marketing demands and are not the result of a targeted strategy to optimize the biological efficacy of the commercial product. While the ultimate composition and possible function of commercial biostimulant products may be partially determined by both the source of raw material and the process by which it is prepared (Traon et al., 2014), there may be manufacturing processes and product treatments utilized that result in compounds that are not present in the initial (primary raw) material. An example of this is the multitude of commercial seaweed extracts, often derived from the same species, that are rarely equivalent (Craigie, 2011). Commercial biostimulant manufactured from similar sources are usually marketed as equivalent products, but may differ considerably in composition and thereby in efficiency (Lötze and Hoffman, 2016). Many manufacturers do not reveal the technology of biostimulant production since that is a commercial secret (Traon et al., 2014).

A diversity of substances contained in raw materials is used for the production of biostimulants. Whereas, primary metabolites are contained in most preparations de facto, the presence of secondary metabolites is more specific and depends to a large extent on the raw material used (species, tissue, growing conditions). Primary metabolites include amino acids, sugars, nucleotides, and lipids (Aharoni and Galili, 2011). Secondary metabolites are formed from different primary metabolic pathways, including glycolysis, the tricarboxylic acid cycle (TCA), aliphatic amino acids (AA), the pentose-phosphate and shikimate pathways which are primarily the source of aromatic AA and phenolic compounds (PC), terpenoids/isoprenoids, nitrogen-containing compounds (alkaloids), sulfur-containing compounds (glucosinolates); (Aharoni and Galili, 2011). Frequently, biostimulants are shown to have a multicomponent composition and may include plant hormones or hormone-like substances, amino acids, betaines, peptides, proteins, sugars (carbohydrates, oligo-, and polysaccharides), aminopolysaccharides, lipids, vitamins, nucleotides or nucleosides, humic substances, beneficial elements, phenolic compounds, furostanol glycosides, sterols, etc. (Table 4). While many articles have attempted to describe the composition of complex biostimulants, these descriptions are frequently incomplete since the vast majority of biological molecules that would be present in crude extracts of complex origin, have not yet been characterized and the mere presence of a specific compound does not a priori demonstrate that compound is functional. The composition of most biologically derived biostimulant feed stock will also vary with the season of production, species, physiological state of the source organism and growth conditions. Indeed, there is an implication in the marketing of many biostimulants that stress conditions experienced by the plant or microbe utilized to produce the biostimulant, results in the production of stress metabolites and amino acids with consequent beneficial effects on plant response. In the absence of knowledge of the functional component of a biostimulant, changes in composition of a biostimulant over time and between batches and commercial sources cannot be interpreted. In the most rigorously prepared biostimulants from leading companies, high-throughput analytical methods have been employed to ensure consistent product quality (Sharma et al., 2012b). Methods such as chromatography, mass spectrometry, NMR spectroscopy, elemental analysis, ELISA, spectrophotometry, etc. are typically used for this purpose (Table 4). The complexity of this challenge is illustrated in the analysis of a four-year algae composition sequence using a profile or fingerprint technique employing NMR (Craigie et al., 2009).

Biostimulants have been used at all stages of agricultural production including as seed treatments, as foliar sprays during growth and on harvested products. The mode/mechanisms action of “biostimulants” is equally diverse and may include the activation of nitrogen metabolism or phosphorus release from soils, generic stimulation of soil microbial activity or stimulation of root growth and enhanced plant establishment. Various biostimulants have been reported to stimulate plant growth by increasing plant metabolism, stimulating germination, enhancing photosynthesis, and increasing the absorption of nutrients from the soil thereby increasing plant productivity (Table 4). Biostimulants may also mitigate the negative effects of abiotic stress factors on plants and marked effects of biostimulants on the control of drought, heat, salinity, chilling, frost, oxidative, mechanical, and chemical stress, have been observed (Table 4). Alleviation of abiotic stress is perhaps the most frequently cited benefit of biostimulant formulations. The following text describes the primary modes/mechanisms of action that have been demonstrated or claimed for biostimulants in the primary scientific literature.

