2026 Volume 14 Issue 2 Pages 30-49
Post-dispersal seed predation by invertebrates may contribute to weed suppression as part of integrated weed management strategies. Invertebrate seed predators can reduce weed seed abundance through direct seed removal, thereby diminishing soil seed banks and ultimately lowering subsequent weed emergence. Although post-dispersal seed predation has been extensively documented in agricultural landscapes—particularly in annual cropping systems—its relevance to perennial cropping systems, such as tea, has not been comprehensively investigated. This review synthesizes findings from annual cropping systems to identify key invertebrate seed predators and the weed species they consume, and highlights knowledge gaps relevant to perennial cropping systems. Focusing on tea agroecosystems, the review summarizes current evidence on potential seed predators occurring in tea fields and discusses how post-dispersal seed predation may function as an alternative approach to biological weed control. By integrating insights from annual cropping systems with emerging evidence from tea fields, this review provides a conceptual foundation and a research framework for evaluating the role of invertebrate-mediated seed predation in perennial agroecosystems.
In agricultural fields, weeds are commonly defined as undesirable plants that grow in inappropriate places and/or at inappropriate times during crop production. Abundant weed populations can substantially limit crop yield and quality [1].
Chemical herbicides are widely used for weed control in numerous agricultural systems. However, excessive reliance on herbicides has led to the rapid evolution of herbicide resistance. So far, 273 weed species with confirmed herbicide resistance have been documented in 102 crops across 75 countries [2]. In addition to the emergence of resistant biotypes, long-term herbicide use contributes to the accumulation of chemical residues in crop products, soils, and surrounding environments, raising concerns about potential toxicity to humans and animals [3, 4, 5]. These issues highlight the need for more environmentally sustainable weed management approaches that minimize negative impacts on both crops and agroecosystems. Within this context, biological weed control represents a promising alternative strategy.
Biological weed control relies on living organisms, such as insects, nematodes, bacteria, and fungi, to suppress weed populations. Weed seed banks determine future weed populations and weed competition, making them important targets for weed management [6]. Predation by seed predators can directly affect weed seed mortality. Seed predation includes pre- and post-dispersal predation. Pre-dispersal seed predation occurs before the seeds are shed by the parent plants, while post-dispersal seed predation is the consumption of seeds after their release from parent plants, typically on the soil surface [7]. Post-dispersal seed predation has attracted attention as a potential weed management approach, as it can directly reduce seed banks and, in turn, influence plant regeneration and population dynamics [8]. Van Klinken and White [9] demonstrated that free seeds located on or within the soil surface are more vulnerable to predation than seeds that remain on the parent plant. Both vertebrates and invertebrates participate in post-dispersal seed predation; generally, invertebrates preferentially consume relatively small weed seeds, whereas vertebrates tend to feed on relatively large or hard-coated seeds [10].
A substantial body of research has documented post-dispersal weed seed predation in grasslands and crop fields [11, 12, 13]. In addition, a recent review highlighted the importance of investigating weed seed predation in agricultural ecosystems [14]. In these systems, invertebrate seed predation has been shown to reduce weed emergence by depleting seed banks and increasing seed mortality [15, 16, 17]. For example, Perthame et al. [17] demonstrated that carabid beetles contribute to weed management by reducing weed infestation levels and enhancing crop yields. The diversity and abundance of invertebrate seed predators are strongly shaped by agricultural management practices [18], and crop type has also been reported to influence the level of seed predation [19, 20]. Most studies on post-dispersal seed predation as a potential biological weed control strategy have focused on annual cropping systems or in annual–perennial intercropping systems. In contrast, relatively few studies have examined this process in perennial monocultures. Annual crop systems are typically characterized by intensive field management, including tillage and harvest operations, and crop cycles that only last for a single growing season or do not extend throughout the year. Conversely, perennial crops are present year-round and generally require relatively little management, leading to more stable ecosystems with greater accumulation of organic matter, such as leaf litter. These contrasting ecological and management conditions influence the diversity and availability of seed predators and can consequently affect seed predation levels.
The first objective of this review is to synthesize current knowledge on post-dispersal seed predation, particularly the types of seed predators, the weed species they consume, and their effects, mainly in annual cropping systems. The second objective is to offer implications for future research on post-dispersal seed predation in perennial cropping systems, with a focus on tea agroecosystems, which exhibit high ecosystem stability. Tea is a globally important commodity, and effective management practices, including weed control, are essential for maintaining both yield and quality. Therefore, the aim of this review is to provide foundational information for evaluating seed predation as an alternative biological weed control approach in tea agroecosystems and to outline a framework for future research on invertebrate-mediated seed predation as a weed management strategy in this system.
Supplementary Table 1 provides an overview of post-dispersal seed predation by invertebrates in cropping systems.
