As mentioned above, cover crops are usually planted in the fall, early or late winter, or summer depending on the type of cash crop to be planted after cover crop termination in annual cropping systems [7,27,28,29]. In orchards and vineyards, cover crops are usually planted in fall or early winter [8,26,30]. The cover crops are generally terminated before the planting of the cash crop in annual cropping systems or before the orchard crops or grapevines resume active growth after dormancy [30] because the termination date can negatively affect the emergence of cash crops [31] and result in crop yield losses [9,18]. As a general rule, under north American conditions, specifically in the southeast, it was suggested that cover crops should be terminated two to four weeks prior to cash crop planting [32]. Such research-based recommendations generated from studies on the effect of cover crop termination time on trees or grapevines in orchards and vineyards does not seem to exist.

Timing of cover crop planting can have direct implications for the growth rate and amount of biomass accumulation by the cover crop because of a longer growing season, amount of nitrogen fixed in the case of legumes [33], and amount of weed suppression [29,30,34]. For example, it was reported that cover crops produced 40% less biomass, and less nitrogen production by the legumes, when they were planted in mid-October compared to early-September [33]. Haring and Hanson [30] attributed some suppression of weed biomass by cover crops to early planting compared to late planting. However, a longer growing season may also mean more biomass accumulation in both the cover crop and the weeds [28,29]. Studies have also reported that cover crop planting density can also be a factor in weed suppression. For example, Brennan and Smith [34] documented a positive correlation between the amount of weed suppression and cover crop plant density and stated that early-season canopy development by cover crops was important in weed suppression. However, in a study in Australia, it was observed that there was no relationship between cover crop density and weed suppression [35]. Most of the published reports, globally, seem to indicate that suppression of weeds is more in terms of biomass accumulation of the weeds than the density of the weed per se [18,30,36,37,38,39].

A study conducted in central Spain concluded that the termination method for cover crops can be critical in optimizing cover crop benefits because it can impact cash crop productivity in annual cropping systems and the ecosystem services from cover crop usage [40]. However, this may not be the case for crop productivity in perennial cropping systems but there are very few studies showing the effect of the cover crop termination method on crop productivity. For example, in an annual crop system in central Italy, a study compared the termination of cover crops at different times with a roller-crimper and glyphosate applications and observed that the sunflower (Helianthus annuus) yield was similar between the two systems when the cover crop was rolled late but not when it was rolled at an earlier stage of the cover crop and hence, the authors suggested that early termination of cover crops with a roller-crimper may have to be combined with glyphosate applications [41]. A study compared the chemical, mechanical, and chemical + mechanical termination methods of cover crop termination and their effect on cotton (Gossypium hirsutum) emergence and yield, and reported some differences in the crop emergence but no effect on yield [42]. Another study compared two different mechanical methods of cover crop termination on mulch, weed cover and nitrogen but not on crop yield [43]. Kornecki and Kichler [44] compared different cover crop terminations with different roller-crimper types and their effect on the cantaloupe (Cucumis melo) yield, but no comparison was made with other cover crop termination methods. However, it can be argued that regrowth of cover crops by an inappropriate termination method could be an issue for crop productivity, especially if the cover crop is still growing actively during the bud break of grapevines in vineyards or the onset of active growth after dormancy in orchards.

The most common methods for cover crop termination include herbicide application, tillage/incorporation, rolling/crimping, burning, mowing [45], and natural winterkill [46]. Generally, broad-spectrum postemergence herbicides, such as glyphosate, paraquat, glufosinate, 2,4-D, etc., are used for the termination of cover crops [46,47], depending on the type of cover crop. For example, it was reported that glyphosate was effective in terminating cereal rye and wheat (Tricticum aestivum) cover crops, but not as effective in terminating legumes.

Cover crop residue incorporation with tillage has been shown to be effective in cover crop termination in studies conducted in the US [52] and Denmark [53]; this practice, however, can cause shifts in soil microbial communities and cause damage to the soil structure [54,55,56,57]. Studies from Spain [40] and Italy [58] concluded that cover crop rolling with a roller-crimper was becoming more popular, and this practice enhanced soil health and beneficial biological activity compared to tillage methods. However, rolling methods used alone seem to negatively affect the cash crop yield, as they are less efficient in controlling weed and cover crop populations [52,53]. Studies in Europe have suggested that rolling methods in combination with flaming or herbicide treatments can improve shortcomings of the sole reliance on rolling [40,59].

A study in France reported that frost, as a termination method, had benefits on soil characteristics when compared to rolling and herbicide methods of cover crop termination [60]. However, this method would only be effective in climates characterized by cold winters and when using winter-sensitive cover crop species.