Understanding the modes of action of an agricultural chemical has been a fundamental requirement for effective marketing and frequently a regulatory requirement for manufactured products used in agriculture. Mode of action is used here to mean “a specific effect on a discrete biochemical or regulatory process,” thus the “mode of action” of Glyphosate is to inhibit the activity of the enzyme enolpyruvylshikimate-3-phosphate synthase (EPSPS). Biostimulants frequently do not meet this standard of specificity and indeed there are few biostimulant products for which a specific biochemical target site and known mode of action has been identified. For a small subset of biostimulants, however, a demonstrated impact on general biochemical or molecular pathways or physiological processes, termed here as a “mechanism of action,” has been identified even though the explicit “mode of action” may not be known. An example of a “mechanism of action” would be a stimulation of photosynthesis or the down regulation of a plant stress signaling pathway without an understanding of the explicit biochemical or molecular “mode of action.”

For many biostimulant products, however, neither a specified mode of action, nor a known mechanism of action, has been identified. The presence of some spurious products in the marketplace compromises the market for all players resulting in the assumption by many, that biostimulants as a whole, are “snake oils” (Basak, 2008), a pejorative term implying the product is of no value. Multicomponent biostimulants are particularly difficult to reconcile since they may have constituents for which the mode of action is known and components of no known functional benefit. Furthermore, multicomponent biostimulants will frequently contain measureable but biologically irrelevant concentrations of known essential elements, amino acids, and plant hormones etc., for which the mode of action is known but the concentrations are irrelevant when used at recommended rates. Thus, for many of the multicomponent biostimulant in the marketplace today, we propose that a demonstration of a clear “mechanism of action” is a more rationale and attainable regulatory goal than requiring an unequivocal demonstration of the “mode of action.”

Insight into the use of the terms “mode and mechanism” of action can be drawn from the pesticide science and pesticide development. In pesticide science, the “mechanism of action” describes the integral of all the biochemical events following application while the “mode of action” characterizes the main features of a bioactive molecule and its specific biochemical action leading to its effect in treated plants (Aliferis and Jabaji, 2011). In reference to plant bioregulators, Halmann (1990) suggests that ideally an understanding of the mode of action of plant bioregulators on the molecular level requires the identification of the receptor site for each regulator, as well as the elucidation of the subsequent reactions. In reality this standard is often not met in biopesticide or biostimulant products where the identification of the molecular targets of all bioavailable (and frequently uncharacterized) compounds within a given extract cannot be easily achieved. The identification of the target binding sites of the natural biomolecules has, however, proven to be helpful in the design of new insecticidal molecules with novel modes of action (Rattan, 2010).

At the present time, given the difficulty in determining a “mode of action” for a complex multicomponent product such as a biostimulant, and recognizing the need for the market in biostimulants to attain legitimacy, we suggest that the focus of biostimulant research and validation should be upon determining the mechanism of action, without a requirement for the determination of a mode of action. This is the standard of practice for many pharmacological products. With the development of advanced analytical equipment, bioinformatics, systems biology and other fundamentally new methodologies a more complete understanding of the mechanisms and even possible modes of action of these materials may be achieved in the future. While this proposal suggests that the development and marketing of a biostimulant may not require a demonstration of the mode of action, it is still in the interest of the manufacturers of these products to pursue an understanding of the mode of action so that the product can be improved and the use can be optimized for various environments and cropping systems.

The mechanisms of action of all but a few biostimulants remain largely unknown (Rayorath et al., 2008; Khan et al., 2009; Rathore et al., 2009). This is primarily due to the heterogeneous nature of raw materials used for production and the complex mixtures of components contained in biostimulant products which makes it almost impossible to identify exactly the component(s) responsible for biological activity and to determine the involved mode(s) of action (Parađiković et al., 2011). Therefore, focus should be upon the identification of the “mechanisms of action” of biostimulants as indicated by general positive impacts on plant productivity through enhancement in processes such as photosynthesis, senescence, modulation of phytohormones, uptake of nutrients and water, and activation of genes responsible for resistance to abiotic stresses and altered plant architecture and phenology (Khan et al., 2009; Sharma et al., 2012b). An example of this process is the advances in use of protein-based biostimulants for which recent studies have identified the target metabolic pathways and some of the mechanisms through which they exert their effects on plants (Nardi et al., 2016).