Many animals are responsible for post-dispersal seed predation, with invertebrates being among the most important. These invertebrates are mostly ground-dwelling species or organisms that inhabit the soil surface and prey on seeds after dispersal. Regarding invertebrate taxa, studies from the 1970s to the 1990s mainly focused on Carabidae (ground beetles), particularly Harpalus spp. (e.g., H. pensylvanicus and H. rufipes)[21, 35], as well as Gryllidae (field crickets) [22]. In addition, studies examined other invertebrates, such as slugs from the families Arionidae (Arion subfuscus) and Agriolimacidae (Deroceras reticulatum) [36]. Since the 2000s, studies have consistently reported members of the Carabidae (e.g., Harpalus spp., Amara spp.) and Gryllidae (Gryllus spp.) as seed predators, and additional taxa—e.g., Teleogryllus emma, Modicogryllus siamensis (Gryllidae), Anisodactylus signatus, and A. punctatipennis (Carabidae)—have also been reported [32] (Supplementary Table 1). In addition, Formicidae (ants), isopods from Armadillidiidae and Porcellionidae, and annelids from Lumbricidae have been documented [12, 26, 30, 37, 38]. Interestingly, studies show that these invertebrates are not primarily granivorous; rather, many are polyphagous predators [39, 40, 41, 42, 43]. For example, larvae of the carnivorous species Amara similata, which feed on small insects, were first reported to consume Capsella bursa-pastoris seeds in 1997 [40]. These larvae exhibited higher fecundity and survival rates when fed weed seeds than when fed insect prey [40].
The weed species consumed by invertebrate seed predators vary considerably among studies. Nevertheless, Supplementary Table 1 indicates that numerous investigations have examined post-dispersal seed predation on common weeds in annual crop fields, such as Amaranthus retroflexus (redroot pigweed), Chenopodium album (lambsquarters), Abutilon theophrasti (velvetleaf), Setaria faberi (giant foxtail), S. glauca (yellow foxtail), and Echinochloa crus-galli (barnyardgrass), which are widespread weeds in maize and soybean fields in several countries, including Poland and the United States [44, 45, 46]. In addition, Lolium multiflorum (Italian ryegrass), a major weed in wheat fields in Japan, is preyed upon by invertebrate seed predators [32].
Considering the types of weed seeds consumed by seed predators across crop fields listed in Supplementary Table 1, Fig. 1 illustrates the relationship between seed predator genera and the weed seed families they prey upon. This figure was constructed from the information summarized in Supplementary Table 1. Among them, three genera, Harpalus, Amara, and Armadillidium exhibit a particularly broad prey spectrum. Harpalus spp. are widely documented as predators of weed seeds from the Poaceae, Amaranthaceae, and Violaceae families. Amara spp. are reported to prey on seeds from the Poaceae, Malvaceae, Amaranthaceae, Caryophyllaceae, and Violaceae families. In addition, members of the genus Armadillidium have been reported to consume seeds from Poaceae, Amaranthaceae, Asteraceae, Caryophyllaceae, and Plantaginaceae families.
Post-dispersal seed predation results in seed removal and seed loss [31]. Quantitative studies further demonstrate the importance of seed predation in weed population dynamics. Invertebrate predation (by ants, beetles, and crickets) removed 25–57% of Veronica persica, Taraxacum officinale, Stellaria media, and Cuscuta campestris (field dodder) seeds within two weeks [34]. Westerman et al. [47] estimated that annual seed losses due to predation in cereal fields ranged from 32% to 70% at two-week intervals. In a three-year experiment, seed predation by invertebrates on weeds in maize and soybean fields varied from 6.1% to 27.2% at three- to four-day intervals [48]. Modeling analyses that include estimation of velvetleaf (Abutilon theophrasti) seed predation by Westerman et al. [49] show that weed seed predation after dispersal can contribute to weed suppression by increasing seed losses in a four-year crop rotation (maize–soybean–triticale + alfalfa–alfalfa) than in a two-year crop rotation (maize–soybean).
2.3 Important factors influencing post-dispersal seed predation 2.3.1 Biotic (1) Crop typeCrop type influences seasonal patterns of seed predation [19, 20]. For instance, in maize and soybean fields, predation on Abutilon theophrasti and Setaria faberi is low in autumn and spring but relatively high in summer [20]. In triticale intercropped with legumes, seed predation increases in autumn, likely owing to the formation of a denser crop canopy. O’Rourke et al. [19] further reported that invertebrate predation on S. faberi is higher in maize–soybean rotations than in triticale–alfalfa or alfalfa monocultures. In addition, predation rates declined in soybean fields by late September but remained high in maize fields. Collectively, these findings suggest that post-dispersal seed predation follows crop-specific and season-specific patterns, reflecting differences in crop structure and phenology.
(2) Vegetation coverVegetation cover and crops also influence seed dispersal and availability. Inagaki et al. [50] reported that carabid beetle densities were higher under dense vegetation than in bare soils, and the abundance of H. pensylvanicus is particularly high in grassy vegetation [51]. Vegetation structure may also affect seed predation [25]. However, some species prefer open spaces to densely vegetated areas. For example, Cyclotrachelus sodalis requires sunlight to regulate its body temperature [51]. Ants also prefer to feed in open spaces rather than areas with dense vegetation [52].
