Modelling the effects of herbicide dose and weed density on rice‐weed competition

The effects of herbicide dose on rice‐weed competition were investigated to develop a combined model, which can be utilised to estimate an optimum herbicide dose for a given weed density in paddy rice cultivation. Field studies were conducted in Suwon for rice‐Echinochloa crus‐galli competition and Iksan for rice‐Eleocharis kuroguwai during 2007. The competitive effect of the weeds E. crus‐galli and E. kuroguwai decreased with increasing doses of flucetosulfuron and azimsulfuron, respectively, in the same manner as the standard dose–response curve. The combination of the rectangular hyperbolic model and the standard dose–response curve adequately described the complex effects of herbicide dose and weed competition on rice yield. Parameter estimates were used with the model to predict rice yield and estimate the doses of flucetosulfuron and azimsulfuron required to restrict rice yield loss caused by E. crus‐galli and E. kuroguwai, respectively, to an acceptable level. For a rice yield of 5.0 t ha^?1, the model recommended flucetosulfuron doses of 8.7, 13.4 and 20.1 g a.i. ha^?1 when infested with E. crus‐galli at 12, 24 and 48 plants m^?2 respectively. For a rice yield of 5.2 t ha^?1, the model recommended azimsulfuron doses of 3.9, 7.5 and 12.6 g a.i. ha^?1 when infested with E. kuroguwai at 24, 48 and 96 plants m^?2 respectively. The theoretical outputs of the combined model appear robust and indicate there are opportunities for reduced herbicide use in the field. These now require evaluation under field conditions.     

Tillage and herbicide treatments with inter‐row cultivation influence weed densities and yield of three industrial crops

Field experiments were conducted in northern Greece in 2003 and 2004 to evaluate effects of tillage regimes (moldboard plowing, chisel plowing, and rotary tilling), cropping sequences (continuous cotton, cotton‐sugar beet rotation, and continuous tobacco) and herbicide treatments with inter‐row hand hoeing on weed population densities. Total weed densities were not affected by tillage treatment except that of barnyardgrass (Echinochloa crus‐galli), which increased only in moldboard plowing treated plots during 2003. Redroot pigweed (Amaranthus retroflexus) and black nightshade (Solanum nigrum) densities were reduced in continuous cotton, while purple nutsedge (Cyperus rotundus), E. crus‐galli, S. nigrum, and johnsongrass (Sorghum halepense) densities were reduced in tobacco. A. retroflexus and S. nigrum were effectively controlled by all herbicide treatments with inter‐row hand hoeing, whereas E. crus‐galli was effectively reduced by herbicides applied to cotton and tobacco. S. halepense density reduction was a result of herbicide applied to tobacco with inter‐row hand hoeing. Yield of all crops was higher under moldboard plowing and herbicide treatments. Pre‐sowing and pre‐emergence herbicide treatments in cotton and pre‐transplant in tobacco integrated with inter‐row cultivation resulted in efficient control of annual weed species and good crop yields. These observations are of practical relevance to crop selection by farmers in order to maintain weed populations at economically acceptable densities through the integration of various planting dates, sustainable herbicide use and inter‐row cultivation; tools of great importance in integrated weed management systems.     

Vertical distribution, size and composition of the weed seedbank under various tillage and herbicide treatments in a sequence of industrial crops

Investigations were conducted during the 2003, 2004 and 2005 growing seasons in northern Greece to evaluate effects of tillage regime (mouldboard plough, chisel plough and rotary tiller), cropping sequence (continuous cotton, cotton–sugar beet rotation and continuous tobacco) and herbicide treatment on weed seedbank dynamics. Amaranthus spp. and Portulaca oleracea were the most abundant species, ranging from 76% to 89% of total weed seeds found in 0–15 and 15–30 cm soil depths during the 3 years. With the mouldboard plough, 48% and 52% of the weed seedbank was found in the 0–15 and 15–30 cm soil horizons, while approximately 60% was concentrated in the upper 15 cm soil horizon for chisel plough and rotary tillage. Mouldboard ploughing significantly buried more Echinochloa crus‐galli seeds in the 15–30 cm soil horizon compared with the other tillage regimes. Total seedbank (0–30 cm) of P. oleracea was significantly reduced in cotton–sugar beet rotation compared with cotton and tobacco monocultures, while the opposite occurred for E. crus‐galli. Total seed densities of most annual broad‐leaved weed species (Amaranthus spp., P. oleracea, Solanum nigrum) and E. crus‐galli were lower in herbicide treated than in untreated plots. The results suggest that in light textured soils, conventional tillage with herbicide use gradually reduces seed density of small seeded weed species in the top 15 cm over several years. In contrast, crop rotation with the early established sugar beet favours spring‐germinating grass weed species, but also prevents establishment of summer‐germinating weed species by the early developing crop canopy.     

