Yield loss and competition

The first issue is the per cent yield loss in relation to the duration of the competition. Even though crop–weed competition commences with crop emergence and continues to maturity (Thakral et al., Reference Thakral, Pandita, Khurana and Kalloo1989), CPWC considers only one-third of the duration in the life cycle of the crop and is based on a 5% acceptable yield loss level (Bukun, Reference Bukun2004) only. However, this is not universal. Yield loss from weeds is a function of crop species, growing environment, soil weed seed banks, etc. For example, a 10% yield loss due to weeds was suggested for aerobic rice (Anwar et al., Reference Anwar, Juraimi, Samedani, Puteh and Man2012) and spring canola (Martin et al., Reference Martin, Vanacker and Friesen2001) since a 5% yield loss level would not be practical from an economic viewpoint. The onset of crop–weed competition in wet, direct-seeded rice (DSR) begins much earlier because of wide adaptability, quick germination and rapid growth of weeds compared to a rice crop (Elliot et al., Reference Elliot, Navarez, Estario and Moody1984; Rao, Reference Rao2011). Yield loss calculations are considered only for the current crop. Weeds that emerge late in the crop season after the CPWC are often considered less important (Bastiaans et al., Reference Bastiaans, Paolini and Baumann2008) but could add to the soil weed seed bank (Reisinger et al., Reference Reisinger, Lehoczky and Komives2006) and would become a yield-limiting factor in the subsequent crops in that field. A large proportion of the weed seed bank remains on or close to the soil surface after seed rain (Mário, Reference Mário and Li2017), particularly in no-till (NT) systems (Chahal et al., Reference Chahal, Brar and Walia2003; Morris et al., Reference Morris, Miller, Orson and Froud-Williams2010). Weed seeds from weed seed banks present in the top layer of the soil germinate and seedlings grow faster than the crop, particularly in aerobic rice. The length of the CPWC in aerobic rice is expected to be longer than other systems of rice production (usually one-third of the life cycle), including conventional transplanted rice since the flooded soil environment hinders the germination of several weed species (Singh et al., Reference Singh, Singh, Dhyani, Banga, Kumar, Satyawali and Bisht2016; Raj and Syriac, Reference Raj and Syriac2017). For example, the CPWC in wet DSR was from 12 to 60 days after sowing (DAS) (Azmi et al., Reference Azmi, Abdul Shukor and Mohamad Najib2007a, Reference Azmi, Juraimi and Najib2007b), so the first 60 days are accounted for crop–weed competition (Singh, Reference Singh, Singh, Singh, Chauhan, Orr, Mortimer, Johnson and Hardy2008), while it was 20–40 days after transplanting (DAT) for transplanted rice (Mukherjee et al., Reference Mukherjee, Sarkar and Maity2008). In Sahel (West Africa), CPWC for lowland irrigated rice was 29–32 DAS in the wet season, while 4–83 DAS in the dry season (Johnson et al., Reference Johnson, Wopereis, Mbodj, Diallo, Powers and Haefele2004). Hence, CPWC is based on a 5% acceptable yield loss level, which can be 10% in some crops. It changes according to the crop production system (conventional, conservation or organic), weed types, crop types, soil types, climate and agronomic practice followed in that crop.

Comprehensive information on CWFP and CPWC for different crops is presented in Table 1. Genotypic differences [e.g. wheat (Huel and Hucl, Reference Huel and Hucl1996), maize (Saito et al., Reference Saito, Phanthaboon, Shiraiwa, Horie and Futakuchi2010) and sorghum (Wu et al., Reference Wu, Walker, Osten and Robinson2010)] do modify the CPWC due to their differential weed-competitive nature (Ramesh et al., Reference Ramesh, Matloob, Aslam, Florentine and Chauhan2017a, Reference Ramesh, Rao and Chauhan2017b; Chauhan, Reference Chauhan2020). Mahajan et al. (Reference Mahajan, Ramesha and Chauhan2014) found that rice genotypes PR-115 and H-97158 could not compete with weeds and were regarded as the worst weed competitors while PR-120, IR88633 and IR83927 were able to compete with weeds. Although Campos et al. (Reference Campos, Barroso, Silva Junior, Gonçalves and Martins2016) noticed a temporal variation in CWFP in maize between years; 54 days after emergence (DAE) in the first year v. 27 DAE in the second year. In general, a long CPWC is an indication of more competitive weeds or less competitive crops (Ghosheh et al., Reference Ghosheh, Holshouser and Chandler1996).