(3) Natural enemies of seed predatorsThe presence of natural enemies of seed predators is another factor that may influence the level of seed predation. Natural enemies are commonly classified as either specialists or generalists [53]. Specialist natural enemies may reduce the abundance or activity of seed predators and thereby potentially weaken weed seed suppression by these predators. In contrast, generalist natural enemies do not rely solely on weed seed predators as a food source [54]. One relevant mechanism is intraguild predation, in which natural enemies and weed seed predators exploit shared food resources, potentially altering the overall intensity of weed seed predation. In some cases, the establishment of specialist natural enemies has been reported to result in higher prey populations compared with systems dominated by generalist enemies alone. Interactions between generalist or specialist natural enemies and their prey, including weed seed predators, may therefore be positive, negative, or neutral, depending on ecological context [53, 54]. Davis and Raghu [55] reported that the activity density of crickets and granivorous carabids was strongly and weakly positively correlated with invertebrate seed predation, respectively. In contrast, the activity density of spiders, classified as generalist natural enemies, was strongly negatively correlated with invertebrate seed predation in maize fields in Illinois, USA. Despite these potentially complex effects, the presence of natural enemies is generally considered essential for maintaining ecosystem balance and biodiversity.
2.3.2 Abiotic (1) Cultivation managementField management practices, such as tillage, can bury seeds and reduce their availability [49, 56]. Chemical applications during the cultivation management may also affect predation levels [49]. No tillage application in a 4-year crop rotation (maize–soybean–triticale + alfalfa–alfalfa) with reduced herbicide use resulted in higher seed loss due to predation (32% vs. 17% per two days), compared with a two-year rotation (maize–soybean) with regular herbicide use. In no-tillage systems, weed seed production is higher than in tillage systems, and seeds are therefore more likely to be exposed to predators, resulting in a higher seed loss to predation [49].
In addition to affecting the seed availability, the factors mentioned above also affect the seed predators in the fields. For example, beetles show higher activity in strip-tillage systems compared with full tillage [48], and other arthropods, such as sow bugs and millipedes, are more abundant in no-tilled than in tilled fields [57]. With respect to seed predation, however, Cromar et al. [57] reported that predation rates peaked in no-till and moldboard plow systems, whereas chisel-plowed fields exhibited the lowest rate of predation. These results suggest that management practices do not directly affect seed predation, but instead interact with additional factors, such as the availability of alternative food resources within tillage systems, as well as foraging behavior and food preferences of predators [36, 57].
(2) SeasonPost-dispersal seed predation by invertebrates has been shown to vary temporally and seasonally and to be associated with crop type and management practices. Seed predation was higher in maize–soybean rotations than in triticale–alfalfa rotations or alfalfa monocultures; moreover, predation rates only declined in soybean fields in late September but remained high in maize fields [19]. This may be because the season affects the availability of seed predators. According to Briese and Macauley [58], foraging ants are higher in spring than in summer, where spring temperature is relatively favorable for them. Furthermore, Daouti et al. [8] stated that optimal predation occurs when seed predator activity overlaps with weed seed availability. Amara spp., Harpalus spp., and Pseudoophonus rufipes are active from April to October in Czech pear orchards, with peak activity in June and July [16]. Consequently, weed seeds that are dispersed and remain on the soil surface during periods when seed predators are active are more likely to be encountered and subjected to higher predation pressure.
To determine the invertebrates with the potential to play roles as weed seed predators, some criteria may aid in their incorporation into this function. In this section, three criteria for classifying invertebrates as potential weed seed predators are outlined based on studies of post-dispersal seed predation in annual cropping systems.
3.1 Weed seed predatorsInvertebrates are considered potential seed predators because they consume weed seeds. Numerous studies show that invertebrates in the families Carabidae (beetles) and Gryllidae (crickets) are among the most commonly reported seed predators of weeds [14, 15, 19, 22, 24, 28, 31]. Species of Carabidae, particularly Harpalus spp. and Amara spp., and Gryllidae, such as Gryllus spp., have been reported to prey on weed seeds in laboratory and field experiments. These invertebrates, particularly Harpalus and Amara, function as generalists, feeding on seeds from multiple weed families (Fig. 1).
A feeding experiment conducted in laboratory chambers by Deroulers and Bretagnolle [43] showed that more than half of the 28 beetle species fed on seeds of Viola arvensis (European field pansy), including Harpalus affinis, H. tardus, H. distinguendus, A. similata, A. apricaria, Pseudoophonus rufipes, and P. calcaetus. Male and female Gryllus pensylvanicus feed on seeds of Setaria faberi, Digitaria ciliaris (southern crabgrass), Abutilon theophrasti, and Amaranthus retroflexus, with females consuming more than males [22]. In field experiments, the carabid beetle Harpalus pensylvanicus and Amara aenea and the cricket Gryllus pennsylvanicus were reported to feed on seeds of A. retroflexus, S. faberi, and A. theophrasti [15, 19, 20, 21, 22, 24, 25, 28, 30, 31]. Anisodactylus spp. such as A. signatus and A. punctatipennis consumed seeds of Lolium multiflorum [32].
Other arthropods, such as ants (Formicidae), including Solenopsis invicta, feed on Amaranthus retroflexus, Urochloa platyphylla (broadleaf signalgrass), and Senna obtusifolia (sicklepod) [30], whereas S. geminata preys on Digitaria ciliaris, Eleusine indica (indian goosegrass), and Echinochloa colona (awnless barnyard) seeds [12]. Isopod in the family Armadillidiidae (Armadillidium vulgare), a natural detritivore, also feeds on the seeds of Capsella bursa-pastoris (shepherd’s purse), Poa annua (meadow grass), Veronica persica (common-field speedwell), and Stellaria media (Common chickweed) in laboratory seed consumption tests [26] and also confirmed to feed on the seeds of Taraxacum agg. (dandelion) in field experiments [16].
