The effect of crop,cultivation and seed addition for birds on surface weed seed densities in arable crops during winter

Weed seeds present an agronomic threat, but are also an important food resource for wildlife in winter. Weed seed densities on the soil surface in winter were examined from 1999 to 2002 in 105 fields on three different farms in UK. The effect of the preceding crop, cultivation, position within the field and the application of seed for birds (bird seed) on surface seed abundance and species composition was tested. Six or fewer species comprised c. 80% of the weed seeds. By January of each study year, the densities of seeds important for farmland birds (key seeds) were 73% or 87% lower compared with early winter on two of the farms, but were stable on the third where seeds were incorporated through cultivation. At the edge and mid‐field, seed densities only exceeded 400 m^?2 in 17%, 10% and 12% of fields for total, key and dicotyledonous seeds respectively. The preceding crop only affected seed densities at one site; stubbles of winter barley had fewer seeds compared with winter wheat or spring barley. Seed densities varied between the edge and mid‐field, but trends were inconsistent between sites. The density of the larger seeds (Atriplex patula, Viola arvensis, Polygonum aviculare and Chenopodium album) were reduced in fields receiving bird seed. The objectives of weed control and conservation may not be mutually exclusive because seed return was most reduced where the ground remained uncultivated through the winter, yet this also provided the best foraging opportunities for surface feeding seed predators.     

Fates of Setaria faberi and Abutilon theophrasti seeds in three crop rotation systems

Weed seeds in and on the soil are the primary cause of weed infestations in arable fields. Previous studies have documented reductions in weed seedbanks due to cropping system diversification through extended rotation sequences, but the impacts of different rotation systems on additions to and losses from weed seedbanks remain poorly understood. We conducted an experiment in Iowa, USA, to determine the fates of Setaria faberi and Abutilon theophrasti seeds in 2‐, 3‐ and 4‐year crop rotation systems when seed additions to the soil seedbank were restricted to a single pulse at the initiation of the study. Over the course of the experiment, seedlings were removed as they emerged and prevented from producing new seeds. After 41 months, seed population densities dropped >85% for S. faberi and >65% for A. theophrasti, but differences between rotation systems in the magnitude of seedbank reductions were not detected. Most of the reductions in seedbank densities took place from autumn through early spring in the first 5 months following seed deposition, before seedling emergence occurred, suggesting that seed predation and/or seed decay was important. For S. faberi, total cumulative seedling emergence and total seed mortality did not differ between rotation systems. In contrast, for A. theophrasti, seedling emergence was 71% lower and seed mortality was 83% greater in the 3‐ and 4‐year rotation systems than in the 2‐year system. Results of this study indicate that for certain weed species, such as A. theophrasti, crop rotation systems can strongly affect life‐history processes associated with soil seedbanks.     

Comparison of growth,survivorship, seed production and shedding of eight weed species in a wheat crop in Western Australia

Development of integrated weed management strategies is dependent on a thorough knowledge of the demography of individual species. The current research established eight winter or summer weed species in a winter annual wheat cropping system at Wongan Hills, Western Australia, and investigated emergence of the first cohort of each species, survivorship, plant size, seed production and seed shedding over three years (2016–2019). The winter weeds Bromus diandrus and Lolium rigidum emerged at the same time as the wheat crop, and the initial cohort of marked plants had 100% survival to seed production in each year. By comparison, other winter weed species like Hordeum leporinum, Rumex hypogaeus, Sonchus oleraceus and Polygonum aviculare frequently emerged later than the crop and had a lower percentage of plants surviving to seed production. However, individual S. oleraceus and P. aviculare plants had the greatest seed production compared to other species. All winter weeds had variable patterns of seed shedding between years, with the exception of L. rigidum. Summer weed species emerged at the same time, but plants in the initial cohort of each species did not always survive to produce seed. The early emergence and high survivorship of B. diandrus indicates high competitive ability, but shedding commenced at a similar time to L. rigidum and harvest weed seed control may be a viable control method for this species.     

