nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Species biodiversity in ecosystems is crucial in providing ecosystem services and their stability and resilience (Oliver et al. 2015a, b). It has been proposed that increased biodiversity prevents the devastation of species by herbivores and pathogens in an ecosystem (Pautasso et al. 2005, Jactel & Brockerhoff 2007). However, epidemics, or recurrent generalized insect attacks in cycles of several years may occur on a few species of trees in natural ecosystems (Hicke et al. 2006, Robert et al. 2020). Devastating attacks may also fall upon a species that can lead to its local extinction when the pathogen is an introduced invasive species, as in the case of Dutch elm disease (Gibbs 1978, Davidson et al. 1999).
The replication of natural ecosystem processes and patterns in agroecosystems is a strategy to improve the stability and resilience of cultivated systems, particularly in crop health (Altieri et al. 1995, Gliessman et al. 1998). However, replicating structures and processes from natural ecosystems to agricultural systems has been questioned since natural ecosystems’ structure, nutrient and energy flow, and regeneration differ fundamentally from agroecosystems (Denison & McGuire 2015).
Increased biodiversity of species and genotypes in agroecosystems has been implemented after epidemics, or disastrous pests have wiped out genetically homogeneous monocultures (Tooker & Frank 2012). Crop rotation, polycultures, and weed tolerance within and outside of crop plots are also practiced to increase biodiversity to control pests and pathogens in agroecosystems (Risch et al. 1983, Landis et al. 2000, Kremen & Miles 2012). Some of those agricultural practices have been inspired by natural ecosystems (Gliessman et al. 1998, Gliessman 2011). Other biodiversity-enhancing practices have been inspired by traditional agricultures or empirical developments of farmers (Wezel et al. 2014, González Jácome 2021). Despite the criticisms and the source of inspiration used to implement practices in cultivated systems, there is agreement that agroecosystems that are more biodiverse in space and time provide more ecosystem services than agroecosystems with low biodiversity (Letourneau et al. 2011, Kremen & Miles 2012, Oliver et al. 2015b). However, most studies on the benefits of biodiversity in the interactions between plants and the species that consume them have been conducted at the species scale and the level of varieties, cultivars, or genetic variation of a crop, almost always without specifying the resistance mechanisms involved in those interactions (Ode 2006, Letourneau et al. 2011, Tooker & Frank 2012). Thus, it is unclear if phytochemical diversity is one of those resistance mechanisms and if the patterns and processes in which it is involved can be replicated or adapted for the management of sustainable agroecosystems. This paper aims to review the role of phytochemical diversity in ecosystems under the premise that part of the structure and processes in which phytochemical diversity is involved in natural ecosystems can be adapted to agroecosystems. First, I briefly review the spatial and temporal structure of phytochemical diversity in ecosystems and its effect on plant consumers. Then, I summarize how the phytochemical diversity structure is generated and maintained by interactions between plants and other species. Finally, I briefly examine the current or potential role of phytochemical diversity in agroecosystem management, considering that there is a high diversity of agroecosystems along a gradient from industrial or conventional agriculture to traditional agriculture (González Jácome 2003, 2021) and that crops have undergone domestication processes that affect their phytochemical diversity.
Phytochemical diversity in ecosystems: structure and function
Phytochemical diversity can be analyzed at different spatio-temporal scales (García-Rodríguez et al. 2012, Wetzel & Whitehead 2020) (Table 1). At the scale of structural diversity, the nearly 200,000 secondary metabolites described so far represent a variety of molecular arrangements ranging from isoprene (68.2 g mol-1 to polymers such as natural rubber latex (600,000 g mol-1). Many secondary metabolites affect more than one metabolic target (Hu & Bajorath 2013, Singh et al. 2021), so that one molecule can affect several species, and the effect depends on the concentration of the molecule to which a plant antagonist or agonist species is exposed (Gershenzon & Dudareva 2007, Hadacek et al. 2011, García-Rodríguez et al. 2012, Moore et al. 2014). Thus, the functional diversity of secondary metabolites exceeds their structural diversity and is critical in the interactions between plants and their agonists and antagonists, as they are involved in plant defense and signaling between plants and other organisms.