3.2 High prevalence and distributionPrevalence and distribution are associated with habitat and crop-type specificity or generality. Crickets (Gryllus spp.) and beetles (Harpalus and Amara spp.) have been found to be present in more than one type of crop, which suggests that they have a high prevalence and distribution. However, at the species level, these criteria may vary among countries. Field crickets Gryllus pennsylvanicus and Allonemobius allardi, and the beetle Harpalus pensylvanicus are abundant in fields with maize, soybean, and triticale rotations, with G. pennsylvanicus having a higher activity density, and are recorded as abundant in maize and soybean fields in Iowa, USA [19]. The activity density of species of Harpalus (e.g., H. affinis, H. tardus, H. distinguendus), Amara (e.g., A. aenea and A. similata), Pterostichus melanarius, and Pseudoophonus rufipes has been recorded in winter wheat fields in Wageningen, the Netherlands [59]. Pterostichus melanarius is recorded as abundant in arable crops such as organic spring wheat, barley, beans, and conventional winter wheat, barley, and oilseed rape in Northumberland, the UK, with some other species from Amara spp. such as Amara eurynota and A. similata, and Harpalus rufipes reported to be lower in number [60]. The Isopod Armadillidium vulgare (family Armadillidiidae) is abundant in organic olive groves but less abundant in maize fields, whereas carabid beetles such as H. distinguendus are abundant in vineyards and organic maize fields in Greece [61].
3.3 High stabilityDynamic environments, such as annual cropping systems with crop rotation, involve management practices such as tillage and chemical inputs, including fertilizers and herbicides, which may change invertebrate habitats, food resources, and suitable microclimates, ultimately affecting their abundance [61]. Stable environments, exemplified by perennial cropping systems, apply fewer field disturbances, including no crop rotation, and generally necessitate less frequent tillage application, which minimizes changes in invertebrate habitat.
Species of Harpalus (e.g., H. pensylvanicus and H. rufipes), and crickets (Gryllidae) tend to maintain their populations even in dynamic environments and are commonly recorded as abundant in such systems. The activity density of Pterostichus melanarius has been reported to be higher in conventional (using inorganic fertilizers and herbicides) and organic (compost and slurry fertilizers with no herbicides) wheat and barley fields in the UK, with the conventional field reporting slightly higher number [60]. O’Rourke et al. [19] showed that the abundance of beetles H. pensylvanicus and crickets Gryllus pennsylvanicus varied in different crop type rotations; more abundant during triticale-alfalfa cropping in a four-year rotation of maize–soybean–triticale + alfalfa–alfalfa that applies higher management input, such as fertilizers and herbicide use, than in a two-year rotation of maize–soybean that applies lower management input. In contrast, G. pennsylvanicus exhibited higher activity-density in maize–soybean rotations, and lower activity density in triticale–alfalfa rotations. Fox et al. [30] reported that the activity density of H. pensylvanicus in hay fields is higher than in maize fields. Isopod, including Armadillidiidae such as Armadillidium vulgare, is found in olive groves, both conventional and organic, with relatively high weed flora. This may be attributed to the availability of more suitable microclimate habitats owing to the presence of olive trees with a relatively high weed flora [61]. However, this species may be less stable in dynamic environments, because it is particularly sensitive to soil disturbance, with lower abundance in conventional tillage than in no-tillage fields [62].
In winter wheat fields, species such as Amara aenea, A. familiaris, A. anthobia, and Pseudoophonus rufipes are captured as field-edge species, where more varied grasses surround the field edge, especially in fields with narrow weedy strips along the boundary. The weeds may serve as their habitat, and some may serve as their food [59]. This suggests that the populations of these invertebrates may be stable even in the absence of crops and that they use the weedy strips as an alternative habitat.
Based on the criteria discussed in this section, Harpalus spp., Amara spp., and members of the family Gryllidae are frequently observed in annual crop fields, such as maize, soybean, wheat, hay, and oilseed rape. These invertebrates can sustain their populations in dynamic environments. Invertebrates from these groups may serve as potential seed predators in both annual and perennial crop systems. Although members of the family Armadillidiidae, including A. vulgare, are prevalent in agricultural land; they are easily affected by soil disturbances, particularly tillage practices. Consequently, their role as seed predators may be more prevalent for perennial crops.
Seeds are dispersed and incorporated in the soil either naturally or with the assistance of wind or animals [63, 64]. Seeds that successfully reach the soil surface contribute to the formation of the seed bank, which consists of both surface and buried seeds. Burial occurs through cultivation practices such as tillage [57, 65], as well as through animal activities that often bury seeds and thereby contribute to seed bank development [66].
4.2 Seed exposure, detection, encounter, and exploitationThe location of seed dispersal strongly influences seed exposure, detection, encounter, and exploitation. Seeds remaining on the soil surface are highly exposed and thus more likely to be detected by predators [29, 67]. For instance, predation by invertebrates such as Stator vachelliae (Bruchidae) is significantly lower at greater distances from parent plants, indicating that predators typically forage closer to the canopy [68].