Table 1. Critical weed-free period (CWFP) and critical period of weed control (CPWC) for different crops^a

DAT, days after transplanting; DAE, days after emergence; DAS, days after sowing; GDD, growing degree days; NA, not available.

^a For 5% yield loss.

^b For 10% yield loss.

CTWR for different crops is presented in Table 2. While Nedeljković et al. (Reference Nedeljković, Knežević, Božić and Vrbničanin2021) could establish a narrow CTWR window of 16–19 DAE, Campos et al. (Reference Campos, Barroso, Silva Junior, Gonçalves and Martins2016) recorded a broader window of up to 25 DAE in maize. This is so pertinent that an increase in the duration of weed interference delayed silking in maize (Page et al., Reference Page, Cerrudo, Westra, Loux, Smith, Foresman, Wright and Swanton2012).

Table 2. Critical time of weed removal (CTWR) for different crops

DAE, days after emergence.

^a For 5% yield loss.

^b For 3% yield loss.

Type of weeds and emergence patterns

Parasitic weeds, such as Striga spp. (Jamil et al., Reference Jamil, Kanampiu, Karaya, Charnikhova and Bouwmeester2012) and Orobanche spp. (Westwood and Foy, Reference Westwood and Foy1999), remain unaccounted for in CPWC, as their germination pattern depends on the soil fertility and host presence (Raju et al., Reference Raju, Osman, Soman and Peacock1990). Weed emergence patterns are unpredictable; they may emerge over an extended period dictated by the prevailing weather, edaphic and crop factors (Vleeshouwers, Reference Vleeshouwers1997). Weed emergence patterns in spring hardly affected wheat yield but had a significant effect (4–20% yield loss) in autumn (Lotz et al., Reference Lotz, Kropff and Groeneveld1990). The weed flora in wet DSR may vary from transplanted rice due to differences in environmental conditions (Singh et al., Reference Singh, Singh, Mortimer, Johnson, Singh, Singh, Chauhan, Orr, Mortimer, Johnson and Hardy2008; Kumar and Ladha, Reference Kumar and Ladha2011), and accordingly, CPWC varies. Vegetative propagules, particularly non-dormant, have a competitive advantage over crops. For example, Cyperus esculentus propagation is exclusively by tubers in cultivated cropland, and the deepest tubers survive the longest (Stoller et al., Reference Stoller, Wax and Slife1979). Application of glyphosate will kill the weed above the ground but may not always kill the underground reproductive organs (ICID, 2002). Systemic herbicides (e.g. glyphosate) would be expected to be less effective where a large proportion of older tubers of Cyperus rotundus are present in the soil, and these herbicides fail to limit the regenerative capacity and tuber viability in the long term. Presence of Cyperus spp. and/or perennial weeds with vegetative propagules provide unrelenting competition for irrigated crops. The underground resource competition posed by these perennial vegetative propagules is not accounted for in the calculation of the CPWC, since mechanical weed removal is practised in CPWC calculation studies. Dhammu and Sandhu (Reference Dhammu and Sandhu2002) concluded that the critical period of Cyperus iria, an annual weed propagated by seeds, competition with transplanted rice is between 30 and 40 DAT akin to the general CPWC for transplanted rice, 20–40 DAT (Mukherjee et al., Reference Mukherjee, Sarkar and Maity2008). The proliferation of vegetative propagules of perennial weeds even after manual removal is noticed (Schimming and Messersmith, Reference Schimming and Messersmith1988; Lemieux et al., Reference Lemieux, Cloutier and Leroux1993), negating the concept of CPWC.