Long‐term changes in weed occurrence,yield and use of herbicides in maize in south‐western Germany,with implications for the determination of economic thresholds

In 393 field experiments in Baden‐Württemberg region in south‐western Germany, herbicide efficacy, yield loss and crop tolerance of maize (Zea mays) were investigated between 1981 and 2011. The collected data served to determine changes in weed frequencies, in herbicide use, yield loss functions and economic thresholds (ETs). Over 60 weed species were reported. Chenopodium album and Galium aparine were the most frequent broad‐leaved weeds, the former becoming more frequent over time. Species of the genera Lamium, Polygonum, Veronica and Matricaria occurred in about every fifth trial. Alopecurus myosuroides and Echinochloa crus‐galli were the most frequent grass weeds; the former declining in frequency by 1.1% per year, the latter increasing by 1.5%. Results suggest a weed population shift towards thermophilic species. aceto‐lactate‐synthase and 4‐HPPD‐inhibitor herbicides became important in the 1990s. Pendimethalin and bromoxynil have been integral components of weed control since the 1980s. ETs, the point at which weed control operations provide economic returns over input costs, ranged between 3.7% and 5.8% relative weed coverage. Without weed control, no yield increase was found over 24 years. Yield increased by 0.2 t ha ^{? 1} year ^{? 1}, if weeds were controlled chemically. Despite intensive use of effective herbicides in maize, problematic weed species abundance and yield losses due to weed competition have increased in Baden‐Württemberg over a period of 30 years.     

Effects of ferrous iron (Fe) on the germination and root elongation of paddy rice and weeds

Laboratory experiments were conducted to analyze the iron (Fe) tolerance of paddy weeds and rice varieties (Oryza sativa) for germination and root elongation. Under a waterlogged soil condition, the Fe(II) content in a soil solution increased with an increase in the ratio of rice straw to the soil. In the presence of 0.9% (w/w) straw to soil, which corresponds approximately to 8 t of straw applied to an area of 1 ha × 10 cm depth in the field, ～80 mg L^?1 of Fe(II) was produced in the soil solution. Based on this result, the seeds of rice and the weeds were incubated in a solution with <100 mg L^?1 of Fe(II). The presence of 100 mg L^?1 of Fe(II) suppressed the germination of Echinochloa crus‐galli var. crus‐galli, Cyperus serotinus, Cyperus difformis, and Monochoria korsakowii. However, it had no effect on the germination of Echinochloa oryzicola, Schoenoplectus juncoides (= Scirpus juncoides var. ohwianus), and Monochoria vaginalis. This level of Fe tolerance was the same as that of rice. These findings suggest that E. oryzicola, S. juncoides, and M. vaginalis can grow under more severe conditions than E. crus‐galli, C. serotinus, C. difformis, and M. korsakowii. In relation to seminal root elongation, the order of tolerance of Fe toxicity was O. sativa cv. Dunghan Shali > O. sativa cv. Hoshinoyume > E. oryzicola > M. vaginalis > S. juncoides. Thus, the results show that the tolerance of rice is greater than that of E. oryzicola, which had a comparatively strong tolerance among the weeds examined, and also that there are differences in tolerance among the rice varieties. These findings suggest that the difference in Fe tolerance is involved in weed control systems when organic materials are applied. If this difference is an important factor in the weed control system, Fe‐tolerant rice varieties, like cv. D. Shali, could facilitate weed control systems due to their higher Fe tolerance ability.     