To further our understanding of modes/mechanisms of biostimulant action we have systematized the stages of biostimulants action on plants after their application: (1) penetration into tissues, translocation and transformation in plants, (2) gene expression, plant signaling and the regulation of hormonal status, (3) metabolic processes and integrated whole plant effects.

The penetration of amino acids and peptide based biostimulants into plant tissues has been investigated using radiolabeled amino acids (Maini, 2006) and mathematical modeling (Kolomazník et al., 2012; Pecha et al., 2012). The components of a biostimulant preparation of animal origin, labeled with ¹⁴C proline and glycine, were shown to penetrate rapidly into treated leaves and where subsequently distributed to other leaves (Maini, 2006). The mathematical model based on the “mechanism of diffusion” allows the estimation of the time required for the absorption of a minimal amount of the active component of a biostimulant. Furthermore, it describes the process of its transport from the moment of penetration into the leaf until the arrival at more distant tissues (Kolomazník et al., 2012; Pecha et al., 2012). The penetration of protein hydrolysates into a plant tissue occurs via diffusion of protein molecules through membrane pores (Kolomazník et al., 2012) and is energy-dependent (Parrado et al., 2008). Biostimulants must have a good solubility in water or other suitable solvents. This is a precondition for most types of application and for sufficient penetration of active ingredients into internal structures of treated plants. Surfactants and other additives may be required to overcome solubility and uptake limitations including lipophilicity and molecular size of active components (Kolomazník et al., 2012; Pecha et al., 2012).

Ultimately a full understanding of the biological activity of complex biostimulant preparations will require a detailed understanding of the mechanism of action and effects on plant productivity and the identification of the biologically active molecules and their molecular mode of action (Henda and Bordenave-Juchereau, 2014). A wide array of molecular methods has been used to attempt to discern the active compounds found in biostimulants including microarrays, metabolomics, proteomic, and transcriptomics methods. These technologies have been applied to biostimulants to probe changes in gene expression following the application of biostimulants (Jannin et al., 2012, 2013; Santaniello et al., 2013). Further research on the effects of complex biostimulants and their components on the complete genome/transcriptome of plants will be required to understand the mechanisms of action involved in growth responses and stress mitigation (Khan et al., 2009). The search for the mode of action of biostimulants is complicated by the observation that many biostimulants have been shown to induce genes and benefit productivity only when plants are challenged by abiotic and biotic stress. Experimental methods must therefore be developed to produce relevant and reproducible stress conditions so that the application of any molecular tool to probe gene function produces results that are relevant to the purported effects on plant productivity.

The role of signaling molecules in plant response to environmental cues has been an area of active research in plant biology. The process of signal transmission involves the synthesis of signaling molecules (ligands), their translocation, their binding to receptors, the resulting cellular responses, and, finally, the degradation of the signaling molecules (Zhao et al., 2005; Wang and Irving, 2011). When the signaling molecule binds to its receptor, the initial cellular response is the activation of secondary messengers, or intracellular signaling mediators, which cause a further series of cellular responses. Among the substances that may act as secondary messengers are: lipids, sugars, ions, nucleotides, gases, Ca²⁺, cAMP, cGMP, cyclic ADP-ribose, small GTPase, 1,2-diacylglycerol, inositol-1,4,5-triphosphate, nitric oxide, phosphoinosides, and others (Zhao et al., 2005; Wang and Irving, 2011). Generally, a membrane-mediated action is typical for water-soluble compounds, while cytosol-mediated activity is primarily triggered by lipophilic compounds.