Several authors (Swanton and Weise, Reference Swanton and Weise1991; Baziramakenga and Leroax, Reference Baziramakenga and Leroax1994; Wiliams, Reference Wiliams2006) have endorsed that CWFP (i.e. the end of CPWC) would ensure maximum yield, as late-emerging weeds would not impair the crop productivity. Albeit, late-emerging weeds would still create weed problems through their seed input in the subsequent crops (Furlong, Reference Furlong2016). Under organic farming situations, cultural weed management (Bastiaans et al., Reference Bastiaans, Paolini and Baumann2008) is the prime mode of minimising weed pressure including, but not limited to, (i) enhancing crop competitive ability to weeds, and (ii) focusing on weed seed banks by either curtailing weed seedling recruitment and/or reducing the weed seed bank size (Schonbeck, Reference Schonbeck2011). Such positive effects on crop growth without weed interference, if properly translated from CPWC, might have implications in weed management. Dryland cotton growers in low-income countries resort to either pre-emergence (Deshpande et al., Reference Deshpande, Pawar, Mankar, Bobde and Chimote2006) or pre-plant incorporation of herbicides with the available soil moisture. Subsequent herbicide sprays are precipitation-dependent. In some instances, the left-over weed populations interfere with the cotton harvest, resulting in yield and quality losses (Smith et al., Reference Smith, Baker and Steele2000). Similarly, in peanut, the second flush of emerged weeds compete with the crop and infest the land with weed seeds (Kanagam and Chinnamuthu, Reference Kanagam and Chinnamuthu2009), resulting in heavy weed infestation in the subsequent crop.

Agro-ecology and crop management

Environmental factors or site-specific factors, dominant weeds in the region (Van Acker et al., Reference Van Acker, Swanton and Weise1993), tillage (Doll et al., Reference Doll, Doersch, Proost and Kivlin1992; Fortin and Hamill, Reference Fortin and Hamill1994), and soil salinity levels (Hakim et al., Reference Hakim, Juraimi, Musa, Ismail, Rahaman and Selamat2013) affect the duration of the critical period. Variations have been reported for mixed weed species, species to species, perennial weed to annual weed, and low to high weed pressure. For example, the CPWC for potatoes varied from 2 to 4 weeks after planting (Ivany, Reference Ivany1984, Reference Ivany1986) to 9 weeks after planting (Saghir and Markoullis, Reference Saghir and Marhoullis1974). Intensive and non-intensive production systems do modify the CPWC, for example, 28–117 days for intensive and 38–163 days after planting for non-intensive sugarcane production systems (Kouamé et al., Reference Kouamé, Orega, Touré and Kouabenan2014).

CPWC and row spacing have been found to influence the weed seed return to soil (Chandler et al., Reference Chandler, Shrestha and Swanton2001). The wider the row spacing, the higher the weed seed rain. Chandler et al. (Reference Chandler, Shrestha and Swanton2001) found greater weed seed return in soybean if the row spacing was 76 cm instead of either 38 or 19 cm (twin rows). Although the CWFP was similar across locations and years, CTWR varied among locations and between years in soybean (Van Acker et al., Reference Van Acker, Swanton and Weise1993). The alternate wetting and drying in wet DSR favoured several flushes of weeds and extended CPWC (Raj and Syriac, Reference Raj and Syriac2017) beyond the CPWC duration limit. Variations in CPWC for rice under varied growing ecologies are presented in Table 3. For example, Johnson et al. (Reference Johnson, Wopereis, Mbodj, Diallo, Powers and Haefele2004) noticed a CPWC of 4–83 DAS for the dry season lowland rice while only 29–32 DAS for the wet season lowland rice, underpinning that water regime as the chief determining factor of CPWC.

Table 3. Variation in the critical period of weed control (CPWC) for rice under varied growing ecologies

DAS, days after sowing; DAT, days after transplanting.