Weed species composition of conventional soyabean crops in Hungary is determined by environmental,cultural, weed management and site variables

The goal of this study was to identify factors determining weed species composition in soyabean crops in Hungary, where its expanding production faces difficult weed problems. The abundance of weed flora was measured in 262 fields across the country, along with 38 background variables. Using a minimal adequate model containing 24 terms with significant net effects, 21.6% of the total variation in weed species data could be explained. Plot location (edge vs core position, the single site variable in our analysis) was found to be the most important explanatory variable that was followed by a set of environmental (temperature, precipitation, altitude, soil texture, pH, Ca, K, Na and humus content), cultural (cultivar maturity, organic manure, fertiliser P and N, row spacing) and weed management (flumioxazin, pendimethalin, dimethenamid, propaquizafop, bentazone, quizalofop‐p‐ethyl, quizalofop‐p‐tefuril, linuron, thifensulfuron) factors. Variation partitioning revealed that environmental variables accounted for about four times more variance than cultural and about two and half times more than weed management variables. Chenopodium album, Ambrosia artemisiifolia, Hibiscus trionum, Echinochloa crus‐galli and Convolvulus arvensis were the most dominant and frequent weeds, but their abundance was influenced by different factors. The responses of weed species to the studied variables provide new information about their ecological behaviour, and our findings also can be used to develop better weed management strategies.     

Use of phytotoxic rice crop residues for weed management

There is a general perception among Cambodian rice (Oryza sativa) farmers that, after harvesting, rice crop residues that are incorporated into the field benefit the growth of the subsequent rice crop. However, the effect of this action upon weed establishment and growth has not yet been considered. A series of pot and field trials were conducted to determine whether such action could inhibit weed establishment and/or growth. The pot studies first evaluated the response of the test plant (rice line ST‐3) and three weed species, barnyardgrass (Echinochloa crus‐galli), small umbrella sedge (Cyperus difformis), and water primrose (Ludwigia octovalves), to the residue of 16 rice lines and the field trials were later conducted to evaluate the response of the same test plants to the residue of seven putatively allelopathic rice lines and one non‐allelopathic rice line. The residue of all the studied rice lines, depending on how long they had been incorporated into the soil, reduced the establishment and growth of all three weed species, as well as the rice crop. However, if the residue's incorporation was delayed by 2 weeks or only a proportion of the residue was incorporated, the rice crop could withstand the growth‐inhibiting effect, while the inhibition of the establishment and growth of the three weed species was retained. These responses of rice and the weeds to rice crop residues might provide a basis for a weed management strategy, particularly in the resource‐poor rice‐production systems of Cambodia.     

Crop–weed competition rather than temperature limits the population establishment of two annual C4 weeds at the edge of their northern range

Climate change is predicted to affect range expansion of harmful C₄ weeds into the boreal region, given that they are able to successfully colonise both C₃ and C₄ crops. We studied the impact of a 3°C elevation in temperature on the establishment and maintenance of populations of two annual C₄ weeds (Amaranthus retroflexus and Echinochloa crus‐galli) with and without a competing C₃ (barley) or C₄ (maize) crop. Data obtained from field and glasshouse experiments were modelled using a periodic matrix population model. Competition of a weed with a crop appeared to be a more important factor for limiting the maintenance of weed populations than elevation in temperature, as neither of the weed species was able to maintain populations in competition with crops. Even an increase in the frequency of warm years did not result in viable weed populations establishing. However, A. retroflexus was able to form persistent populations in competition with maize when released from competition every fifth year. Simulations parameterised from glasshouse data predicted that both weed species would persist without competition in the current climate, whereas simulations parameterised from field data suggested only A. retroflexus to be able to persist. These results demonstrate that competition affects the range expansion of arable weed species more than elevation in temperature, necessitating the inclusion of crop–weed interactions in models of range shifts as a consequence of climate change.     

Chloroplast DNA markers for Echinochloa taxa

The genus Echinochloa (Poaceae) includes numerous problematic weeds over a wide range of ecoregions in the world. To date, molecular markers for species identification and assessing phylogenetic relationship are still limited in the genus Echinochloa. In this study, we developed seven chloroplast molecular markers based on divergent chloroplast regions of E. crus‐galli and E. oryzicola. Furthermore, Marker #1 (psbA) was examined in more than 200 Echinochloa accessions and a phylogenetic tree grouped these Echinochloa accessions into four clades. Additionally, two different E. crus‐galli varieties (E. crus‐galli var. crus‐galli, E. crus‐galli var. praticola) and E. colona were successfully distinguished by this marker. The developed molecular markers contribute to better identification of Echinochloa taxa.     