Whereas, enzymes interact with their substrates in a geometrical way (“lock and key”), signaling molecules are thought to have a topochemical affinity to their receptors. It is assumed that the interaction of such components at the receptor site is cooperative and quantized (Gafurov and Zefirov, 2007). The bioactive compounds in some biostimulants are also proposed to display signaling activity in plants or induce signaling pathways. Various amino acids (Forde and Lea, 2007; Arbona et al., 2013), and peptides (Ivanov, 2010) function as signaling molecules in the regulation of plant growth and development (Ertani et al., 2009; Mochida and Shinozaki, 2011). Peptide signaling is important in various aspects of plant development and growth regulation including meristem organization, leaf morphogenesis, and defense responses to biotic and abiotic stress (Schiavon et al., 2008). Specific signaling peptides contained in a plant-derived protein hydrolysate have been shown to affect plant growth and development, defense responses, callus growth, meristem organization, root growth, leaf-shape regulation, and nodule development (Matsubayashi and Sakagami, 2006; Colla et al., 2013). Protein hydrolysates from soybean and casein have been shown to act as elicitors to enhance grapevine immunity against Plasmopara viticola (Lachhab et al., 2014).

Proteins may also contain hidden peptide sites, “cryptides” or “crypteins” in their amino acid sequence, which may have their own biological activities, distinct from its precursor (Ivanov, 2010; Samir and Link, 2011). Evidence that cryptides can trigger plants defense reactions have recently been demonstrated (Yamaguchi and Huffaker, 2011) and there are reports of the isolation of cryptides by hydrolysis of proteins from marine organisms, including seaweeds, and cryptides may be present naturally in a variety of biological derived products (Henda and Bordenave-Juchereau, 2014; Hayes et al., 2015).

Many small molecular weight substances are known to participate in signaling cascades in vivo. Exogenous amino acids may affect biological processes by acting directly as signal molecules or by influencing hormone action via amino acid conjugation (Tegeder, 2012). It has been suggested that amino acid based biostimulants are readily absorbed and translocated by plant tissues and once absorbed, they have the capacity to function as compatible osmolytes, transport regulators, signaling molecules, modulators of stomatal opening, and may detoxify heavy metals among other benefits (Kauffman et al., 2007). Sugars (Smeekens, 2000; Eveland and Jackson, 2012) and fatty acids and plant lipids (Kachroo and Kachroo, 2009) are also known to act as signaling molecules and mitigators of stress response in plants (Okazaki and Saito, 2014). Animal based lipid soluble fractions, have also been observed to produce an auxin-like response (Kauffman et al., 2007), while sugars, sucrose, and its cleavage products (hexoses), are also known to act as signaling molecules through regulation of gene expression and by interaction with other hormone signals including auxins. In a sunflower meal hydrolyzate, amino acids, humic substances, microelements, and sugars present in the biostimulant appeared to coordinate, with auxin-like compounds in complex signaling cross-talk promoting plant growth, enhancing plant transplanting success and increasing final crop yield (Ugolini et al., 2015).

Hormones are of central importance for the regulation of metabolic processes and plant development in a complex system of interacting hormones and cofactors, the functions of which are closely intertwined and mutually dependent (Wang and Irving, 2011). Biostimulants developed from humic substances, complex organic materials, seaweeds, antitranspirants, free amino acids (Du Jardin, 2012), and crude extracts of lower (Rathore et al., 2009) and higher plants (Yakhin et al., 2012) have been frequently demonstrated to have an effect on plant hormonal status (Kurepin et al., 2014). While hormone-like compounds may be present in biostimulants, it is also possible that de novo synthesis of hormones may be induced by such preparations in treated plants (Jannin et al., 2012) and amino acids, glycosides, polysaccharides and organic acids are contained in many biostimulants and may act as precursors or activators of endogenous plant hormones (Parađiković et al., 2011). Hormones or hormone-like effects could therefore be responsible for the action of natural biostimulants derived from microorganisms, algae, higher plants, animal, and humate based raw material (Table 4).