Table 1 Classification of the phytochemical diversity. Translated and modified from García-Rodríguez et al. (2012).
| Type of diversity | Spatial scale | Time context |
|---|---|---|
| Alpha phytochemical diversity PDα | That which occurs in mixtures of secondary metabolites (SM) present in plant tissues. Its description includes the number of SM structures (richness) together with the abundance of each of them in a plant tissue. This results in the profile or chemical phenotype of the tissue. The set of tissue chemical phenotypes constitutes the chemical phenotype of the individual. |
(a) The ontogeny of an individual or plant tissue; (b) a moment in time; (c) in different seasons; (d) the same tissue before and after the attack of a biotic or abiotic stress agent. |
| Beta phytochemical diversity at the individual scale PDβi | The degree of turnover in SM composition among the chemical phenotypes of the different tissues of a plant. | (a) Among the tissues or organs within an individual plant; (b) differences in the same tissue over time or along environmental gradients; (c) comparing the same tissue before and after the effect of a biotic or abiotic stress agent. |
| Beta phytochemical diversity at the population scale PDβp | The degree of turnover in the composition of individual chemical phenotypes of a species: (a) within a population; or (b) between populations of a species. |
(a) At a point in time; (b) across generations |
| Gamma phytochemical diversity PDγ | The richness of individual secondary metabolites or chemical phenotypes in the plant communities that make up a landscape. Degree of differentiation between individuals of different species that make up a community or landscape. |
(a) Across seasons. (b) Across generations |
Plant secondary metabolites (PSM) are found in complex mixtures within and among plant tissues and vary according to the ontogeny of individuals and their organs, the fitness value of those tissues or organs, and in response to biotic and abiotic environmental stress (Langenheim 1994, Courtois et al. 2012, Gershenzon et al. 2012). Moreover, all that variation is regulated by the genotype, and the PSM mixtures found in an individual at a given time differentially affect the species interacting with plants. Thus, an additional level of functional phytochemical diversity is attributable to PSM mixtures, which act on species interacting with plants, including natural enemies of herbivores and pathogens (Ode 2006, Züst et al. 2012, Bravo-Monzón et al. 2014).
There may be compounds within a mixture that by themselves intoxicate or repel a plant antagonist, although it is more likely that several compounds act against the attacker in an additive or synergistic manner (Espinosa-García & Langenheim 1991b; García-Rodríguez et al. 2021, Macel et al. 2005, Richards et al. 2016). The relative importance of the activity of single compounds and mixtures varies with the host range of plant attackers. Generalist attackers are more inhibited by medium or high concentrations of mixtures than specialists, which can override the toxicity of mixtures and single compounds and exploit them in the localization of their plant host (Richards et al. 2016, García-Rodríguez et al. 2021). However, nullifying the toxicity of secondary metabolites is costly unless herbivores sequester them and use them in their defense; hence, high concentrations of mixtures will adversely affect specialist herbivores (Cipollini et al. 2014, Petschenka & Agrawal 2016).
At the population or plant community scale in the tropics, the diversity of secondary metabolite mixtures is associated with high herbivore diversity and low herbivory, although the predominance of specialists or generalists in the herbivore community varies with the plant taxonomic group examined (Richards et al. 2015, Salazar et al. 2018).
Structural and functional phytochemical variation in space and time is also combined with variation in nutrient content for plant-consuming species (Au et al. 2013, Wetzel et al. 2016). This combination forms “dynamic chemical mosaics” (Whitham 1983) of phenotypes in tissues, individuals, populations, communities, and landscapes that select consumer species that can take advantage of mosaics and exclude those that cannot take advantage of them. Thus, dynamic mosaics’ variation can be considered a defensive trait (Nyman 2010, Wetzel & Thaler 2016, Pearse et al. 2018, Glassmire et al. 2019). The phytochemical mosaics prevalent in one population commonly differ from those of other populations ( Züst et al. 2012, Bravo-Monzón et al. 2014), and those characterizing one species differ from those of other species in plant communities (García-Rodríguez et al. 2012, Coley et al. 2018, Salazar et al. 2018).