Impacts of cropping system and management practices on the assembly of weed communities

Understanding how weed communities assemble as a function of biotic and abiotic filters and transform through time has important implications for the sustainable management of agronomic systems. In a three‐year study, we evaluated weed community responses to lucerne (Medicago sativa, perennial) vs. continuous spring wheat (Triticum aestivum, annual, CSW) and weed management practices where weeds in the CSW system were managed with three contrasting approaches (herbicide, tillage or sheep grazing). Our results indicated no differences in weed diversity between the perennial and annual crops or across the different management practices in CSW. However, there were differences in weed community composition. Lucerne, with the exception of the establishing year, impeded the growth and reproduction of several annual weeds, including Amaranthus retroflexus, Thlaspi arvense, Lamium amplexicaule and Chenopodium album, but favoured perennial broad‐leaved weeds such as Taraxacum officinale and Cirsium arvense. The replacement of herbicide treatments in pre‐plant and post‐harvest in CSW with soil tillage or sheep grazing selected for different weed communities beyond the second year of establishment. The weed species driving the differences in CSW systems were Androsace occidentalis, more common in CSW managed chemically; Asperugo procumbens, more common in CSW managed with tillage; and T. officinale and Lactuca serriola, more common in CSW managed with sheep grazing. Understanding how cropping systems modify weed communities is a necessary step to shift from reactive weed control programmes to predictive management strategies.     

A novel screening method for rice allelopathic potential: the inhibitory‐circle method

Screening crop accessions for allelopathic activity is of paramount importance for crop allelopathy research. Previous bioassays often did not use a mixed culture of donor and target plants, did not use soil and were not conducted under natural conditions. In this study, we designed an inhibitory‐circle method in which a rice accession (donor plant) and Echinochloa crus‐galli (target plant) were cultured together in paddy soil under natural conditions. First, we determined that the highest allelopathic activity of allelopathic rice accession PI312777 was at the 5‐leaf stage, and the suitable distance of rice seedlings and E. crus‐galli was 12 cm apart. This method was then validated by a field test. A further 40 rice accessions were evaluated for allelopathic activity to E. crus‐galli using this method. Two rice accessions, PI312777 and Taichung Native 1, had highly allelopathic activity to E. crus‐galli (inhibitory rate > 50%), while another accession, Lemont, had non‐allelopathic activity. These experimental results were in accordance with previous studies using direct field experiments. The inhibitory‐circle method integrated three necessary conditions, that is donor and target plants grown together, with soil as the medium and under natural conditions for reliable results. The ‘inhibitory‐circle method’, which combined donor and target plants, soil medium and field conditions, can give reliable results in one step, compared with laboratory screening methods. Also, the ‘inhibitory‐circle method’ gave results in 30‐35 days, thereby substantially reducing the requirements for time, labour and cost.     

Restriction site‐associated DNA sequencing allows for the rapid identification of simple sequence repeat markers in Echinochloa crus‐galli

Echinochloa crus‐galli is a serious weed worldwide. Microsatellite markers (simple sequence repeats, SSRs) are important molecular markers that are used widely for studying genetic diversity in plants. However, a limited number of SSRs is available for E. crus‐galli. The restriction site‐associated DNA (RAD) sequencing approach was combined with Illumina DNA sequencing for the rapid and mass detection of SSRs in E. crus‐galli. The RAD tags were generated from the genomic DNA of E. crus‐galli and were sequenced in order to produce 6921.6 Mb of high‐quality sequences with 45.1% guanine–cytosine content. In total, 3081 putative SSRs were detected, of which 82.2% were dinucleotide motif‐repeats. AT was the most frequent motif, accounting for 35.0% of the SSRs. In order to test the validity of the SSRs that were developed here, eight SSRs that were selected from putative SSRs were used to study the genetic diversity and structures of 20 E. crus‐galli populations that had been collected from rice fields in eastern China. Ninety‐seven alleles were amplified from the eight microsatellite loci among the 20 E. crus‐galli populations. These populations showed low genetic diversity and were classified on the basis of their genetic structures into three distinct groups that corresponded to the three regions of population sampling. The SSRs that were identified in this study represent a valuable resource for studying the genetic diversity, population biology and evolution of E. crus‐galli.     