Further, CPWC has grossly ignored nutrient management, particularly the basal application of nutrients to crops. There are two probable situations under rainfed farming. First, a pre-emergence herbicide is ineffective due to either a lack of moisture or excess moisture after sowing under rainfed farming which when applied to the soil requires incorporation by rainfall, irrigation, tillage, etc. (Khalil et al., Reference Khalil, Flower, Siddique and Ward2019). Inthaphan and Thanomsak (Reference Inthaphan and Thanomsak1980) observed a complete loss of effectiveness of applying the herbicide to a dry rice seedbed if rains fail within 3–4 days of application. Second, post-emergence herbicides are applied around 20 DAS or slated for manual weeding around the end of CPWC. By that time, most of the soil-applied nutrients might have been extracted by the associated weed species. Particularly under rainfed cultivation, farmers either skip the basal application of fertilizers or resort to weeding at the end of CPWC. Even if applied, herbicides can be exhausted by weeds or lost in the soil–plant–atmosphere system before the next weeding. Therefore, understanding the influence of nutrient management on CPWC (Knezevic et al., Reference Knezevic, Evans, Blankenship, Van Acker and Lindquist2002) warrants further investigations.

In conservation agriculture systems, rotations of crops inhibit the buildup of weed seedbanks (Kassam et al., Reference Kassam, Friedrich, Shaxson and Pretty2009; Kassam and Friedrich, Reference Kassam and Friedrich2011); however, non-selective, non-residual herbicides, such as glyphosate and paraquat, are used for weed management before sowing a crop (Beckie et al., Reference Beckie, Flower and Ashworth2020) but they kill only emerged weed seedlings and have no effect on the weed seed bank. The interaction of weed–crop system becomes too complex under conservation agriculture (Ramesh, Reference Ramesh2015). The seeds from the weed seed bank would still germinate and compete with the crop. Wherever pre-emergence herbicides are used for killing a wide range of weed species, some species would remain unaffected, and the escaped weeds enrich the soil's weed seed bank (Singh, Reference Singh, Singh, Singh, Chauhan, Orr, Mortimer, Johnson and Hardy2008), or herbicide use in each cropping sequence would produce a shift in the weed seed bank in favour of species less susceptible to applied herbicides (Ball, Reference Ball Daniel1992). In the perturbed agricultural ecosystem, where only a single species is allowed to perpetuate, certain weeds would naturally become adapted to exclusion mechanisms and survive and reproduce even in the presence of herbicides. Non-selective herbicides employed in herbicide-tolerant crops destroy the total weedy vegetation. Broad-spectrum post-emergence herbicides with an extended period of weed control (up to 20 days after application) may not prevent the germination of weed seeds. As a result, weeds that emerge in the later stage of the crop (i.e. after the end of the CPWC) may cause damage to the system as a whole. The normal and predictable outcome of natural selection expressed as herbicide resistance (Heap, Reference Heap2013) and the herbicide-resistant weeds (Sosnoskie and Culpepper, Reference Sosnoskie and Culpepper2014) would add to the weed seed bank.

Weed seed bank

A final factor promoting greater attention is the weed seed bank. One among the chief omissions of the CPWC concept is the weed seed rain from the escaped weeds and the damage to the succeeding crop barring the standing crop productivity. Inevitably, leaving weed seed banks in the CPWC will sooner or later derail the concept of CPWC. For long-term weed management, in addition to the recommended pre-emergence and post-emergence herbicides, either one additional post-emergence herbicide might be required, as suggested by Martin et al. (Reference Martin, Vanacker and Friesen2001), or the weeds should be manually removed. Though the escaped weeds do not cause a significant yield loss in the standing crop, it increases the chance of higher weed infestation in the next season (Shrestha, Reference Shrestha2004). For example, allowing late-emerging E. colona and E. crus-galli plants (45 days after rice emergence) to produce even a few seeds may cause these weeds to be an increasing problem in the subsequent seasons through seed rains to the soil seed bank (Chauhan and Johnson, Reference Chauhan and Johnson2010; Bagavathiannan et al., Reference Bagavathiannan, Norsworthy, Smith and Neve2012). Knezevic and Datta (Reference Knezevic and Datta2015) revisited the data analysis for CPWC, but the weed seed bank remained unaccounted for.