The level of plant resistance to pests and pathogens has been related to the heterogeneity of the dynamic mosaic composition of structural, functional, and nutritional phytochemical diversity (Wetzel et al. 2016, Glassmire et al. 2019). However, the homogeneity or heterogeneity in phytochemical diversity in terms of resistance will depend on the specific pest or pathogen (Espinosa-García et al. 2021). For example, heterogeneity for a generalist consumer, such as Phytophthora cinnamomi, which can attack about 5,000 plant species belonging to more than 40 families (Hardham & Blackman 2018), will be different from that of a specialist such as the Abert’s squirrel, which discriminates the chemotypes of pine trees of the same species when feeding on them (Snyder 1992). The heterogeneity of phytochemical diversity also affects natural enemies of herbivores or pathogens, as the plant’s chemical composition can positively or negatively affect natural enemies (Ode 2006, Pearse et al. 2018, Hauri et al. 2021). Thus, phytochemical diversity affects top-down and bottom-up herbivore population regulation acting directly on herbivores or their natural enemies (Peñaflor & Bento 2013, Bálint et al. 2016, Massad et al. 2017, Gasmi et al. 2019).
At the community scale, odors emitted by different species may hinder or facilitate the location of herbivore hosts or insect predators and parasitoids depending on the composition of the species neighborhood (Barbosa et al. 2009). Moreover, the complex network of volatile signals emitted by plants and their herbivores in seasonal forests results from an “arms race” type of coevolution. Plants make themselves difficult to locate, and herbivores refine their host detection ability by utilizing plant volatiles (Gershenzon et al. 2012, Zu et al. 2020).
The increase in phytochemical diversity has been assumed to be beneficial for plant defense: the higher the phytochemical diversity, the more effective the defense ( Kubo & Hanke 1985, Jones et al. 1991, Richards et al. 2015, Espinosa-García et al. 2021, García-Rodríguez et al. 2021). The few tests of this hypothesis show that the increased diversity of the phytochemical makeup of populations and plant lineages is associated with less herbivory (Bravo-Monzón et al. 2014, Becerra 2015), but in a low-elevation plant community, the negative relationship between increased phytochemical diversity and herbivory holds, whereas at a high elevation community the relationship reverses (Fernandez-Conradi et al. 2021). Moreover, such a relationship was not found at a clonal crop scale (Espinosa-García et al. 2021). In experimental settings where phytochemical mixtures of varying diversity are assayed on herbivores or pathogens, mixtures perform better against generalist fungal pathogens but not always against specialists (García-Rodríguez et al. 2021); increased phytochemical richness in the mixtures did not affect herbivores and pathogens, but single compounds of some mixtures were active against one or more plant consumers. Then, the more phytochemical diversity, the more plant consumers were negatively affected, and the compounds’ activities depended on the consumer identity (Whitehead et al. 2021). Therefore, a high phytochemical diversity at the individual, populational and community scales should protect plants against the local assemblage of herbivores. However, the degree of phytochemical diversity and its activity on plant consumers may vary depending on environmental gradients and the identity of the species that consume plants in a locality (Glassmire et al. 2019, Bakhtiari et al. 2021, Fernandez-Conradi et al. 2021). Nevertheless, the random addition of plant secondary metabolites to a mixture does not assure its efficacy because antagonist interactions among the PSM may occur, and the plant consumers may detoxify that mixture (Whitehead et al. 2021). Likewise, the sole addition of new phytochemical variants in a population does not imply increased stability or resistance against the local assemblage of consumers for the same reasons that may operate at the mixture scale (Espinosa-García et al. 2021). Thus, the novel phytochemical combinations must undergo natural selection, leaving the combinations that add up to the plants’ fitness or the population’s persistence.