Consequences of narrow crop row spacing and delayed Echinochloa colona and Trianthema portulacastrum emergence for weed growth and crop yield loss in maize

Echinochloa colona and Trianthema portulacastrum are weeds of maize that cause significant yield losses in the Indo‐Gangetic Plains. Field experiments were conducted in 2009 and 2010 to determine the influence of row spacing (15, 25 and 35 cm) and emergence time of E. colona and T. portulacastrum (0, 15, 25, 35, 45 and 55 days after maize emergence; DAME) on weed growth and productivity of maize. A season‐long weed‐free treatment and a weedy control were also used to estimate maize yield and weed seed production. Crop row spacing as well as weed emergence time had a significant influence on plant height, shoot biomass and seed production of both weed species and grain yield of maize in both years. Delay in emergence of weeds resulted in less plant height, shoot biomass and seed production. However, increase in productivity of maize was observed by delay in weed emergence. Likewise, growth of both weed species was less in narrow row spacing (15 cm) of maize, as compared with wider rows (25 and 35 cm). Maximum seed production of both weeds was observed in weedy control plots, where there was no competition with maize crop and weeds were in rows 35 cm apart. Nevertheless, maximum plant height, shoot biomass and seed production of both weed species were observed in 35 cm rows, when weeds emerged simultaneously with maize. Both weed species produced only 3–5 seeds per plant, when they were emerged at 55 DAME in crop rows spaced at 15 cm. Infestation of both weeds at every stage of crop led to significant crop yield loss in maize. Our results suggested that narrow row spacing and delay in weed emergence led to reduced weed growth and seed production and enhanced maize grain yield and therefore could be significant constituents of integrated weed management strategies in maize.     

Weed seed rain interception by grass cover depends on seed traits

Weed management requires a better understanding of the dynamics of the weed seedbank, which is a primary source of weeds in a field. Seeds reaching the ground after seed rain replenish the seedbank and therefore contribute to future weed infestations. Our investigation is based on the hypothesis that a permanent vegetation cover, such as a grassland, can prevent weed seeds from reaching the ground. Therefore, we developed an innovative experimental device to simulate in controlled conditions the seed rain of 12 weed species (Capsella bursa‐pastoris, Conyza canadensis, Myosotis arvensis, Papaver rhoeas, Poa annua, Polygonum aviculare, Ranunculus sp., Rumex obtusifolius, Sonchus asper, Stellaria media, Taraxacum officinale and Veronica persicaria). We quantified the interception of weed seeds by a grass cover. Grass cover height, seed size and seed appendage (e.g. pappus, wing or awn) increased seed interception, in contrast to seed weight and shape index. From these results, we established a linear model to predict weed seed interception by a grass cover as a function of their seed trait values. The relationship between the predicted interception and weed community dynamics observed in grasslands was negative for some species, indicating that other processes may be involved depending on weed species. The weed seed interception model will be incorporated into an existing model of weed population dynamics to simulate the impact of grassland insertion into arable crop rotations.     

Invasiveness of agronomic weed species in wheat in Western Australia

Weed seeds are introduced to agronomic systems naturally or through human-mediated seed dispersal, and introduced seeds have a high chance of being resistant to selective, in-crop herbicides. However, colonisation (invasion) rates for a weed species are usually much lower than rates of seed dispersal. The current research investigated colonisation of a winter annual wheat cropping system in Western Australia by a range of winter or summer annual weed species. The weed seeds were sown (at 100 seeds/m²) directly before seeding the crop in 2016 and allowed to grow in the following 3 years of wheat. Selective herbicides were not applied, to simulate growth of weed populations if the initial seed had been resistant to herbicide. Bromus diandrus, Hordeum leporinum, Rumex hypogaeus, Sonchus oleraceus, Polygonum aviculare, Lolium rigidum, Citrullus amarus and Tribulus terrestris colonised the crop, while Dactyloctenium radulans, Chloris truncata and Salsola australis failed to establish over 3 years. The most successful weed was B. diandrus, with a plant density of 1,170/m² by the third year and seed production of 67,740/m². The high density of B. diandrus reduced wheat density by 76% in the third year and reduced average yield by 36%. Lolium rigidum reduced average yield by 11%, and the other weed species did not affect crop yield. Further research is required on the invasiveness of these species in other regions, but it is clear that the spread of B. diandrus to new areas or the introduction of resistant B. diandrus seeds via contaminated grain should be avoided.     