Seed density also plays a key role: high-density seed patches are more readily detected than sparse ones [67, 69]. In contrast, burial reduces the probability of encounters [70]. Moreover, seed size and density interact to influence predation risk, with larger or more numerous seeds being easier to locate [70, 71]. Seed encounter and subsequent exploitation may lead to seed removal, a process that is also density-dependent [71].
Predator behavior further shapes these interactions. Certain invertebrates transport and cache seeds [68]. For example, ants collect seeds and carry them to nests or storage sites [72], whereas Harpalus larvae have been observed to collect and bury seeds at depths of 8–20 cm [73].
4.3 Seed predationOnce encountered, predators selectively consume preferred seeds (Fig. 2). Predator feeding preferences are largely determined by seed morphological and physiological traits. Small seeds are preferred by Armadillidium vulgare, and this preference increases with predator body size [26]. Hard seed coats (e.g., Geranium pratense) are unsuitable for arthropods but may be consumed by slugs that can swallow them intact [74].
Some invertebrates transport seeds before consumption, occasionally leaving them uneaten. Partial predation is also common: for instance, Gryllus pennsylvanicus consumes the endosperm of Setaria faberi and Digitaria sanguinalis, leaving the pericarp intact [23]. Such partial damage reduces germination vigor compared to intact seeds. Post-dispersal seed predation varies across space and time, influenced by various factors [52, 71, 75]. Local environmental conditions, including soil type, canopy cover, and predator communities, create variability in predation intensity. Therefore, a nuanced understanding of these factors is essential for integrating seed predation into weed management strategies.
4.4 Seed survival and weed establishmentNot every seed that predators detect is destroyed. Some non-preferred seeds survive (Fig. 2). The presence of buried seeds, whether owing to predator activity or cultivation, also increases the probability of survival. For example, White et al. [15] found that buried seeds experienced lower carabid predation and a higher chance of survival than surface seeds. Likewise, Traveset [68] demonstrated that seeds dispersed and later defecated by frugivores had greater survival against bruchid beetles than fully exposed seeds.
Surviving seeds may germinate and establish seedlings unless dormancy is maintained. However, they still face intra- and interspecific competition [76]. Larger seeds can physically suppress smaller ones during germination [77], and microhabitats, such as canopy shading, also affect seedling growth and development [78].
These interactions influence plant community structure. Janzen and Connell's theory of negative density dependence posits that higher conspecific seed densities reduce individual survival due to intensified predation, thereby limiting dominance [64, 74, 79, 80]. Therefore, under ecological conditions where (i) seed density is high, thereby increasing density-dependent mortality, (ii) predator abundance and diversity are sufficient to exert measurable pressure on seed banks, (iii) environmental conditions such as soil exposure or favorable climate promote predator activity, (iv) seed traits (e.g., size, hardness) influence susceptibility to predation, and (v) agricultural management practices, such as tillage or weed control are present, seed predation may function as a weed population regulation and shape community structure.
Understanding arthropod communities in tea ecosystems provides valuable insights into their diversity and ecological roles and may help clarify the potential seed predators in these systems. Accordingly, this subsection discusses arthropod communities reported from tea fields across several countries.
Tea fields in China harbor diverse arthropod communities, including herbivorous insects, predatory insects, and neutral arthropods [81, 82]. According to Ye et al. [81], 390 species of predatory insects have been found in Chinese tea fields, belonging to 45 families and 10 orders. Some of these insects are reported to be natural enemies of mites, whiteflies, and aphids. Among them, Chilocorus kuwanae Silvestri, Coccinella septempunctata L., Leis (Harmonia) axyridis (Pallas), and Propylaea japonica (Thunberg) (family Coccinellidae), as well as Parena rufotestacea (family Carabidae), are predatory insects.
In Iranian tea fields, 17 species, 16 genera, and 7 families of predatory and parasitoid insects have been recorded. These include Brachinus crepitans L., Harpalus affinis (Schrank), H. griseus (Panzer), Poecilius lepidus (Leske), and Pterostichus niger (Schaller) from the family Carabidae. Additionally, C. septempunctata L., Cryptolaemus montrouzieri Mulsant, and Adalia bipunctata L. (family Coccinellidae), as well as Paragus tibialis (family Syrphidae), are present [83].
In Japanese tea fields, diverse ground-dwelling invertebrates such as beetles, spiders, earwigs, house centipedes, crickets, millipedes, and woodlice have been reported. Among them, Amara sp., members of the subfamily Harpalinae (beetles), members of Gryllidae (crickets), Diplopoda (millipedes), and Armadillidium vulgare (woodlice) are recognized as seed predators [84]. In addition, Toyoshima et al. [85] reported several carabid beetle species from Japanese tea fields, including Synuchus arcuaticollis, S. dulcigradus, S. nitidus, Anisodactylus signatus, A. punctatipennis, A. tricuspidatus, Harpalus jureceki, H. calcaetus, and H. tinticulus. Similarly, arthropods observed in Indonesian tea fields include beetles (Carabidae), crickets (Gryllidae), ants (Formicidae), leafhoppers (Cicadellidae), and aphids (Aphididae) [86, 87].