Generation and maintenance of phytochemical diversity structure in populations and communities
Phytochemical diversity is maintained and structured according to evolutionary processes (coevolution and adaptation) fostered by interactions between plants and the agonist or antagonist species that consume them (Becerra et al. 2009, Speed et al. 2015, Whitehead et al. 2021). The modes of natural selection (diversifying, frequency-dependent, and stabilizing) in those processes may promote, limit, or stabilize phytochemical polymorphisms in populations (Cates et al. 1983, Sturgeon & Mitton 1986, Snyder 1992, Moore et al. 2014, Bracewell et al. 2018, Kessler & Kalske 2018).
Those natural selection modes imply diffuse coevolutionary processes in which phytochemical diversity (α, β, and γ) is shaped by the species that depend on plants, and the community of plant enemies and mutualists is shaped by phytochemical diversity (Glassmire et al. 2016, 2020, Kessler & Kalske 2018). The shaping of phytochemical diversity by plant enemies can be different, and even opposite, depending on the specialization of the enemies (generalists vs. specialists), if they are keystone enemies or the identity of each one of them (Macel et al. 2005, Stam et al. 2014, Poelman & Kessler 2016, Massad et al. 2017, Glassmire et al. 2020, Cuny et al. 2021). Thus, phytochemical polymorphisms, with several or many variants are expected in plant populations, where individual plants in a population will have phenotypes varying in their resistance and vulnerability to different sets of plant enemies (Squillace et al. 1985, Langenheim 1994, Linhart et al. 2005, Iason et al. 2012).
A mechanism that promotes and maintains phytochemical diversity is diffuse coevolution (Bakhtiari et al. 2021, Volf et al. 2018), which contributes to the variation in chemical profile among populations of a plant species (Macel et al. 2004, Linhart et al. 2005, Bravo-Monzón et al. 2014, Martínez-Díaz et al. 2015) as the herbivore assemblages attacking that species vary among populations, causing local adaptation of plants to those assemblages (Züst et al. 2012, De-la-Cruz et al. 2020). Selection pressure from local consumers acts on phytochemical variation produced by gene drift, founder effects, or that correlated with plant adaptation to local abiotic factors (Linhart & Grant 1996, Defossez et al. 2021). That process is consistent with the geographical mosaic hypothesis of coevolution in which the coevolutionary outcome of the interaction of a plant species with a particular species of a plant consumer can be agonistic, antagonistic or neutral depending on the local set of herbivores, pathogens, and mutualists (Thompson 2001). The local non-focal plant consumers are affected by the changes in the chemical profile of the focal plant and local diffuse coevolutionary process can occur. Hence, those local coevolutionary processes would promote the phytochemical differentiation among populations of a plant species (Bravo-Monzón et al. 2014, 2018, Martínez-Díaz et al. 2015).
The processes producing differences in the phytochemical makeup among populations are relevant for crops that are disseminated by people away from their original place of domestication; the crops in new agroecosystems may have different sets of agonists and antagonists that may change the crops’ phytochemistry. As domestication in situ is a continuous process (Casas et al. 2007), where peasants foster the adaptation of crops to the new local biotic and abiotic conditions producing new varieties (Johannessen et al. 1970), new phytochemical variants are expected.
A critical issue in the persistence of plant populations and communities is the degree of phytochemical differentiation among individuals or plant species, as perceived and processed by plant consumers (Becerra 2007, Sedio et al. 2017, Volf et al. 2018). In agroecosystems, polymorphisms of individual phenotypes are thought to be relevant for the persistence of crops preventing the evolution of resistant strains or biotypes of plant enemies capable of annihilating plant populations (Pimentel & Bellotti 1976, Gershenzon & Dudareva 2007, Gershenzon et al. 2012).