Post-dispersal seed predation of non-target weeds in arable crops

Field experiments were conducted to quantify the natural levels of post-dispersal seed predation of arable weed species in spring barley and to identify the main groups of seed predators. Four arable weed species were investigated that were of high biodiversity value, yet of low to moderate competitive ability with the crop. These were Chenopodium album, Sinapis arvensis, Stellaria media and Polygonum aviculare. Exclusion treatments were used to allow selective access to dishes of seeds by different predator groups. Seed predation was highest early in the season, followed by a gradual decline in predation over the summer for all species. All species were taken by invertebrates. The activity of two phytophagous carabid genera showed significant correlations with seed predation levels. However, in general carabid activity was not related to seed predation and this is discussed in terms of the mainly polyphagous nature of many Carabid species that utilized the seed resource early in the season, but then switched to carnivory as prey populations increased. The potential relevance of post-dispersal seed predation to the development of weed management systems that maximize biological control through conservation and optimize herbicide use, is discussed.     

Factors impacting the detection of weed seed contaminants in seed lots

BACKGROUND

The setting and following of phytosanitary standards for weed seeds can lessen the impacts of weeds on agriculture. Standards adopted by seed companies, laboratories and regulators ensure the contamination rates do not exceed some thresholds. Globally sample size standards are set based on the amount needed to obtain a contaminant in a random sample of the seed lot, not detectability. New Zealand requires a 95% confidence that the maximum pest limit of 0.01% of quarantine weed seed contamination is not exceeded in an imported seed lot. We examined 24 samples each containing approximately 150 000 seeds of either perennial ryegrass (12 samples) or white clover seeds (12 samples) that were then spiked with seeds (contaminants) from 12 non-crop species (3–8 seeds of each). We considered factors that may impact detection rates: shape, color, size, and texture relative to the crop, and technician (including a commercial seed laboratory).

RESULTS

A linear mixed model fitted to the data indicated significant observer, crop, and seed color, shape, and size effects on detection. Detectability increased by 20% ± 7.7 (± standard error) when seeds had a distinct shape or color (28% ± 8.1), or were larger (23% ± 8.7) rather than smaller, relative to the crop. Commercial laboratory identifications were usually correct at the level of genus, and species for common weeds, but some misidentifications occurred.

CONCLUSION

Sample sizes for border inspections should be based on detectability of regulated weed seeds in the crop in combination with weed risk for the crop and location. © 2022 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.     

Contribution of allelopathic effects to the overall weed suppression by different cover crops

Cover crops can suppress weeds within agricultural fields due to competitive and allelopathic effects. Glasshouse experiments were conducted to evaluate the relative proportions of allelopathic effects to the total weed inhibition. Six different cover crop species were combined with three weed species in the presence or absence of active carbon over a period of four weeks. Active carbon was used as an adsorbent for allelopathic substances in the soil. Our study revealed that the competition between cover crops and weeds shifted, possibly due to the minimisation of allelopathic effects by active carbon in the soil. We assume that the degree of cover crops allelopathic effects on weeds is species‐specific, both on the side of cover crops and on the weed side. The cover crops Raphanus sativus, Fagopyrum esculentum and Avena strigosa showed the highest allelopathic weed suppression with up to 28%. Additionally, Stellaria media turned out to be the most sensitive weed against allelopathic effects induced by all cover crops, except for Linum usitatissimum and Guizotia abyssinica. The knowledge about the contribution of competitive and allelopathic effects by cover crops would help to create cover crop mixtures with high weed suppressive ability.     