5.1.2 Weed communitiesIn addition to examining arthropods in tea fields, it is also crucial to investigate the weed flora. Identifying which weeds are present helps clarify weed diversity in tea fields and indicates which species may be targeted by seed predators. In addition, understanding the weed flora in tea fields provides information for the community structure within the system aside from the crop itself and may provide alternative habitats for invertebrates. The common weeds documented in tea fields are summarized in Table 2.
| Family | Species | Weed type | Country | References |
|---|---|---|---|---|
| Asteraceae | Ageratum conyzoides | Br | India | [90] |
| Thailand | [107] | |||
| Indonesia, China | [89] | |||
| Ageratum houstonianum | Br | India | [90] | |
| Bidens pilosa | Br | India | [90] | |
| Thailand | [107] | |||
| Indonesia, Japan, China | [89] | |||
| Erigeron canadensis | Br | India | [90] | |
| Erigeron annuus | Br | Turkey | [91] | |
| Erechtites hieraciifolia | Br | Japan | [89, 95, 97] | |
| Crassocephalum crepidioides | Br | Thailand | [107] | |
| Japan | [95] | |||
| Sri Lanka | [92] | |||
| Ageratina adenophora | Br | Thailand | [107] | |
| Chromolaena odorata | Br | India | [93] | |
| Galinsoga parviflora | Br | Ethiopia | [94] | |
| Conyza albida | Br | Turkey | [91] | |
| Ethiopia | [94] | |||
| Ambrosia artemiisifolia | Br | Turkey | [91] | |
| Artemisia vulgaris | Br | Japan, China, Georgia | [89] | |
| Poaceae | Digitaria ciliaris** | Gr | Thailand | [107] |
| Japan | [95] | |||
| Digitaria sanguinalis** | Gr | Turkey | [91] | |
| Sri Lanka | [92] | |||
| Georgia | [89] | |||
| Digitaria adscendens* | Gr |
Indonesia, Japan, China |
[89] | |
| Eleusine indica** | Gr | Thailand | [107] | |
| India | [93] | |||
| Japan | [95, 97] | |||
| Cynodon dactylon | Gr | India | [88, 90, 93,] | |
| Turkey | [91] | |||
| Japan, China, Indonesia, Sri Lanka, East Africa, Georgia | [89] | |||
| Poa annua** | Gr | Iran | [108] | |
| Japan | [89, 104] | |||
| Imperata cylindrica | Gr | India | [90] | |
| Thailand | [107] | |||
| Sri Lanka, Indonesia, China, Japan, East Africa | [89] | |||
| Axonopus compressus | Gr | India | [88] | |
| Setaria glauca* | Gr | India | [88] | |
| Turkey | [91] | |||
| Setaria viridis | Gr | Iran | [108] | |
| Lolium perenne** | Gr | Turkey | [91] | |
| Lolium multiflorum** | Gr | Japan | [97, 104] | |
| Echinochloa crus-galli* | Gr | Turkey | [91] | |
| Ethiopia | [94] | |||
| Paspalum spp. | Gr | Thailand | [107] | |
| India | [88] | |||
| Turkey | [91] | |||
| Sri Lanka | [92] | |||
| Japan, Georgia | [89] | |||
| Cyperaceae | Cyperus spp. | S | Thailand | [107] |
| Ethiopia | [94] | |||
| Turkey | [91] | |||
| Sri Lanka | [92] | |||
| Georgia | [89] | |||
| Kyllinga brevifolia | S | Thailand | [107] | |
| Kyllinga bulbosa | S | Ethiopia | [94] | |
| Amaranthaceae | Amaranthus retroflexus** | Br | Turkey | [91] |
| Iran | [108] | |||
| Amaranthus hybridus* | Br | Ethiopia | [94] | |
| Amaranthus dubius* | Br | Ethiopia | [94] | |
| Chenopodium album** | Br | Turkey | [91] | |
| Plantaginaceae | Plantago lanceolata | Br | Ethiopia | [94] |
| Plantago major | Br | Turkey | [91] | |
| Veronica persica** | Br | Turkey | [91] | |
| Iran | [108] | |||
| Brassicaceae | Capsella bursa-pastoris** | Br | Turkey | [91] |
| Brassica oleracea | Br | |||
| Fabaceae | Trifolium repense* | Br | Turkey | [91] |
| Trifolium pratense** | Br | |||
| Mimosa pudica | Br | Sri Lanka | [92] | |
| Oxalidaceae | Oxalis corniculata | Br | India | [93] |
| Turkey | [91] | |||
| Japan | [104] | |||
| Thailand | [107] | |||
| Oxalis barrelieri | Br | Sri Lanka | [92] | |
| Polygonaceae | Polygonum spp.* | Br | India | [90] |
| Turkey | [91] | |||
| Iran | [108] | |||
| Japan | [89] | |||
| Caryophyllaceae | Stellaria media** | Br | Turkey | [91] |
| Japan | [89, 104] | |||
| Stellaria aquatica* | Br | Japan | [89] | |
| Drymaria cordata | Br | Thailand | [107] | |
| Lamiaceae | Prunella vulgaris | Br | Turkey | [91] |
| Iran | [108] | |||
| Verbenaceae | Lantana camara | Br | India | [88, 93, 109] |
Abbreviations: Br, broadleaf; Gr, grass; S, sedge.