Phytochemical differentiation, or dissimilarity, is produced by the change in the chemical phenotypes in the expression of novel PSM or through new configurations of the ensemble of preexisting PSM in a population or community. The new arrangements may include changes in the relative concentrations of the PSM and the absence of compounds in a phenotype expressed by other individuals (e.g., Becerra 1997). The phytochemical differentiation among of co-occurring woody plant species is much greater in a tropical rainforest than in temperate forests (Sedio et al. 2018) and in North American temperate forests, the differentiation among co-occurring species increases with precipitation and temperature (Sedio et al. 2021).
The phytochemical differentiation within (Robinson et al. 2022) and among individuals in a population, and its differential effects on plant consumers and their natural enemies, has been extensively documented with chemotypes, polymorphic chemical phenotypes, induced changes in the chemical phenotype after the attack of a plant antagonist, or the dynamic mosaics discussed previously (Espinosa-García & Langenheim 1991a, Wheeler 2006, Padovan et al. 2013, Bustos-Segura et al. 2015, Richards et al. 2015, Bálint et al. 2016, Meléndez-González & Espinosa-García 2018, Espinosa-García et al. 2021). Likewise, the differentiation among plant species in communities affecting directly or indirectly the structure of the second or third trophic levels is well documented (Becerra 2007, Richards et al. 2015, Glassmire et al. 2016, Salazar et al. 2018, Volf et al. 2018, Fernandez-Conradi et al. 2021). As dissimilarity increases among plant taxa, the host breadth gets reduced and specialists predominate in the herbivore community (Becerra 2007, Volf et al. 2018) but generalists may increase their activity (Massad et al. 2017). The degree of dissimilarity among individuals or species can be assumed to be functional for the survival of individuals or the persistence of species facing the assemblage of consumers predominant in their communities. However, that differentiation may not be enough to face plant consumers that did not evolve with a plant species, as in the case of Ophiostoma novo-ulmi, causal pathogen of the Dutch elm disease (Gibbs 1978, Brasier 1991). Also, the sensitivity of plant consumers to the differentiation among taxa may vary.
Summarizing the two previous sections: the α-, β-, and γ-phytochemical diversity are very high in ecosystems and landscapes; phytochemical diversity gets arranged in dynamic mosaics of mixtures of secondary metabolites that vary in their concentration and composition within (Robinson et al. 2022) and among individuals with particular chemical morphs in populations, species with unique chemical phenotypes, or in communities whose species differ in their array of phytochemical phenotypes (Defossez et al. 2021). The phytochemical mosaics select consumer species that can take advantage of them and exclude those that cannot. The degree of phytochemical differentiation among individuals and plant species is related to the susceptibility to herbivores and pathogens. Phytochemical diversity is fostered by evolutionary or coevolutionary processes, mainly under an arms-race scenario. Those processes are context-dependent, and their outcome depends on the species’ identity and the local abiotic and biotic conditions.
Potential role of phytochemical diversity in agroecosystem management
In an agroecosystem, phytochemical diversity is structured by two interaction networks: the interactions between the domesticated crop and its associated agonists and antagonists, and the other by the interactions among the non-crop species adapted to that agroecosystem. The crop’s network is structured by domestication-produced changes in the crop’s morpho-physiology and phytochemistry that affect the crop’s agonist and antagonist species and their natural enemies (Lindig-Cisneros et al. 1997, Benrey et al. 1998, Turlings & Wäckers 2004, Rasmann et al. 2005, Ode 2006, Kappers et al. 2011, Chen et al. 2015, Gaillard et al. 2018, Turlings & Erb 2018, Gasmi et al. 2019, Hauri et al. 2021). The non-crop species interaction network is constituted by the species that prosper in agroecosystems, such as microbes, arthropods, and weeds, which do not depend on the crop, although they may compete with the cultivated species. The complexity of those biotic interaction networks depends on their biodiversity, which is determined by the type of agriculture (traditional or industrialized) utilized. Therefore, to analyze which patterns or processes of phytochemical diversity in natural ecosystems might be helpful in agroecosystem management, we have to consider the effect of domestication on the phytochemical makeup of crops or semi-domesticated plants and the variability of agroecosystem types, as practices based on natural ecosystems might be fit only for some of them.