Seed survival and patterns of seedling emergence studies of Rottboellia cochinchinensis (Lour.) W.D. Clayton in cultivated soils

The longevity of buried Rottboellia cochinchinensis (Lour.) W.D. Clayton seed represents a major survival mechanism for the weed, enabling the persistence of a continuing source of weed seeds in crop land. The pattern of seed persistence and depletion of R. cochinchinensis in cultivated maize soils was investigated by means of (1) studies on the effect of depth and duration of burial on the viability of the weed seeds, (2) quantitative estimation of the seed population and viability in cultivated fields, and (3) the periodicity of emergence and effects of cultivation on seed germination both in the field and in the greenhouse. The results indicated that the mode of persistence was innate (8.5%) and enforced (35%) dormancy after 1 year of burial, and that the persistency component of the seed population on cultivated soils ranged from 40.60%. The weed was able to remain viable at depths of 45 cm, indicating an excellent mechanism of escaping the effects of most soil-applied herbicides, and it was shown that tillage increases the depletion rate of the weed seed reserve by 32% per year.     

Impact of Brassica juncea biofumigation on viability of propagules of pernicious weed species

Biofumigation may be a promising tool for depletion of persistent weed seedbanks/bud banks. This technique is based on the incorporation of chopped glucosinolate‐rich plant biomass into the soil, upon which isothiocyanates with herbicidal properties are released. To gain acceptance by farmers and foster its implementation, the biofumigation process should be further optimised. This study elucidated the impact of biological (species), technical (burial depth, ground cover) and pedohydrological (temperature and moisture content) factors on efficacy of Brassica juncea biofumigation under (semi‐)natural conditions. In a first experiment (field experiment), seeds and vegetative propagules of various weed species were buried at different depths and exposed to different doses of fresh fine‐chopped B. juncea biomass in the presence or absence of a plastic ground cover. In a second experiment (container experiment), buried seeds of ten species were subjected to biofumigation at diverging soil organic matter content, soil moisture content and soil temperature. In a third experiment (dose–response Petri dish bioassay), unburied seeds of eight species were subjected to various doses of rehydrated B. juncea powder. Biofumigation efficacy was determined by analysing viability of treated and untreated propagules. In general, efficacy of biofumigation increased with decreasing burial depth and increasing B. juncea dose. Biofumigation was highly effective (mortality >85%) against small‐seeded species but less effective (mortality 0%–20%) against hard‐seeded and large‐seeded species at 200 t ha^?1. Vegetative propagules of Sonchus arvensis, Equisetum arvense and Convolvulus sepium were highly sensitive (mortality >90%) to biofumigation. Efficacy was most pronounced under moist warm incubation conditions, in the presence of a plastic ground cover.     

Seed retention of ten common weed species at oat harvest reveals the potential for harvest weed seed control

We assessed the seed production and shedding pattern of 10 common weed species in two oat fields in Denmark. The aim was to evaluate the possibility of harvesting retained seeds on weeds at crop harvest by a combine harvester based on estimation of weed seed retention. Before flowering, ten plants of each weed species were selected and surrounded by a seed trap comprising of a porous net. When the plants started shedding seeds, the seeds were collected from the traps and counted weekly until oat harvest. Just before oat harvest, the retained seeds on the plants were counted and the ratio of harvestable seeds and shed seeds during the growing season was determined. The seed production and shedding patterns varied between the 2 years. Across both years, Anagallis arvensis L., Capsella bursa-pastoris L. Medik, Chenopodium album L., Geranium molle L., Persicaria maculosa Gray, Polygonum aviculare L., Silene noctiflora L., Sonchus arvensis L., Veronica persica Poir. and Viola arvensis Murray retained on average 61.6%, 52.7%, 67.2%, 58.4%, 32.05%, 59.5%, 95.7%, 23.5%, 51.7% and 33.9%, respectively, of their produced seeds at crop harvest. Silene noctiflora was classified as a good target for harvest weed seed control; C. bursa-pastoris, C. album, G. molle, P. maculosa, S. arvensis and V. arvensis were classified as intermediate targets; and A. arvensis, P. aviculare and V. persica were classified as poor targets.     

