Grasslands and forests are integral components of the global ecosystem, totally covering about 70% of the earth’s terrestrial area. Both function as the crucial global pool of biodiversity to supply a wide range of species, and their productivity and sustainability modulate global changes ^[1][2][3][1,2,3]. Importantly, grasslands and forests play substantial roles in diverse ecological services to generate tremendous benefits for humans, such as water conservation, sand fixation, carbon sequestration, oxygen release and global biogeochemical cycles ^[4][5][4,5]. Understanding the biodiversity, productivity and sustainability of grasslands and forests and their underlying mechanisms has been of great interest to ecologists for decades.

The biodiversity, productivity and sustainability of grasslands and forests are the net outcomes of various biotic versus abiotic feedbacks between plants and their environment. These can arise through a variety of mechanisms such as resource partitioning, niche divergence, plant–soil and other species-specific interactions ^[6][7][8][6,7,8], but the central driver must be interspecific and intraspecific plant–plant interactions that can be neutral (consummation and recognition), positive (facilitation and kin selection) and negative (competition and allelopathy) to allow local coexistence. The interactions, either beneficial, harmful or commensal, eventually contribute to the biodiversity, productivity and sustainability of grasslands and forests. While most studies have focused on resource competition, environmental factors and global changes, relatively little is known about the importance of allelopathy in grassland and forest ecological processes [9].

A plant may interfere with the growth and establishment of neighboring plants through competition, allelopathy or both. Differing from competition for resources, allelopathy is an interference mechanism in which living or dead plants release allelochemicals exerting an effect (mostly negative) on co-occurring plants ^[10][11][10,11], even within a species (i.e., autotoxicity or intraspecific allelopathy). Four ecological processes, volatilization, leaching, litter decomposition and root exudation, can bring allelochemicals into air or soil. When allelochemicals contact or approach the associated plants, they directly demonstrate allelopathic action by disturbing the systems of photosynthesis, respiration, and metabolism, or indirectly affect target species by altering environmental conditions, particularly for soil physicochemical properties and microbial communities ^[12][13][14][12,13,14]. In fact, allelopathy originates from interspecific and intraspecific plant–plant interactions in grasslands and forests. The first classical case is black walnut (Juglans nigra), which produced and released a 1,4-naphthoquinone (juglone) to interfere with the growth of understory plants thousands of years ago [15]. The allelopathic interference of shrubs in grass through the release of volatile terpenes into southern California coastal grassland was reported in the 1960s [16]. Subsequently, an increasing number of studies have shown that many ecological events occurring in grasslands and forests are associated with allelopathy and certain allelochemicals ^{[9][10][11][17][18]}[9,10,11,17,18].

2. Allelopathy in Grasslands

2.1. Allelopathy Drives Plant Invasion in Grasslands

The occurrence of invasive plants threatens the structure and function of grassland ecosystems, especially in biodiversity and stability [17]. Several plant species have been confirmed to invade grasslands with an allelopathic mechanism. Spotted knapweed (Centaurea stoebe), native to Europe and introduced into North America, is an example of an invasive plant in western American grasslands. Spotted knapweed can take advantage of root-secreted allelochemicals against local grassland species and alter nutrition availability and underground microbial community composition ^[18][19][18,19]. However, the allelopathy of spotted knapweed is conditional, and there is discrepancy between geographical sites. Spotted knapweed does not exhibit allelopathic invasion in eastern American grasslands [20]. Additionally, sufficient light or infection with fungal endophytes can enhance the allelopathic invasion of spotted knapweed in American grasslands ^[20][21][20,21]. Allelopathic invasion of spotted knapweed in American grasslands results in the novel weapons hypothesis (NWH) that the success of plant invasion can be attributed to the allelochemicals of invaders [22]. Generally, allelochemicals of invasive species have little effect on their original neighbors due to long-term mutual adaptation, but as they are novel to the species of the invaded habitat, they exert a strongly allelopathic interference on the native species [22]. Much evidence has demonstrated that allelochemicals appear to confer a competitive advantage to the invasive plants ^{[23][24][25][26]}[23,24,25,26]. However, some studies did not fully support the NWH, and questioned the necessity of secondary metabolites for nonnative species to ensure invasive success ^[27][28][29][27,28,29]. Another hypothesis, the biochemical recognition hypothesis (BRH), postulates that plant seeds can adaptively detect phytochemicals released from potential competitors and respond by extending their period of dormancy until better establishment conditions occur [30]. Leachates from spotted knapweed reduced the germination rate of grassland species. Importantly, they had no effect on seeding biomass, implying that the allelochemicals in the leachates are non-phytotoxic and do not impede plant growth [30]. Although both the NWH and BRH focus on plant-derived chemicals and predict similar results that phytochemicals released from invasive plants inhibit the emergence of native plants, their fundamental mechanisms are distinct. This can be explained either a negative exposure to toxic chemicals by NWH or a positive recognition of facilitative chemicals by BRH ^[22][30][22,30]. Nevertheless, whether the success of invasive plants is attributed to allelochemicals has been debated. Actually, allelopathy is pervasive in invasive plants [31]. Interestingly, allelopathy of native grassland communities seems to increase their resistance to invasion by introduced plants [32], but there was no evidence that native plant communities’ tolerance to allelopathy contributes to the degree of invasiveness of introduced plants. A more vital linkage between allelopathic traits and invasive performance needs to be explored in further studies.