Information on currently available biostimulants gives some insight into the possible biochemical and molecular genetic effects of biostimulants derived from different natural raw materials (Table 4). Many published reports are available suggesting various biostimulants improve plant productivity through increased assimilation of N, C, and S (Jannin et al., 2012, 2013), improved photosynthesis, improved stress responses, altered senescence, and enhanced ion transport (Gajic, 1989; Khan et al., 2009; Parađiković et al., 2011). Biostimulants are also reported to increase free amino acids, protein, carbohydrates, phenolic compounds, pigment levels, and various enzymes (Table 4). The protective effect of many biostimulants against biotic and abiotic stresses has been associated with a reduction of stress-induced reactive oxygen species, activation of the antioxidant defense system of plants, or increased levels of phenolic compounds (Ertani et al., 2011a, 2013a).

While it is clear that many biologically derived biostimulants contain small molecular weight compounds that are involved in signaling events and may directly influence plant metabolic processes, it remains unclear how an exogenous soil or foliar application of an uncharacterized product can have predictable and beneficial responses in plants. It is well-known, for example, that application of exogenous plant hormones or compounds that disrupt hormone function (PGR's) can have markedly negative effects on plants and that optimization of PGR materials and their applicaitons requires precise information on dosage and timing. Application of biostimulants for which the dosage and efficacy of the functional compounds is unknown, cannot, therefore, be expected to result in predictable plant responses and identification of molecules with effects on plant metabolic processes is not, in of itself, a sufficient explanation for the function of a biostimulant. It is also uncertain why the application of a biostimulant with purported function as a PGR, signaling molecule or other discrete compound would be superior to, or more easily controlled, than a direct application of the purified product itself.

Modern crop production requires a balance of high and consistent productivity with maximum safety for consumers, agricultural workers, and the environment (Rathore et al., 2009; Jannin et al., 2012; Pecha et al., 2012). While some biostimulants have been analyzed with regard to unwanted side effects including negative impact on the natural environment (Janas and Posmyk, 2013) most biostimulants have not been fully characterized but have been regarded as generally recognized as safe (GRAS in the US) on the basis of the biological origin of their constituents (Thomas et al., 2013). Generally, biostimulants are assumed to be biodegradable, non-toxic, non-polluting and non-hazardous to various organisms. While this may be a rational conclusion for many formulations derived from biological materials such as seaweed extracts and their components (Turan and Köse, 2004; Dhargalkar and Pereira, 2005; Rathore et al., 2009; Michalak and Chojnacka, 2014; Stadnik and de Freitas, 2014), higher plants (Onatsky et al., 2001; Abdalla, 2013; Yakhin et al., 2013), chitin and chitosan (Bautista-Baños et al., 2006; Cabrera et al., 2013) it is not clear that this is a valid assumption for microbial products or products that would not normally be present in agricultural fields.

Biostimulants have been utilized as bioremediants and have been shown to improve ATP levels and phosphatase and urease activity (Tejada et al., 2011a), and hence increase the rate of degradation of xenobiotics in the soil (Tejada et al., 2010, 2011b) and to enhance beneficial soil microbial communities under semi-arid climates (Tejada et al., 2011b). Biostimulants may also help reduce the amount of potentially risky agrochemicals (Kolomazník et al., 2012) including reducing the use of fertilizers and pesticides (Hamza and Suggars, 2001). Most compounds contained in biostimulants are natural constituents of terrestrial and aquatic ecosystems (Jannin et al., 2012) and metabolites of plant and microbial origin and as such most are generally regarded as safe, particularly at the low rates at which they are typically applied. Thus, it has been proposed that biostimulants can be positioned as eco-friendly products for sustainable agriculture (Mladenova et al., 1998; Ertani et al., 2011a; Ghannam et al., 2013; Vijayanand et al., 2014). In many countries, however, biostimulants are not subject to rigorous toxicological screening (Traon et al., 2014) and there remains the potential for the persistence of human pathogens in materials of animal origin and for the synthesis of novel compounds of unknown function or toxicology during the manufacturing process.