Mowing and tillage are commonly used for cover crop termination in perennial systems [25,61,62]. However, there seems to be less research and publications involving the impacts of different cover crop termination methods in perennial systems likely because cover crop adoption in perennial cropping systems is less common [21,24]. A study reported using flail mowers to terminate cover crops in almond (Prunus dulcis) and walnut (Juglans regia) orchards in California [30]. Another study reported mowing and allowing the cover crops to senesce as a termination method in vineyards [63]. Perhaps mechanical means of cover crop in orchards and vineyards may be safer than herbicides because the chemicals could drift to the crops and cause phytotoxicity. Usually, as mentioned earlier, cover crops are terminated in spring in orchards and vineyards to avoid competition during the stage when the crops are just resuming active growth after winter dormancy.

One of the main goals of cover cropping is to enhance soil health properties but cover crops can aid in weed suppression because of the interactions between the cover crops and weed species. Such interactions that may result in weed suppression could occur during the actively growing phase of the cover crop or after the cover crop dies or is terminated and left as a surface mulch/residue. The various possible interactions are summarized in a conceptual diagram. Plant interactions that aid in weed suppression include direct competition for plant growth resources, allelopathy, facilitation, and indirect interactions [64]. According to the competitive production principle, a species in a shared niche will influence the environment and cause a negative reaction in the other species [65]. Cover crops and weeds may share specific niches in certain cropping systems, causing competition and the suppression of one group by the other.

Direct competition by manipulation of the seeding rate and method of a rye cover crop was reported to suppress weeds [66]. Biomass and traits, such as plant height, canopy area, and leaf shape also affect the outcome of plant competition [67]. Thus, biomass produced by cover crops can affect light transmittance by creating shaded areas, reduced moisture availability, and reduced soil temperature which in turn can affect the germination of weed seeds [68,69]. Reduced light availability to the weeds in the understory by a taller canopy of subterranean clover (Trifolium subterranean) was attributed as a mechanism of weed suppression in a study conducted in the Netherlands [70]. While biomass and leaf area affect competition for light, root length affects nutrient competition [71]. The aboveground plant parts are a direct result of belowground root growth, so plants with rapid root expansion and colonization of root zones are more competitive [72,73]. If cover crops decrease the resource capture of weeds through adjustments to the microclimate, they may out-compete weeds thereby reducing weed pressure in agricultural production systems. Weed suppression by cover crops due to modifications in the soil microclimate has also been reported [74]. Similarly, a study attributed weed suppression in the form of the colonization of weed seeds by bacteria and fungi brought about by soil microbial changes by cover crops [75].

While competition is a major mechanism of plant interaction, the physiological properties of cover crops can also influence weed population dynamics; non-competitive interference, such as the chemical interaction of plants, i.e., allelopathy, can cause harm between plant species [76]. It has been reported that allelochemicals produced by certain cover crop species can have a suppressive effect on weeds, and the study documented a linear relationship between allelochemicals produced by a rye cover crop and percent weed inhibition [14]. Several other studies conducted in North America have reported allelopathic weed suppression by a rye cover crop [77,78,79,80,81]. Other cover crop species, such as sunn hemp (Crotalaria juncea), cowpea (Vigna unguiculata), and velvet bean (Mucuna deeringiana) have also been reported to suppress weed germination and growth by allelopathic processes [82]. Similarly, there are several reports of allelopathic weed suppression by sorghum (Sorghum bicolor), barley, and wheat [83,84]. Some studies have reported allelopathic effects of the cover crops on the following cash crop [85]. Koehler-Cole et al. [86] published a review on the allelopathic effect of winter cover crops on several row cash crops. Legumes, such as velvet bean (Mucuna pruriens), have also been reported to have suppressive effects on weeds in a field experiment in Mexico with corn [87], which perhaps is an example of physical rather than allelopathic suppression. A study in Spain evaluated the allelopathic effects of aqueous extracts from several plant species to explore their potential as a cover crop. Species included Bromus hordeaceus, B. rubens, Festuca arundinacea, Hordeum murinum, H. vulgare, Vulpia ciliata, Medicago rugosa, M. sativa, Trifolium subterraneum, T. incarnatum, Phacelia tanacetifolia, Sinapis alba, and Pinus sylvestris on three weed species Conyza bonariensis, Aster squamatus, and Bassia scoparia. Their results showed differential effects of the extracts in the suppression of the three weed species and concluded that aqueous extracts of some of these species demonstrated that they had potential to be used as cover crops for weed suppression [88].