2.2. Allelopathy Exacerbates Grassland Degradation

Grassland degradation is a phenomenon in which grass struggles to grow or hardly survives, which usually leads to an irreversible reduction in grassland productivity and biodiversity [33]. Many factors have been regarded as the drivers of grassland degradation, of which the main factors are natural climate change and human disturbance ^[34][35][34,35]. One early sign of degraded grassland is that the originally dominant species are gradually replaced by other adaptable plants, such as toxic weeds with allelopathic traits ^[36][37][38][36,37,38]. Toxic weeds in degraded grassland are adapted to extremely harsh environmental conditions and exhibit high aggression toward surrounding plants, even poisoning livestock or humans ^[39][40][39,40]. In the process of grassland degradation, toxic weeds not only vigorously compete with forage plants for water and nutrition resources, but also produce a wide range of secondary metabolites to exert allelopathic effects on the establishment of the co-occurring plants, subsequently reducing species richness and exacerbating grassland degradation ^[41][42][43][41,42,43]. Several studies have shown that extracts of toxic weeds, regardless of plant tissues or growing soil, can reduce the seed germination rate and seedling biomass of the receiving plants ^[38][44][45][38,44,45]. However, the allelopathic effects have distinct differences among the extract concentration, extract source and tested species [44]. Many phytotoxic compounds, such as coumarins, flavonoids and terpenoids, have been isolated and identified from toxic weeds. These potential allelochemicals could jeopardize the photosynthesis, respiration, and metabolic system of plants ^[46][47][48][46,47,48]. Stellera chamaejasme and Artemisia frigida are representatives of toxic weeds and generally serve as bioindicators to characterize the degree of grassland degradation. S. chamaejasme is a common toxic weed in the degraded grasslands of northern China, which can restrict the growth of co-occurring plants via root exudates ^[38][49][38,49]. A. frigida, a perennial dicotyledonous semi-shrub species, has a wide distribution range in the global temperate grasslands, covering Eurasian steppes and northern mixed-grass prairies. Differing from the mainly allelopathic pathway of S. chamaejasme, A. frigida can significantly decrease seed germination and seedling growth by emitting volatile organic compounds (VOCs) as allelochemicals ^[50][51][50,51]. This environmental disturbance may severely influence the composition and abundance of VOCs emitted from A. frigida. Artificial damage can induce A. frigida to release more categories and greater concentrations of VOCs [51]. In particular, grazing activity can enhance the allelopathic effect on the growth of other grassland species, suggesting that allelopathy may interact with over-grazing grassland to accelerate the grassland deterioration by frequently simulating A. frigida [52]. Overall, allelopathy is one of the critical factors driving grassland degradation. Comprehensively understanding of how allelochemicals from toxic weeds mediate intraspecific and interspecific plant–plant interactions would be useful for rehabilitating degraded grassland.