Predispersal seed predation of Amaranthus retroflexus and Chenopodium album growing in soyabean fields

Field experiments were made in 1998 and 1999 to determine the influence of tillage and soyabean (Glycine max) row width on predispersal weed seed predation in Amaranthus retroflexus L. (redroot pigweed) and Chenopodium album L. (common lambsquarters). Soyabean was planted in wide (76 cm) and narrow (19 cm) rows with conventional or conservation tillage. Additional control plots without soyabean were also established. The two objectives were to determine (1) whether predispersal seed predation occurs in A. retroflexus or C. album, and (2) whether disturbance (soil tillage) or microclimate (planting pattern) influence predation level. Mean rates of seed predation were 26% and 4% in A. retroflexus and C. album, respectively. Although these levels were low at the population level, individual plants of both species had predation levels ranging from 0% to 80%, however, very few individuals of C. album had levels of predation above 10%. Differences among tillage and row width treatments occurred for A. retroflexus, but not for C. album. Amaranthus retroflexus and C. album growing within the soyabean crop received less light than those in the no‐crop plots, and produced less above‐ground biomass, smaller terminal inflorescences, and fewer seeds per inflorescence. Plant height, terminal inflorescence weight, and total seeds were correlated with predation in both weed species.     

Ecophysiological study on weed seed banks and weeds in Cambodian paddy fields with contrasting water availability

Weed infestations are a major cause of yield reduction in rice (Oryza sativa) cultivation, particularly with direct‐seeding methods, but the relationship between weed dynamics and water availability in Cambodian paddy fields has not been documented previously. We surveyed the weed abundance and weed seed banks in the soil of paddy fields with inferred differences in their water regime in 22 farm fields in three provinces of Cambodia in the 2005 and 2006 rainy seasons. We studied rain‐fed lowland fields in upslope and downslope topographic positions and fields at different distances from the irrigation water source inside an irrigation rehabilitation area. The weed seed banks were estimated by seedling emergence in small containers and weed abundance and vigor were estimated by a simple scoring system. The estimated weed seed bank in the top 5 cm of soil ranged from 52.1 to 167 × 10³ seeds m^?2 (overall mean of 8.5 × 10³ seeds m^?2) and contained a high proportion (86%) of sedge species, such as Fimbristylis miliacea L. and Cyperus difformis. Several fields had particularly large seed banks, including one near the reservoir. No clear difference was found in the weed seed banks between the irrigated fields that were located close to (upstream) and distant from (downstream) the water source or between the irrigated and rain‐fed lowland fields, but the weed scores were larger in the rain‐fed fields and the downstream fields within the irrigated area. A water shortage during the late growing season in 2005 led to a proliferation of weeds in some fields and an associated increase in weed seedbank size in 2006. However, the weed scores in 2006 were more strongly associated with that year's water conditions than with the weed seedbank size.     

Factors affecting seed germination and emergence of button grass (Dactyloctenium radulans) (R.Br.) P.Beauv.