Suppression of Digitaria sanguinalis and Amaranthus palmeri using autumn-sown glucosinolate-producing cover crops in organically grown bell pepper

Experiments were conducted to compare growth characteristics, biomass production and glucosinolate content of seven autumn‐planted glucosinolate‐producing cover crops that were terminated the following spring. The control of Digitaria sanguinalis and Amaranthus palmeri following cover crop incorporation into soil was characterised and fruit yields of bell pepper transplanted into cover crop‐amended soil were determined. Differences in glucosinolate concentration and composition were noted between cover crop roots and shoots and among cover crops. Total biomass production by cover crops ranged from 103 g m⁻² for garden cress to 894 g m⁻² for Indian mustard (F‐E75), but cover crop biomass was not correlated with D. sanguinalis and A. palmeri control. D. sanguinalis and A. palmeri control in bell pepper varied by cover crop. D. sanguinalis control by cover crops ranged from 38% to 79%, and A. palmeri control was 23% to 48% at 4 weeks after transplanting (WATP) bell pepper in 2004. D. sanguinalis control was positively correlated with total glucosinolate production, but A. palmeri control was not. D. sanguinalis control in 2005 ranged from 0% to 38% at 2 WATP. In the absence of weeds, cover crops did not negatively affect fruit yields which were often higher than in the absence of a cover crop. Glucosinolate‐producing cover crops are not a stand‐alone weed management strategy, but some will provide early season control of D. sanguinalis and A. palmeri without having a negative effect on transplanted bell pepper.     

Are weed patches stable in location? Application of an explicitly two‐dimensional methodology

Field observations were made in three years continuous maize cultivation in the Netherlands to study the spatial pattern and stability of spatial pattern over time in agricultural weeds. Two‐dimensional correlograms were made, using data from single years, to characterise spatial correlation and pattern, while data from two different years were used to calculate correlation over space and time, to characterise the stability of pattern. Weeds that were able to attain high recruitment also exhibited the strongest spatial correlations. These weeds were Echinochloa crus‐galli, Chenopodium album, Chenopodium polyspermum and Solanum nigrum. Weeds that were less successful in attaining high densities in the maize rotation, also showed less spatial correlation. Wind dispersing Compositae, e.g. Taraxacum officinale, had spatially uncorrelated patterns. All weeds that showed spatial correlation also showed stability in space, except E. crus‐galli. The latter species showed marked population increase and the locations and extent of its patches changed over the years. Statistical interpretation of the data is discussed, as are potential consequences for site‐specific management and optimal sampling of weeds.     

Utilization of Bidens pilosa var. radiata (Sch. Bip.) Sherff integrated with water irrigation for paddy weed control and rice yield production

The application of allelopathic plant materials combined with an agricultural production method is a concept that can be used to manage weeds during rice production. This research investigated the effects of the utilization of Bidens pilosa var. radiata integrated with water irrigation for weed control and rice production. B. pilosa and water extract exerted more toxicity against Echinochloa crus‐galli than the extracted residue applied at the same rates, suggesting that weed control was caused by compounds extracted from B. pilosa. The changes in water properties after the application of B. pilosa were explored. Dissolved oxygen was rapidly reduced after application and then slowly increased within 3–5 days after application (DAA), while the pH also decreased at the early stage and then increased back to normal values within 1–3 DAA. Electrical conductivity increased with increasing rates of application, suggesting that allelochemicals were released from B. pilosa. The concept was studied under field conditions. B. pilosa was applied along with irrigation at 7 days after sowing (DAS) on a direct wet seed system. The results showed that application rates of 2, 4 and 6 tons ha^?1 decreased the number of weeds by 52.16, 86.73 and 95.18% at 30 DAA and reduced dried weight of weeds by 17.65, 34.69 and 86.82%, respectively. The rice yield showed an increase of 72.73, 81.03 and 73.66% when compared with the nonweeded crop, while herbicide treatment increased by 84.68%. The findings demonstrate that the application of B. pilosa integrated with water irrigation was able to control weeds in a direct seeded rice production system.     

The impact of management on weeds and aquatic plant communities in Hungarian rice crops

This study assessed the cultural and weed management factors influencing the weed communities of Hungarian rice fields. Hungary is situated at the northern limit of rice production with a history of about 300 years of rice culture. We surveyed the weed flora and 25 background variables in 100 active rice fields. Using a minimal adequate model containing 11 terms, 48.5% of the total variation in weed species data could be explained. The net effects of nine variables on species composition were significant. Crop cover was found to be the most important explanatory variable, which was followed by the herbicides penoxsulam and azimsulfuron, tillage depth, phosphorous and potassium fertilisers, years after last rotation, water depth in May, sowing type, pendimethalin and water conductivity. Filamentous algae, as the most abundant group of weeds, were positively associated with deep tillage, deep water and surface sowing. Echinochloa crus‐galli, one of the most troublesome grass weeds, was associated with low rice cover, shallow water and later years after crop rotation, while weedy rice favoured high crop cover, deep water and soil sowing. These findings can be used to design improved weed management strategies. The occurrence of red list species and charophytes in diverse micro‐mosaic patterns deserves attention from a conservation perspective, as well. The maintenance of these unique charophyte communities can be facilitated by shallow tillage without soil inversion.     