2.3. Allelopathy in Pasture Management

A pasture is a piece of grassland that mainly grows forage grass for livestock. Its quantity and quality are closely related to grassland ecosystem health and animal husbandry development. Hence, the management of pasture, whether natural or managed, is essential to ensure adequate forage grass and to support livestock production. Allelopathy-based interspecific and intraspecific interactions have ecological consequences for the productivity and biodiversity of a pasture. Particularly in a managed pasture, pasture weeds can immensely decrease forage yield and quality, negatively affecting livestock production. Fortunately, some forage species can take full advantage of allelopathy and allelochemicals to retard the emergence and growth of co-occurring weeds, from which they will obtain growing benefits ^[53][54][53,54]. For example, rye (Secale cereale) is a cool-season forage species with high frost and drought resistance; it is generally planted in infertile or acid soils due to its strong adaptability. Rye can produce and release benzoxazinoids to selectively inhibit broadleaf weeds, modifying the spectrum of weed species in the pasture ^[55][56][55,56]. Therefore, some fine forage cultivars with allelopathic traits can be used for weed control. In particular, natural allelochemicals released from allelopathic forage cultivars may act as biological herbicides to a large extent, lowering the consumption of chemical herbicides and the cost of pasture management ^[57][58][57,58]. Many studies have shown that the application of allelopathic forage cultivars can effectively control pasture weeds and increase pasture productivity ^[55][57][59][55,57,59]. Notably, allelopathic forage species such as rye not only suppressed the pasture weeds but also succeeding forage species. To avoid failure in rotation systems, it is warranted to select resistant succeeding forage species [60]. Autotoxicity (intraspecific allelopathy) is ubiquitous in pastures. Autotoxicity in pasture has been well verified in alfalfa (Medicago sativa) ^[61][62][63][61,62,63]. Alfalfa is a major forage legume used as a high-quality livestock feed and cultivated in pastures throughout the world. Several phytotoxic phenolics, saponins and medicarpin in alfalfa can remarkably suppress their own seed germination. To attenuate the autotoxicity, the most obvious solution is to develop a new autotoxicity-tolerant alfalfa cultivar. A recent study has picked out the most autotoxicity-tolerant alfalfa from 22 cultivars based on a technique for order of preference by similarity to ideal solution analysis [64], which provides a theoretical basis for the breeding of autotoxicity-tolerant alfalfa cultivars. However, a long-term and large-scale field verification is needed to assess the tolerance of different alfalfa cultivars to autotoxicity. A mixture of diverse forage species is considered as another option to experimentally prove effectiveness in improving forage productivity ^[65][66][65,66]. Directly, some highly allelopathy-tolerant forage seeds can be used as a subsequent alternative for restoring sparse natural grassland caused by allelopathy [67]. Additionally, the pattern of mixing species also has another benefit for pastures. The mixture of rye with berseem clover (Trifolium alexandrinum) may promote rye pathogen-resistant capabilities [68]. In the coexistence system of Artemisia adamsii with Stipa krylovii, volatiles emitted by A. adamsii can strengthen photosynthesis of S. krylovii by enhancing stomatal conductance even with water deficiency [69]. When grown with the P-mobilizing species Filifolium sibiricum, Leymus chinensis exhibited greater shoot and root P content [70]. These positive interactions are prevalent in pastures and mostly attributed to plant–plant chemical communication.