Traditional and industrial agroecosystems. Agroecosystems have very different characteristics depending on what type of agriculture fostered them. In one extreme, traditional agricultures produce many high-diversity agroecosystems, cultivating diverse crops adjusted to the local conditions (González Jácome 2021). There, the peasants continue a fine-tuning domestication process fostering the variability of crops and sowing selected seeds obtained in the agroecosystem for the following agricultural cycle. That implies that the crop’s biotic interactions network mediated by PSM is both under artificial and natural selection. Additionally, the peasants use weeds and eliminate the undesired ones (Chacón & Gliessman 1982, Espinosa-García & Díaz-Pérez 1996, Vibrans 2016); also, the non-crop species may foster natural enemies to control pests and pathogens (Gliessman et al. 1998, Altieri & Nicholls 2004). Thus, the non-crop species are under artificial and natural selection adjusting all the biotic interactions in the agroecosystem to variable socioeconomic and environmental conditions.
On the other extreme, industrial agriculture produces high-yielding, low diversity agroecosystems adjusted to homogeneous monocultures grown in uniform conditions that exclude most non-crop species and pests and pathogens using pesticides; the crop seeds are purchased for every agricultural cycle from companies that continued the domestication process ex situ directed to increase yields and reduce costs by nullifying particular pests or abiotic factors (Bautista Lozada et al. 2012). Thus, the biotic interaction network for those crops is reduced to few insect pests, weeds, and pathogens, which evolve tolerance or resistance to pesticides, and crop mutualists that are introduced each cycle into the agroecosystem. The patterns or processes of phytochemical diversity in industrial agriculture favor an evolutionary analog of an arms-race scenario in which pests, weeds and pathogens adapt to pesticides or resistant crops and humans develop new pesticides or crop variants.
Therefore, the possible adaptation of the natural ecosystem phytochemical diversity patterns and processes to agroecosystems would vary in the traditional to industrial agroecosystem gradient. Traditional agroecosystems bear more resemblance to natural ecosystems than industrial ones; thus, many patterns and processes in which phytochemical diversity is involved may already operate in those agroecosystems.
For the industrial agroecosystems, the possibility of adapting natural ecosystems’ patterns and processes is very low. However, some of those patterns and processes could be implemented by increasing biodiversity within and among crops, field parcels, and the landscape, thus increasing the chemical interactions implicit in biotic interactions (Valiente-Banuet et al. 2015, Sirami et al. 2019).
The complete reliance on phytochemical diversity to protect crops against pests and pathogens is very difficult, particularly in simplified agroecosystems occupying a large part of the landscape. Pests and pathogens thrive in large homogeneous monocultures with few weed species and impoverished soils. In those agroecosystems, pests and pathogens adapt to synthetic pesticides or the crop varieties resistant to single pests, and the crop antagonists move freely among individual plants and parcels. Thus, biological control should be implemented to suppress pests and pathogens, taking advantage of the volatiles emitted by crops (Peñaflor & Bento 2013).
Plant domestication and phytochemical diversity. Plant domestication occurs under artificial selection to favor traits that fulfill human needs, but natural selection continues operating on the plants under the domestication process eliminating the variants unable to survive after artificial selection in the human-intervened ecosystems. The domestication process may disrupt the coevolved biological interactions between the plants and their antagonists and agonists (Macfadyen & Bohan 2010) by changing and reducing the variability of morphological and phytochemical traits and the genetic makeup that were present in wild plants ( Lindig-Cisneros et al. 1997, 2002, Blanckaert et al. 2012, Albores-Flores et al. 2018, Hernández-Cumplido et al. 2021). The type and degree of disruption vary depending on the domestication stage (Table 2), the life history of the crop, and the ecosystem where the domestication process occurs. For example, for perennial plants domesticated for their fruits in situ (that is, in or near the same ecosystem where the wild relatives occur) and that later could be planted in home gardens or rain-fed plots, the disruption can be minor (Casas et al. 2007, Avendaño-Gómez et al. 2015).