Even though there have been relatively few rigorous demonstrations of the benefit of biostimulants, and to a large extent the mode of action of these products remains uncertain, the industry for biostimulants is substantial and rapidly growing. Though many recent “market” studies show that the market for these products is growing at a remarkable rate, the validity of these analyses must be considered with care as they frequently do not provide an explicit definition of term “biostimulants.” The value of the European biostimulants market ranged from €200 to €400 million in 2011, €500 million in 2013 and may grow to more than €800 million in 2018 with annual growth potential in 10% and more (EBIC, 2011a, 2013; Traon et al., 2014). France, Italy, Spain are the leading EU countries in the production of biostimulants (Traon et al., 2014). In North America, the biostimulant market was valued at $0.27 billion in 2013¹, and is expected to grow at a growth rate of 12.4% annually, to reach $0.69 billion by 2018, the USA is the largest producer and consumer of biostimulants in the region (http://www.micromarketmonitor.com/). In 2014, the USA market was assessed at $313.0 million and is projected to reach $605.1 million by 2019², at a CAGR of 14.1% (http://news.agropages.com/). The biostimulants market in the Asia-Pacific was valued at $0.25 billion in 2013, and is expected to grow at a CAGR of 12.9% annually, to reach $0.47 billion by 2018 (Asia Biostimulants Market, 2015)³. China and India are key countries playing a significant role. The Southeast Asian & Australasian biostimulants market was valued at $233.8 million in 2015, and is projected to reach $451.8 million by 2021 (http://news.agropages.com/)⁴. The market in Latin America was valued at $0.16 billion in 2013⁵, and is expected to grow at a CAGR of 14.4% annually, to reach $0.32 billion by 2018 (http://www.micromarketmonitor.com/). This market is mostly concentrated in Brazil and Argentina. The regional market shares of the global biostimulants market⁶ are: EU—41.7%, North America—21.5%, the Asia-Pacific region—20%, Latin America—12.9%. Globally, it biostimulants were valued at $1402.15 million in 2014 and are projected to have aCAGR of 12.5% reaching $2524.02 million by 2019⁷, largely as a consequence of growing interest in organic products. Wu (2016) summised that “the global biostimulants market is projected to reach $2.91 billion by 2021, with a CAGR (compound annual growth rate) of 10.4% from 2016 to 2021. In terms of area of application, the biostimulants market is projected to reach 24.9 million hectares by 2021 and is projected to grow at a CAGR of 11.7% from 2016 to 2021” (Wu, 2016).”

The biostimulant industry faces many problems and challenges. Until recently biostimulant products based on natural raw materials and particularly waste stream has mainly been developed based on observational and less commonly, empirical data. While many contemporary biostimulants have been shown to be effective in practice, very few biostimulants can claim to understand the mechanisms or modes of action (Khan et al., 2009). Furthermore, while biostimulants can be categorized by source of origin, this is frequently inadequate as very substantial differences can exist between products even within a common feed stock origin. The challenge to biostimulant science is further exacerbated since composition and content of active substances in the original plant raw material can be affected by many factors including the location and growing conditions, season, species, variety, organ, and the phase of growth (Naumov et al., 1993; Dragovoz et al., 2009; Sharma et al., 2012b). Similarly, the response of the target crop can be expected to vary across crops and environments. One solution to this problem is to derive the raw materials for the biostimulant under highly regulated conditions. This approach has been successfully implemented by leading seaweed producers and fermentation based products that have developed harvesting and manufacturing processes that ensure uniformity of product performance through time. The development of a product with uniformity of response is not, however, a guarantee that the product is optimized for biological efficacy.