3. Allelopathy in Forests

3.1. Allelopathy in Natural Forests

Natural forests usually possess plant diversity and stable productivity. The role of allelopathy and the mechanisms underpinning it remain poorly resolved in species-rich forests, but allelopathy does contribute to natural forest regeneration. Forest regeneration is commonly considered as a critical ecological process that sustains resource reproduction through the establishment of saplings and the replacement of dead trees; it has profound implications for the perpetuation of tree species in the temporal and spatial dimensions. However, long-term exposure to allelochemicals from woody species may create a barrier effect on the understory-regenerated saplings, resulting in forest regeneration failure. In particular, endangered and rare plant species are inherently difficult to generate due to their scarce propagules and low adaptability. Allelopathy additively reduces the likelihood of the sapling establishment and probably leads to locally rare species’ extinction. Cinnamomum migao and Metasequoia glyptostroboides are two endangered woody species. Their regeneration is extremely restrained, and the natural population would be gradually diminished over time without active management. Generally, most natural populations only occasionally have 1~2 saplings in their understories ^[71][72][71,72]. Recent studies found that leaf extracts or litters of C. migao and M. glyptostroboides dramatically impeded their seedling growth by impairing the lipid structure of the cell membrane, suggesting that autotoxicity might aggravate the obstruction of the natural forest regeneration among some endangered tree species ^[72][73][72,73]. Apart from autotoxicity or self-inhibition, allelochemical-mediated interspecific interactions also hinder natural forest regeneration and impact the plant community’s composition. In the context of forests dominated by two tree species, dominant tree species may chemically inhibit the sapling regeneration of the others. For example, Kandelia obovate and Aegiceras corniculatum are two dominant species in mangrove forests. Leaf litter leachates of K. obovate are detrimental to the propagule germination and sapling growth of A. corniculatum, ultimately modulating the natural regeneration of the whole mangrove forest [74]. In the later successional forests of maple-beech codominance (Acer saccharum and Fagus grandifolia), the abundance of beech progressively increases as maple decreases with the years. This result, in part, can be explained by the allelopathic advantage of beech leading to the regeneration failure of maple ^[75][76][75,76]. Monopolistic herbaceous plants grown in the floor layer may inhibit natural forest regeneration. For example, the natural regeneration of sessile oak (Quercus petraea) is often hampered by the dense moor grass (Molinia caerulea) understory [77]. When watered with root exudates of moor grass, a significant decrease in oak biomass occurred, suggesting the allelopathic interference of moor grass in oak growth. Even though this negative impact was lower than that of resource depletion, it demonstrated the crucial contribution of herbaceous allelopathy to natural forest regeneration [78]. Based on the understanding of the allelopathic mechanisms underlying natural forest regeneration, some appropriate methods of forest management are proposed to alleviate the adverse effects of allelopathy and promote long-term natural regeneration. One of the most direct and efficient ways is to reduce the frequency of allelopathic interactions by removing litter, or eradicating the allelopathic species. Prevention of saplings from potential allelochemicals facilitates the sustainability of forest health ^[78][79][78,79]. In addition, attempts to enhance the diversity of the shrub layer and floor layer may be an alternative way to promote natural forest regeneration [80].

3.2. Allelopathy in Tree Plantations

A tree plantation is an artificial forest for the large-scale production of wood; usually, easily established and fast-growing tree species are selected as a monoculture forest. The productivity and sustainability of tree plantations intimately links the economic and ecological benefits of forestry. However, successive rotations of some forestry species may cause a replanting problem or soil disease, resulting in a decline in productivity and the loss of biodiversity in plantations ^[81][82][81,82]. Although the underlying mechanism for this issue is still being disentangled, a growing amount of evidence has shown that allelochemicals enriched in soil are mainly responsible for this problem ^[83][84][83,84]. Eucalyptus is one of the most widely planted forestry genera on the planet, but it has suffered from autotoxicity for a long time. Most studies have demonstrated that allelochemicals of Eucalyptus penetrate into the soil through the decomposition of litter and leachates, exerting an allelopathic effect on understory plants, thus limiting the regeneration of native vegetation ^[85][86][85,86]. However, Zhang et al. (2016) argued that the poor establishment of indigenous vegetation on plantations mainly arose from Eucalyptus roots rather than Eucalyptus litter. Retention of understory litter is more likely to facilitate the performance of native species [87]. Whatever the case is, a consensus is that allelopathy is more crucial than resource competition in the replanting problem of Eucalyptus plantations [88]. Chinese fir (Cunninghamia lanceolata) is another tree plantation severely disrupted by autotoxicity. Regeneration failure and poor establishment have remained critical problems in monocultural plantations of this species [89]. However, root exudates contribute more to soil allelochemicals than the litter in Chinese fir plantations. Root-secreted allelochemicals, therefore, are considered a primary source leading to the decline in the plantation of Chinese fir [90]. The mixture of multiple tree species is an effective way to improve the self-inhibition and soil deterioration caused by allelopathy and allelochemicals in plantations ^[91][92][93][91,92,93]. In Eucalyptus plantations, Albizia lebbeck, an introduced N-fixing species, has been regarded as a ’good partner’ to Eucalyptus. Mixed-species plantations of Eucalyptus with A. lebbeck increase productivity and maintain soil fertility compared with pure Eucalyptus stands [91]. Similarly, the establishment and productivity of autotoxic Manchurian walnut (Juglans mandshurica) can be improved in the presence of larch (Larix gmelini). Larch root exudates and soil in mixed-species plantations greatly stimulated the growth of Manchurian walnut seedlings and rapidly degraded the allelochemical juglone [92]. The growth and regeneration of Chinese fir is improved in Michelia macclurei and Chinese fir mixed-species plantations. One of the explanations for this beneficial promotion is that there may be interspecific facilitation mediated by the root exudates from M. macclurei, which not only attenuate the release of allelochemicals from Chinese fir roots but also induce a microbial shift to accelerate the decomposition rates of allelochemicals [93]. These studies illustrate the importance of mixed-species stands in plantations. However, most successful mixtures were empirically established from traditional practices, or were assessed from haphazard experimental combinations.