Table 2 Proposed stages of domestication according to the levels of intensity of change undergone by plant populations subject to human manipulation and domestication. Slightly modified from Clement (1999). There may be intermediate stages between two stages of domestication. Taken and translated from Bautista Lozada et al. (2012).
| Domestication status | Population changes |
|---|---|
| Wild | Natural population whose phenotypes and genotypes have not been modified by human manipulation. |
| Evolution in systems under intensive human management | Species that grow in environments disturbed by humans, with possible genetic changes not produced by artificial selection (e.g., weeds). In extreme cases some species only grow in these environments. |
| Incipient domestication | Human intervention at least by promotion (propagation) or tolerance in the system, but with the average phenotype of the selected trait still within the range of variation found in wild conditions. The variance of this average is possibly lower in populations managed under this stage and a reduction in genetic diversity begins to occur. This process occurs in situ. |
| Semi-domestication | Significant differences are evident with respect to wild populations due to human manipulation. The average and variance of the selected phenotype differ and increase, respectively, with respect to wild populations. The variance increases because phenotypes appear due to human manipulation that are not found in wild populations, and that will gradually disappear by natural selection in these populations. There is also a reduction in genetic diversity due to the bottleneck effect. Even so, the plant possesses ecological adaptability to reproduce and survive without depending on human care. This process occurs in situ. |
| Domestication | The plant is completely dependent on the agricultural environment and human care to survive and reproduce. Genetic diversity is usually significantly reduced and ecological adaptability is lost. However, when the domestication process occurs and continues in traditional agroecosystems, the crop’s genetic variability is fostered by peasants. |
| Domestication for monoculture in simplified agroecosystems | The domestication process is carried out ex-situ, in laboratories, greenhouses, or experimental fields. The crop is completely dependent on simplified agroecosystems, synthetic agrochemicals and human care to survive and reproduce. Genetic diversity is null or highly reduced and ecological adaptability is lost. |
By contrast, in genetically engineered annual crops for monocultures in intensive agriculture, the disruption occurs in practically all the interaction networks in which their wild relatives participated (Macfadyen & Bohan 2010). The effect of domestication on the phytochemical diversity and the second and third trophic levels has been reviewed several times or subjected to meta-analysis (Bautista Lozada et al. 2012, Chen et al. 2015, Whitehead et al. 2017, Fernandez et al. 2021); the general agreement is that domestication favors herbivores and natural enemies by reducing chemical defense in the harvested parts of the crop but not necessarily in other plant parts. However, recent studies complicate understanding the effect of domestication on phytochemical defense. For example, low phytochemical diversity in monocultures may facilitate plant colonization by insect herbivores but hinder host localization by natural enemies; some compounds emitted after herbivore attack increase the susceptibility of some herbivores to insect pathogens, but other herbivores are not affected by those compounds (Gasmi et al. 2019); cucumber varieties differing in the quality of their emission of volatile mixtures attract generalist carnivore mites differentially (Kappers et al. 2011), but the genotypic variants of Asclepias syriaca emitted similar bouquets of volatiles after herbivore attack, but there, the amount of the mixture determined the natural enemies attack on the herbivore (Wason & Hunter 2014); mixtures of tomato cultivars with different odor chemotypes affected pest growth and survival, but hindered the hunt activity of predators (Hauri et al. 2021); domestication in tomato changed the volatile mixture emission increasing the localization and colonization of the domesticated tomatoes by a specialist herbivore and decreasing the attraction of natural enemies (Li et al. 2018); and although domestication reduces most chemical defenses in maize, herbivory from generalists increases but herbivory from specialist herbivores decreases (Gaillard et al. 2018). The emerging pattern is that domestication reduces chemical defense concentration and the variability of the phytochemical mixtures, increasing herbivory mainly in the plant parts harvested by humans, but natural enemies frequently do not control herbivores. Generalist herbivores benefit from the phytochemical changes caused by domestication, but specialist herbivores do not always benefit from those changes. Those trends are more evident in annual crops than in tree crops, even the clonal ones.