To address these issues, developments in -omics approaches will be critical in accelerating the discovery of mode of action of bioactive compounds (Aliferis and Jabaji, 2011; Craigie, 2011; Jannin et al., 2012) and optimizing their use. Metabolomics, phenomics and agronomics represent the integration of gene expression, protein interactions, and other regulatory processes as they impact on plant productivity and thus are more appropriate tools for discovery in this field than mRNA, transcripts, or proteins analyzed in isolation (Arbona et al., 2013). Integrative, multidisciplinary approaches using tools from transcriptomics in conjunction with metabolomics and biochemical analysis are necessary to establish the mechanism of action and to identify the active components in the extracts (Lee et al., 2012). The difficulty in identifying modes of action and subsequent standardization of composition of multicomponent biostimulants based on natural raw materials will continue to hamper the use, certification and registration of biostimulants. The solution to this problem will require the collaborative efforts of specialists from different fields: chemists, biologists, plant physiologists, industrial manufacture, sales and distribution and those with expertise in practical agricultural production (Raldugin, 2004; Craigie, 2011; Jannin et al., 2012; Lee et al., 2012).

Products with a single active substance represent a simpler construct in which the physiological effects and mechanism of action can be more readily determined and hence certification and registration is simpler. The multicomponent composition of many preparations, however, are much more difficult to characterize (Bozhkov et al., 1996), though they may offer novel insight into biological synergy (Bulgari et al., 2015), multifunctionality and emergence which may be crucial to product efficacy (Gerhardson, 2002). In the absence of a functional rationale for every constituent in a multicomponent biostimulant, it is likely that there will be molecules present that may positively or negatively influence plant productivity. Currently, it is almost impossible using available chemical-synthetic, and genetic engineering approaches to reproduce the full suite of molecules and complexes of biologically active substances (Kershengolts et al., 2008) that are present in most biostimulants.

Many have noted the state confusion in the field of biostimulants (Torre et al., 2013; Traon et al., 2014) and this has resulted in the opinion that much of the biostimulant market is not based on science or efficacy and that many products are little more than recycled waste products sold on the basis of pseudoscience and marketing. Indeed, research on several biostimulant products has shown them to be ineffective or to contain inactive, unstable or inconsistent properties with several showing negative effects compared when contrasted with well-designed controls (Csizinszky, 1984, 1986; Albregts et al., 1988; Di Marco and Osti, 2009; Vasconcelos et al., 2009; Banks and Percival, 2012; Cerdan et al., 2013; de Oliveira et al., 2013; Carvalho et al., 2014). For example, foliar and root application of a product containing amino acids from animal origin have been reported to cause severe plant-growth depression and negative effects on Fe nutrition while a second product containing amino acids from plant origin stimulated plant growth (Cerdan et al., 2013). In another report that tested several biostimulant products it was concluded that “none of the biostimulant products tested achieved a sufficient degree of pathogen control to warrant replacement of or supplementation with conventional synthetic fungicides” (Banks and Percival, 2012), and there have been demonstrated positive and negative impacts and overall questions of the economic feasibility of the use of humic substances for increasing crop yields (Rose et al., 2014). Since biological systems are inherently complex, and given that most biostimulant products have not been characterized and have received relatively little replicated and rigorous independent validation, it is perhaps not surprising that many products are ineffective or highly variable in response. Nevertheless, there are a significant number of rigorous independent reports of benefits from some biostimulant formulations and market growth data demonstrates that there is a good deal of support for these products within agricultural producer communities. That such market growth has occured, even in the absence of a known “mechanism of function” suggests that there are aspects of plant metabolism and productivity constraints that are not understood but are potentially important if we are to achieve the goal of increased global food production.

The market euphoria that is taking place in the biostimulant industry recognizes these unknowns and biostimulants are viewed by many innovators and investors as a mechanism to conduct broadscale, if unfocussed, discovery of novel biologically derived molecules. Much as the exploration of marine organisms, and plants and microbes from diverse ecosystems has led to the discovery of novel pharmaceuticals, so too the development of biostimulants from the broad range of source materials, holds significant promise of discovery. Recent years have seen rapid growth in the number of published studies, increased numbers of scientific conferences and development of legal framework and legislation. These trends will inevitably improve the image of this industry and the efficacy of products. Two significant problems still exist within the industry broadly: (1) preparations of products with highly complex multicomponent and incompletely identified composition make the identification of a primary mode of action extremely difficult and (2) the current classification and legislation/legal framework for regulation of biostimulants is based primarily on source material and not on biological mode of action. Hence there is insufficient capacity to differentiate products, and there is the potential for the successful demonstration of a single product within a biostimulant category, to falsely indicate the efficacy of the group as whole.