3.3. Tree-Understory Vegetation Allelopathic Interactions

The canopy position and soil occupancy of dominant forest trees remarkably reduce light and soil nutrient availability for understory vegetation. Even so, some shrub and herbaceous species in understory vegetation can adapt to these diverse conditions and coexist with trees. Apart from competition for resources, the allelopathy of the trees is an interference mechanism for the growth of understory vegetation ^[94][95][94,95]. The allelopathic trait of some trees is highly associated with forest abundance and biodiversity, particularly for woody invasive species. The presence of allelopathic tree species in forests can reduce the abundance of understory vegetation, ultimately becoming dense monospecific stands and extending to the whole forests ^[96][97][98][96,97,98]. In this process, allelochemicals may act as a meditator [99]. For the allelopathic effect of trees on understory plants, leaf litter and leachates have long been considered the main source ^[100][101][100,101]. Leaf litter and leachates from trees falling into the ground may prevent the colonization and development of understory vegetation ^[102][103][102,103]. This suppression is mainly attributed to their physical and chemical effects ^[104][105][104,105]. However, allelochemicals from leaf litter and leachates also have a measurable effect on understory vegetation ^{[106][107][108]}[106,107,108]. Through the decomposition of leaf litter, allelochemicals can be gradually liberated into the soil and come into effect by altering soil pH, nutrient availability, the nitrogen cycle and microbial community structures ^[109][110][109,110]. Especially intriguing is leaf litter and leachates that may modify plant coexistence in the grass layer. For example, spotted knapweed and Bromus tectorum exhibit strong competition with each other, while leaf litter and leachates of Pinus ponderosa can mitigate the competitive effect of spotted knapweed on B. tectorum. In other words, the presence of P. ponderosa shifted competitive outcomes through physical and allelopathic effects, thereby indirectly facilitating B. tectorum by more strongly inhibiting spotted knapweed [111]. In some cases, leaf litter and leachates cannot solely show allelopathic potential. It must unite other biotic or abiotic factors to jointly impact the ecological process ^[102][112][102,112]. Prosopis juliflora is one of the world’s most aggressive invasive species, the leaf litter of which causes the increase of total phenolics in soil and toxifies understory vegetation [113]. When incubated with similar levels of leaf leachate from P. juliflora, the content of allelochemicals varies in different soil textures. Sandy soil accumulates higher levels of phenolics than sandy loam soil due to the greater absorption of inactive phenolics fettered in sandy loam soil [112]. In addition, the allelopathic effect of P. juliflora is also limited by soil moisture because their water-soluble allelochemicals in the soil are more likely to be washed away by rain. Therefore, P. juliflora could not manifest their allelopathic potential in humid soil. Only in dry environments, P. juliflora can create a depressive impact on understory plants [114]. Dense understory species with highly allelopathic potential, in turn, may directly slow the growth of trees and indirectly cause trouble by dissolving the fungal hyphae of trees. Garlic mustard (Alliaria petiolata) is a typical understory invasive species that may suppress fungal mutualists via allelochemicals, leading to significant declines in a series of physiological and metabolic functions ^{[115][116][117]}[115,116,117]. Nevertheless, arbuscular mycorrhizal fungi (AMF) strains can be quickly selected by the allelopathic stress from garlic mustard. After the initial decline in AMF abundance, resistant AMF strains gradually displace sensitive AMF strains and the abundance rises again after the long-term invasion of garlic mustard ^[118][119][118,119]. Moreover, as an invader, the novelty of allelochemicals to resident species, regardless of the plant or microorganism, diminishes over time. Ultimately, garlic mustard may enter a new coevolutionary relationship with native competitors and slowly be integrated into the native community ^[120][121][120,121].