Genetic relationship between Echinochloa crus‐galli and Echinochloa oryzicola accessions inferred from internal transcribed spacer and chloroplast DNA sequences

Barnyardgrass, hexaploid Echinochloa crus‐galli, is considered to arise from the hybridization between tetraploid Echinochloa oryzicola and an unknown diploid species. The genetic relationship between E. crus‐galli and E. oryzicola was examined to investigate the position of E. oryzicola in the evolutionary process of E. crus‐galli, based on the nuclear DNA internal transcribed spacer (ITS) and the chloroplast cpDNA trnT‐L, trnL intron, and trnL‐F regions. New World E. crus‐galli was clearly separated from Eurasian E. crus‐galli and showed a close relationship to the American taxa, Echinochloa crus‐pavonis and Echinochloa walteri, in both the ITS and chloroplast DNA. The nrDNA ITS sequences indicated no differentiation between the Eurasian E. crus‐galli and E. oryzicola, in contrast to their clear divergence in the cpDNA sequence. The present results suggest that E. oryzicola is the male donor of E. crus‐galli.     

Response of Sorghum halepense demographic processes to plant density and rimsulfuron dose in maize

In spatially heterogeneous weed infestations, variable dose technologies could be used to minimise herbicide use; high doses could be applied to reduce high‐density patches and low doses to maintain weed populations in low‐density portions of a field. To assess the potential short‐ and long‐term effects of variable herbicide dose and site‐specific management, the major weed demographic processes were described and parameterised in this study. Various doses of rimsulfuron (from 0 to 12.5 g a.i. ha^?1) were applied to different densities of Sorghum halepense (0–100 plants m^?2). Contrary to similar studies with other weed species, higher herbicide efficacy was not observed at low densities, suggesting that the same rimsulfuron dose should be applied regardless of the S. halepense density. The highest percentage of control was obtained with the full rimsulfuron dose. However, it did not guarantee a decrease of the infestation in the following season in the field areas where the initial S. halepense density was lower than 60 plants m^?2. Reduced doses of rimsulfuron to control S. halepense cannot be recommended based on our results.     

Spatial correlation between weed species densities and soil properties

The spatial cross‐correlation between weed species densities and six soil properties within fields was analysed using cross‐semivariograms. The survey was carried out in three successive years in two fields. The most consistent relationship between weed species density (numbers m^?2) and soil properties was negative cross‐correlation between the density of Viola arvensis Murray and clay content. This correlation was found in both fields; however, the range of spatial dependence varied between fields. In one of the fields, the density of Lamium purpureum L. was positively cross‐correlated with the phosphorus content in the soil in all years. The density of Veronica spp. and Poa annua L. was negatively cross‐correlated with pH in all three years. Other spatial cross‐correlations that were found in this study were inconsistent over time or field site. The densities of some of the weed species were spatially cross‐correlated with more than one soil property. The results showed that the range of spatial dependence varied not only between fields, but also between weed species and soil properties, as well as between years. This study indicates that the weed pattern is field‐specific and that the spatial variation in soil properties within a field is one of several factors affecting weed patchiness.     

Using panicle dry weight to estimate seed production in Echinochloa crus‐galli

A better understanding of weed seed production is a key element for any long‐term management allowing some weeds to shed seeds. The challenge with measuring seed production in weeds is the large effort required in terms of time and labour. For the weed species Echinochloa crus‐galli, it was tested whether the number of seeds per panicle dry weight or per panicle length can be used to estimate seed production. Experiments were conducted in three maize fields in north‐eastern Germany. The effect of factors that could influence this relationship, such as the time of seedling emergence, the density of E. crus‐galli, the control intensity of other weeds, seed predation and field, was included. A few days before maize harvest, all panicles were removed and weighed, panicle length was measured, and for a subsample of 178 panicles, the number of seeds was counted manually. Panicle dry weight predicted the number of seeds per panicle better (R² = 0.92) than did panicle length (R² = 0.69). The other factors except for ‘field’ and ‘seed predation’ had no effect on these relationships. The relationships between seed number and panicle dry weight found in this study closely resembled those reported in an earlier study. Based on our results, we conclude that both plant traits are appropriate for the estimatation of seed production, depending on required level of precision and availablilty of resources for the evaluation of sustainable weed management strategies.