Indirect interaction between cover crops and weeds includes the cover crop mulch acting as a physical barrier for weed seedling emergence [79] and can also cause shifts in weed populations when cover crops impact the presence of other biocontrol agents, such as omnivorous predators. For example, it was reported that red clover (Trifolium pratense L.) cover crops increased seed predation through the increase of predator activity, density, and frequency; the impact of cover crops in this experiment resulted in weed seed removal [89].

Depending on the species of cover crop, different plant interactions may occur, affecting different species portions in the weed populations. For instance, crimson clover reduced the eastern black nightshade emergence due to physical suppression, while rye reduced yellow foxtail possibly due to allelochemicals produced by the cover crop [90].

The effects of cover crops on weed suppression is highly variable and influenced by many different factors and their interactions. Mainly, in cases where cover crops have been successful in weed suppression, they have been reported to either reduce weed seedling emergence, reduce weed biomass by competing with them, reduce weed seed production, or reduce soil weed seedbanks. The effects could also be a combination of these processes. Although there are more reports of successful weed suppression by cover crops, few studies have reported no effect of cover crops on weeds.

For example, in a study in an orchard in Turkey, it was reported that living cover crops suppressed weed biomass whereas, mowed and incorporated cover crops reduced weed density [92]. There are several reports of correlation of cover crop mulches with a decrease in weed emergence; however, the species used as mulch influenced the rate of weed emergence [79,93]. A field study that was conducted to assess the effect of residues of rye, crimson clover (Trifolium incarnatum), hairy vetch (Vicia vollosa), and barley alone and as mixture of all four observed that they reduced the emergence of eastern black nightshade (Solanum ptycanthum), while the emergence of yellow foxtail (Setaria glauca) was reduced only by rye and barley; hence, suggesting that suppression of emergence not only depended on the cover crop species but also the weed species [90]. Another study compared weed seedling emergence between rye, wheat, and clover residues and observed that while the grain crops suppressed, the clovers stimulated weed seedling emergence [94]. This finding can be explained by a conclusion from a study that cover crop species that contribute to soil nitrogen, such as legumes, may actually stimulate weed seed germination and growth [95]. It has been stated that, during its growth, cover crops reduce both light quantity and quality (red to far red ratio) which in turn will reduce weed seed germination [96]. Therefore, the architecture of the cover crops and the changes it brings about in light quality and quantity may be a factor affecting weed seedling emergence in the case of actively growing cover crops, but there are very few reports of effects of mulch on light quantity and quality, and thereby influence on weed seedling emergence.

Reductions in weed seedbank sizes are also reported as a weed suppressive effect of cover crops. For example, a study in Italy reported that hairy vetch cover crops reduced weed seedling density, while brown mustard (Brassica juncea) showed no effect; the variation in suppressive effects between the cover crop species was not explained by differences in cover crop biomass [97]. A study in Iowa, US reported that winter rye cover crops decreased weed seedbank densities in a maize-soybean farming system.

Moreover, reports exist of either actively growing cover crops or mulches having no effect on weed suppression, weed seedling emergence [26,100,101], or decreases in the weed seedbank [102]. The study [100] used rye as a cover crop in a continental climate of Ontario Canada, characterized by hot humid summers and very cold winters and reported that there was no effect on weed density or species composition. Reddy [101] studied the effect of several crops in a humid environment in Stoneville, MS and reported no suppression of barnyard grass (Echinochloa crus-galli), prickly sida (Sida pinosa), and yellow nutsedge (Cyperus esculentus), but some suppression of browntop millet (Brachiaria ramosa) densities by Italian ryegrass (Lolium multiforum), rye, wheat, hairy vetch, crimson clover (Trifolium incarnatum), or subterranean clover cover crops. Baumgartner et al. [26] observed no significant effects of both perennial and annual cover crops on weed populations in a vineyard study that took place in the dry Mediterranean climate with dry summers and mild, wet winters. Other studies also found similar results. For example, in a study in Ontario, Canada, no effect of rye, triticale, and wheat mulches was observed on the emergence patterns of redroot pigweed (Amaranthus retroflexus) and common lambsquarters (Chenopodium album) [103]. In Oregon, US a study comparing tillage systems with cover crops either lying on the surface or incorporated concluded that tillage type was more important than cover crop mulches in regulating weed seed emergence [104]. Another study in Japan concluded that it was more important to have higher ground coverage at the early stage of the cover crop than using a higher seeding rate of the cover crop for weed suppression [105].