Several topical questions need consideration in the future:

Can living cultures of microorganisms, which have the ability to stimulate the growth of plants be referred to biostimulants?

Modern biostimulants are complex mixtures derived from raw materials of highly diverse origin utilizing highly diverse manufacturing processes and as such can be expected to have a broad spectrum of possible biological activity and safety. To distinguish biostimulants from the existing legislative product categories including essential nutrients, pesticides, or plant hormones a biostimulant should not solely function by virtue of the presence of elements or compounds of known function. We propose, therefore, a definition of a biostimulant as “a formulated product of biological origin that improves plant productivity as a consequence of the novel or emergent properties of the complex of constituents and not as a sole consequence of the presence of known essential plant nutrients, plant growth regulators, or plant protective compounds.” Consistent with this definition, the ultimate identification of a novel molecule within a biostimulant that is found to be wholly responsible for the biological function of that biostimulant, would necessitate the classification of the biostimulant according to the discovered function.

This novel definition is inspired by three observations: (1) that the development of the biostimulant industry will inevitably result in the discovery of novel biologically active molecules and that the identification and classification of these molecules will benefit biological discovery more greatly if these molecules are explicitly described than if they were merely labeled as “biostimulants,” (2) that there is a need for the nascent biostimulant industry to explicitly discourage the inclusion of nutrient elements and known biologically active molecules under the guise of a “biostimulant” and (3) that there is a need to recognize that classic reductionist biology/chemistry may indeed be insufficient to explain biological complexity (Luisi, 2002; Lüttge, 2012; Bertolli et al., 2014).

The definition provided here is important as it emphasizes the principle that biological function can be modulated through application of complex mixtures of molecules for which an explicit mode of action has not been defined. The definition also requires a demonstration of beneficial impacts of the biostimulant on plant productivity. Given the difficulty in determining a “mode of action” for a biostimulant, and recognizing the need for the market in biostimulants to attain legitimacy, we suggest that the focus of biostimulant research and validation should be upon determining the mechanism of action, without a requirement for the determination of a mode of action. This can be achieved through careful agronomic experimentation, molecular or biochemical demonstration of positive impact on biological processes or the use of advanced analytical equipment to identify functional constituents. Given the prerequisite multi-component and emergent characteristics of biostimulants, the discovery of the mode of action is likely to require application of new techniques in bioinformatics and systems biology. While the definition proposed here suggests that the development and marketing of a biostimulant does not require a demonstration of the mode of action, it is still in the interest of the commercial producers of these products to pursue an understanding of these products so that the product can be improved and optimized for use in various environments and cropping systems.

While there is a clear commercial imperative to rationalize biostimulants as a discrete class of products, there is also a compelling biological case for the science-based development of the biostimulant science that is grounded in the observation that the application of biological materials derived from various organisms, including plants, that have been exposed to stressors can affect metabolic and energetic processes in humans, animals, and plants (Filatov, 1951a,b). This hypothesis is based upon the premise that functional chemical communication occurs between individuals or organs that favorably modulate metabolic pathways and networks at different plant hierarchical levels. Inter and intra organism communication and consequent molecular and metabolic regulation are at the heart of the science of systems biology and the tools of systems biology will inevitably be critical to the realization of mode of action of many biostimulants. Continued investments by commercial entities in biostimulant research and product development will serve as a critical driver of discovery in this realm and will inevitably lead to the identification of novel biological phenomenon, pathways and processes that would not have been discovered if the category of biostimulants did not exist, or was not considered legitimate.

All authors OY, AL, IY, PB, contributed equally to this review.