Note: Double asterisks (**) indicate the same weed species. A single asterisk (*) indicates weed species belonging to the same genus as those listed in Supplementary Table 1 whose seeds were reported to be preyed upon by invertebrates.
As shown in Table 2, tea fields are frequently dominated by weeds from the Asteraceae, Poaceae, and Amaranthaceae families, although species from the Cyperaceae, Plantaginaceae, Brassicaceae, Fabaceae, Oxalidaceae, Polygonaceae, Caryophyllaceae, Lamiaceae, and Verbenaceae families are also reported. These weeds include broadleaves, grasses, and sedges. Weed species in tea fields vary across countries; however, several problematic species are consistently observed across regions. For example, Ageratum conyzoides, Bidens pilosa, Artemisia vulgaris, Digitaria sanguinalis, Cynodon dactylon, Cyperus rotundus, Oxalis corniculata, Polygonum spp. (mostly P. longisetum), and Imperata cylindrica are recognized as common weeds in young tea fields in countries such as Indonesia, Sri Lanka, India, Turkey, China, Africa, and Georgia [88, 89, 90, 91, 92, 93, 94]. Young tea fields are particularly vulnerable to growth competition from these weeds. In contrast, mature tea fields often harbor shade-tolerant species such as Drymaria cordata and Oxalis latifolia beneath closed canopies [89]. Regional differences also exist. For instance, in many areas of Shizuoka Prefecture in Japan, Digitaria ciliaris dominates field margins, but in some areas, Eleusine indica, which is suspected to be resistant to glyphosate, is dominant [95], whereas Crassocephalum crepidioides is also commonly found between the furrows of some fields [96].
5.1.3 Post-dispersal seed predation and potential seed predators in tea agroecosystemsResearch on post-dispersal seed predation in tea fields remains limited; however, interest in this topic is increasing. Ichihara et al. [97] investigated post-dispersal weed seed predation in conventional and organic tea fields in Japan, focusing on prevalent weed species in Japanese tea fields such as L. multiflorum, D. ciliaris, E. indica, Bidens pilosa var. pilosa, and Erechtites hieracifolia. The study reported average biweekly seed predation rates ranging from 42.8–80.5% in organic fields and 21.7–59.6% in conventional fields during the summer season. These results indicate that post-dispersal seed predation can substantially reduce the weed seed bank in tea agroecosystems. Seed predation is known to be density-dependent, and temporal overlap between peak seed shedding and high predation activity can enhance weed suppression. For example, synchronization between seed availability and predation has been shown to contribute to population suppression of Alopecurus myosuroides in agroecosystems [56]. In young tea fields in Sri Lanka, weed seed density within the soil seed bank (0–15 cm depth) has been reported to range from 4,168 to 5,520 seeds m-2 [98]. That study further indicated that seed density did not differ substantially between pre-planting and one year after planting; however, weeding intervals significantly influenced seed density. Such variation may result from differences in weed management practices that create temporarily weed-free conditions.
Regarding seed predators, crickets of the genera Velarifictorus and Loxoblemmus, ground beetles of the genera Harpalus, Amara, and Anisodactylus, as well as pill bugs (Armadillidium vulgare), are recorded in Japanese tea fields during summer, with crickets and A. vulgare exhibiting higher activity densities among them [97]. In addition to those invertebrate taxa, species belonging to the family Formicidae may also function as potential seed predators in tea fields, based on the criteria of potential seed predators discussed in Chapters 3 and 5.1.1 and observations from other agricultural systems (Supplementary Table 1, Fig. 1, Chapter 3.1). Crickets, ground beetles, and pill bugs have also been reported to have high prevalence in other Japanese tea fields [84, 85], and, in the case of Harpalus, also in Iranian tea fields [83]. Although some species in the genera Harpalus and Anisodactylus are recognized as seed predators capable of sustaining their populations even in dynamic environments, their occurrence is expected to be higher in more stable ecosystems such as tea fields.
Compared with annual cropping systems, tea agroecosystems provide relatively stable habitats that can support high levels of biodiversity [99]. The persistent presence of tea plants without crop rotation [100], together with minimal tillage practices [101], creates favorable conditions for a wide range of associated organisms. Reduced tillage practices in tea production systems contribute to the conservation of both species abundance and richness. These management characteristics may also influence seed availability following dispersal and within surface soil seed banks. In tea fields, the absence of crop rotation and the application of minimal tillage reduce the likelihood of seed burial, potentially increasing the number of seeds accessible to predators and, consequently, enhancing post-dispersal seed predation rates. However, stable ecosystems may also support natural enemies of weed seed predators, as well as other small animals. Such organisms can serve as alternative food resources for natural enemies, potentially diluting predation pressure on seed predators and altering trophic interactions within the system.