Mechanical weed management and CPWC

Most of the CPWC studies are based on manual/mechanical removal of weeds, not herbicide-based weed removal. Manual removal of weeds does facilitate soil aeration, and thus the yield obtained includes the confounding effects of weed removal and improved physical soil conditions as it appears that repeated tillage creates an environment conducive for better plant establishment (Workayehu, Reference Workayehu2010). Ma et al. (Reference Ma, Xu, Li, Liu and Huang2008) found that root pruning in wheat restrained transpiration of wheat and reduced the consumption of soil water in the early growing stages whereas at later stages, it enhanced the photosynthetic rate, partitioning of more photosynthates to the shoots, and increased harvest index. Root pruning reduced the number of spikes in wheat but increased the grain yield per spike and the 1000-kernel weight (Fang et al., Reference Fang, Xu, Turner and Li2010). It is well-known that these effects are inseparable in the absence of herbicidal weed management, and the physical soil conditions have a significant positive effect on crop productivity. The process also involves root pruning, which is a beneficial practice in some crops for improved productivity through rapid reductions in stomatal conductance, for example, in sugarcane (Meinzer and Grantz, Reference Meinzer and Grantz1990) through the stomatal adjustment to the ratio of root hydraulic conductance to transpiring leaf area. In short, mechanical weed management carried out for CPWC studies may probably enhance the yield of the crops over unweeded or herbicide-applied crops.

General discussion and research needs

Although the concept of CPWC is a rule applied to several crops, it is not relevant to all field crops. For example, (i) initial slow growth in potato (Eberlein et al., Reference Eberlein, Patterson, Guttieri and Stak1997) and pigeon pea may reduce their ability to compete with weeds and only the CTWR is meaningful and not the CPWC, and (ii) perennial weeds may not fall under the CPWC umbrella (e.g. NT systems) (Wrucke and Arnold, Reference Wrucke and Arnold1985). In these crops, weeds grow taller than the crop species. How certain can we be that the CWFP is relevant in these crops or that the effect of CPWC really pertains to yield loss? It seems to be a biased estimate indeed. CPWC is just as descriptive as observed by Weaver and Tan (Reference Weaver and Tan1987) or CTWR should be synonymous with CPWC (Weaver and Tan, Reference Weaver and Tan1983) since CTWR is a relatively fixed estimate of the CWFP which has temporal variations (Karimmojeni et al., Reference Karimmojeni, Barjasteh, Mousavi and Bazrafshan2014). van Heemst (Reference van Heemst1985) has opined that the onset of the sensitive period is generally not very critical, however, the end of the critical period is. In some cases, the onset of the CTWR itself varied between tillage systems, as well as within them, in glyphosate-tolerant soybean (Mulugeta and Boerboom, Reference Mulugeta and Boerboom2000). CPWC would likely be meaningful only if weed seed rain to the weed seed bank is also considered. There are various issues in CPWC for crops and/or the associated weed flora. Even if we could pose some answers for the CPWC, there remains one basic underlying assumption that during the CPWC, the field may remain weed-free either at CTWR or CWFP. There must be significant differences for the weed-free crop at the start and end of CPWC, which is not addressed by the CPWC. It is clear, therefore, that under the above observations, the CTWR seems to be more meaningful than the CPWC. It might be suggested, however, that, CPWC should focus on the weed seed bank and weed seedling recruitment so that the CPWC concept covers the whole spectrum of issues. The CPWC concept should encompass growing degree days (Stagnari and Pisante, Reference Stagnari and Pisante2011; Anwar et al., Reference Anwar, Juraimi, Samedani, Puteh and Man2012), plant ecological concepts and weed seed banks towards an improved and more integrated understanding of CPWC across all crops. The need to establish diverse approaches that can relate weed biology studies to practical weed management is endorsed by Chauhan et al. (Reference Chauhan, Matloob, Mahajan, Aslam, Florentine and Jha2017). Although delineating CTWR for a crop in consonance with growing conditions may appear a herculean task with several experiments, these are needed to get meaningful and practical recommendations for minimizing weed competition in crops. Differences between seasons for the beginning and the end of the CPWC result in a change in weed densities (Tursun et al., Reference Tursun, Bükün, Karacan, Ngouajio and Mennan2007) that reinforces the need for a review of CPWC. A more comprehensive methodology to calculate the CPWC needs to be devised considering the following aspects viz., cultivation seasons, genotypes, weed seed banks, rainfed/irrigated conditions, late-emerging weeds, dormant vegetative propagules, perennial weed density, conservation agriculture systems, slow-growing crops, weed specific CPWC, etc.