Lessons from phytochemical diversity in natural ecosystems. The patterns or processes of phytochemical diversity in natural ecosystems are diverse and variable, mainly determined by biotic interactions and context-dependent; that is, idiosyncratic, because the evolutionary and ecological outcomes of the biotic interactions mediated by phytochemical diversity depend on the identity of the interacting species, the local ensemble of species and the local abiotic environment. Therefore, copying those natural ecosystem patterns and processes to agroecosystems is not viable, but adjustments can be made to get some resemblance. However, five recommendations based on ecosystem patterns and processes in which phytochemical diversity is involved could be helpful in agroecosystem management:
1.- The idiosyncratic nature of the outcomes where phytochemical diversity is involved implies that the responses to the phytochemical diversity of the species interacting with crops in an agroecosystem should be investigated case by case before deciding the best strategy to manage crop agonists and antagonists. That means that the sensitivity to the phytochemical mosaics of the agroecosystems should be investigated for the pest and mutualist species interacting with crops.
2.- Increasing phytochemical diversity at the individual, population, or community scale is desirable in agroecosystem management. However, the sole augmentation of diversity is not enough; the new phytochemical variants at those scales could be favorable to plant antagonists or disadvantageous for the crops and, therefore, eliminated by natural selection. Thus, the phytochemical variants that increase the plant fitness are the ones that should be considered an effective increase in phytochemical diversity.
3.- The phytochemical differentiation among crop individuals or cultivars at local agroecosystems should be investigated to be effective against local pests and pathogens. Also, the number of variants or varieties that will be intercropped should be investigated to prevent the evolution of devastating pests or pathogens. Homogeneous populations of plants with a low level of genetic variation, such as a field planted with an improved cereal cultivar, are assumed to be more susceptible to destruction by pests and pathogens (Pimentel & Bellotti 1976, Tooker & Frank 2012, Chaudhary 2013). The simultaneous cultivation of several cultivars with different resistance factors or cultivars with several genes for resistance to a pathogen or pest can avoid catastrophic attacks (Tooker & Frank 2012, McCarville et al. 2014). Although some of these strategies have worked for some annual crops (Borg et al. 2018, Tratwal & Bocianowski 2018, Grettenberger & Tooker 2020), phytochemical diversity is more important than cultivar diversity in pest and disease control (Hauri et al. 2021).
4.- The design of biopesticides based on phytochemical mixtures should avoid antagonistic interactions among their components and foster synergic interactions. Biopesticides based on a single PSM should be combined with other biopesticides as cocktails to prevent the evolution of resistance in pests and pathogens.
5.- The reduction in concentration and diversity in the chemical plant defense of crops suggests that effective pest control should include the recruitment of natural enemies to protect crops. The recruitment through volatile mixtures emitted by crops should be applied, excluding noxious phytochemicals for the natural enemies (Peñaflor & Bento 2013).
Concluding remarks
The patterns and processes in which phytochemical diversity is determinant are highly variable and context-dependent; moreover, the outcomes of those processes are determined by the identity of the species involved. Also, the function of phytochemical diversity is modulated by antagonist and agonist species that have evolved or coevolved with plants. Thus, copying those patterns and processes to industrial agroecosystems is challenging. However, in traditional agroecosystems, several (adjusted or resembling the originals) are already operating; therefore, there is no need to copy them. Instead, the principles supporting those patterns and processes in traditional agroecosystems could reduce the negative impacts of industrial agroecosystems. Still, some general principles derived from the structure and function of phytochemical diversity can be implemented in agroecosystems.
The knowledge of phytochemical diversity has increased rapidly since the advances in metabolomics and informatics; however, many research areas must be covered to disentangle the principles behind the extreme variability and diversity of plant secondary metabolites.