A comparison of major weed species and predator taxa in tea fields, including those reported as seed predators, is summarized in Table 3 and provides basic information that may facilitate future investigations of post-dispersal seed predation in tea agroecosystems. Ecological conditions that differ from those in annual cropping systems may affect the potential for seed predation due to greater complexity and diversity. Overall, while this discussion highlights potential parallels between annual cropping systems and perennial tea cropping systems, they remain hypotheses that require empirical validation and further investigation. Therefore, the discussion here should be regarded as providing implications, aiming to guide future investigations rather than to draw definitive conclusions.
| a Weed species | References | b Predator taxa | References |
|---|---|---|---|
| Ageratum conyzoides | [89, 90, 107] | Gryllidae* | [84, 97] |
| Bidens pilosa | [89, 90, 97, 107] | Diplododa* | [84] |
| Crassocephalum crepidioides | [92, 95, 97, 107] | Coccinellidae | [81, 83] |
| Erechtites hieraciifolia | [97] | Syrphidae | [83] |
| Digitaria ciliaris | [95, 97, 107] | Chilopoda | [83, 84] |
| Digitaria sanguinalis | [89, 91, 92] | Formicidae* | [86, 87] |
| Eleuisine indica | [93, 95, 107] | Chrysopidae | [83] |
| Lolium multiflorum | [97] | Anisolabididae | [84] |
| Oplismenus undulatifolius var. undulatifolius f. japonicus | [97] | Harpalus* | [97] |
| Cynodon dactylon | [88, 90, 93] | Amara* | [97] |
| Imperata cylindrica | [89, 90, 107] | Anisodactylus* | [85, 97] |
| Poa annua | [89, 104, 108] | Brachinus | [83] |
| Amaranthus retroflexus | [91, 108] | Pterostichus* | [83] |
| Achyranthes bidentata var. fauriei | [110] | Armadillidium* | [84, 97] |
| Cyperus spp. | [89, 91, 92] | ||
| Polygonum spp. | [89, 90, 91, 108] |
Note: This table summarizes ecological surveys of dominant weed species (a) and seed predators (b) in tea fields across various countries. It does not imply predation by predator (b) on weed (a) in the same row. An asterisk (*) indicates that species within the related predator taxa have been reported as seed predators.
Based on previous studies on post-dispersal seed predation in annual cropping systems and on potential seed predators identified in Chapters 3 and 5.1.3 across both annual and perennial cropping systems, including tea agroecosystems, this subsection outlines experimental approaches and future research directions for investigating post-dispersal seed predation in tea fields.
As indicated by Ichihara et al. [97], the natural presence of seed predators—particularly ground-dwelling invertebrates—plays an important role in post-dispersal seed predation in tea fields. Field-based surveys of post-dispersal seed predation can be conducted by combining assessments of invertebrate activity density with seed predation measurements. Activity density can be assessed using sticky traps to capture ground-dwelling insects such as crickets [102], as well as ground beetles and pill bugs [97], and by using pitfall traps [16, 36]. Although pitfall trap studies have been primarily conducted in annual cropping systems, this method is also applicable to tea fields for sampling ground-dwelling organisms [96].
Seed predation can be quantified through seed-removal experiments using seed cards [47, 11, 103]. Seed cards are placed on the soil surface and covered with mesh that allows access by target invertebrates while excluding non-targeted organisms. The number of remaining seeds is then used to estimate the seed predation levels [103]. Considering that weeds commonly occur within or between tea plant rows and along the periphery of tea fields [95, 104], traps and seed cards can be deployed in these locations. Such spatial deployment may help clarify how vegetation structure (tea plants) and food availability (weed seeds) influence invertebrate abundance and seed predation activity.
Because post-dispersal seed predation depends on the availability of seed predators, it is essential to understand the ecological factors—such as field characteristics and management practices—that influence the occurrence and activity of invertebrate seed predators. Tea fields generally involve relatively low soil disturbance, and reduced intensive chemical use, such as insecticides and herbicides may represent a potential strategy for conserving seed predator communities.
Seed predator communities vary considerably across locations and regions. Future research should therefore focus on how region-specific seed predators can be conserved under different field management systems, and on evaluating their contribution to ecosystem functioning, particularly through post-dispersal seed predation, as one of the biological approaches to weed management. In addition, the information compiled in this review on weed families commonly consumed by seed predators in crop fields, including tea fields, may serve as a useful reference for identifying weed taxa that are likely to be preferred by predators and therefore potentially subject to biological control. Together, these steps can contribute to a better understanding of weed–seed predator interactions in agricultural systems and provide a basic framework for evaluating post-dispersal seed predation in tea fields.
Long-term empirical studies are ultimately required to assess the consistency and sustainability of post-dispersal seed predation as a biological weed control mechanism. Such studies should address the extent to which seed predation can contribute to reducing weed pressure and mitigating yield losses in tea fields. By synthesizing existing evidence and proposing a conceptual and methodological framework, this review provides a foundation for future research on invertebrate-mediated seed predation in tea agroecosystems.
Niken Nabilaputri Pranaasri: Conceptualization, Methodology, Investigation, Data curation, Writing – original draft, Visualization. Minoru Ichihara: Conceptualization, Writing – review & editing, Supervision. Masayuki Yamashita: Conceptualization, Project administration, Writing – review & editing, Supervision. Hitoshi Sawada: Writing – review & editing, Supervision.