Button grass (Dactyloctenium radulans) is a native, widely spread summer grass weed species in Australia. However, limited information is available on the seed germination biology of this species. Experiments were conducted to evaluate the effect of environmental factors on the germination and emergence of two populations of D. radulans. The seeds of these populations were collected separately from Dalby, Queensland and Coleambally, NSW. Seeds were germinated at a range of constant and alternating temperatures (25/15, 30/20, 35/25 and 40/30°C day/night). The greatest seed germination was at a constant temperature of 30°C. Seed germination was reduced at the lowest alternating temperature (25/15°C). Germination of both populations was strongly stimulated by light, suggesting a great amount of emergence of D. radulans on bare ground, such as crop seedbeds. Germination of the D. radulans population collected from a northern cotton farming system (Dalby) was more tolerant to a greater range of salt stress than the population sourced from the south (Coleambally). Seeds of both populations germinated over a wide range of pH, between 4 and 10. However, germination was the greatest in a high pH buffer solution, indicating that the species prefers to germinate in alkaline soil. These results showed that D. radulans seeds possess a wide range of tolerance mechanisms to different environmental stresses. Information obtained in this study will help in developing more sustainable and effective integrated weed management strategies for the control of this weed and weeds with similar responses in summer cropping systems, such as cotton.     

Comparative seed predation of woolly cupgrass (Eriochloa villosa) and yellow foxtail (Setaria pumila) along a field border in Canada

The seed predation of woolly cupgrass (Eriochloa villosa), an invasive weed of East‐Asian origin, was compared to that of yellow foxtail (Setaria pumila), an established and widespread weed, along a field border in Canada. The seeds of both species were glued to sandpaper cards and their removal in field conditions was recorded from July to September during the 3 years of 2009, 2010 and 2011. Predator exclusion cages allowed the evaluation of seed loss to be attributed to invertebrates, vertebrates and total seed loss due to predation for both species. Pitfall traps were installed and collected once in 2010 and during all the seed sampling dates in 2011 in order to evaluate predatory pressure from ground beetles and crickets. The average amount of seed loss that could be attributed to invertebrates was higher for S. pumila than for E. villosa, while the amount of seed loss that could be attributed to vertebrates was higher for E. villosa than for S. pumila in 2009 and 2010. The level of S. pumila seed loss to invertebrates was correlated with the sum of crickets (Gryllus pennsylvanicus) and beetles (Harpalus spp. and Ophonus rufibarbis) trapped during the 2011 season. During the same period, correlations between E. villosa seed loss to invertebrates and trapped insects were not observed. Reduced postdispersal seed predation by invertebrates could increase the local fall survival of the unburied seeds of invasive E. villosa populations, compared to other naturalized, smaller‐seeded weedy grasses.     

Digitaria sanguinalis seed dormancy release and seedling emergence are affected by crop canopy and stubble

Digitaria sanguinalis is a troublesome annual weed that causes important yield losses in different crops. Despite this, there is scarce information about different aspects of its biology under field conditions. New knowledge about the establishment process of this species will be of paramount importance in order to maximise the effectiveness of weed management. The aims of this paper were to evaluate the effect of stubble found on the surface on seed dormancy levels through the season, the effects of stubble and soyabean crop canopy on seedling emergence and to determine the field emergence pattern as a consequence of seed dormancy level at dispersal time. Seeds on the soil surface, which showed a high dormancy level at the beginning of autumn, were released from dormancy by low winter temperatures and germinated during spring as temperatures rose, showing a transient surface seedbank. Seeds covered by stubbles had delayed the emergence in the field due to lower alternating temperatures perceived by the surface seedbank. On the other hand, the presence of a soyabean crop and stubble together reduced the number of seedlings. Seeds with a high dormancy level at dispersal time showed a delayed emergence in the next season when compared with seeds with a lower dormancy level. However, the final number of seedlings was similar. Both stubble on surface and crop canopy are useful factors to lessen and delay the seedling emergence allowing the design of weed management strategies in order to diminish the population levels of this species.     

The demography of Datura ferox (L.) in soybean crops

The emergence, survival, seed production and seed dispersal of Datura ferox was studied in soybean fields in 1982–1983 and 1984–1985. Most seedling emergence occurred within one month of crop sowing. Later germination, associated with inter-row cultivation, represented 4–26% of the total seedlings, and none survived to seed production. Only 5% of the first cohort in 1982–1983, and 7% in 1984–1985, survived to seed production; it was independent of initial density. Although weed density was greatly reduced by control measures, negative relationships were observed between peak seedling density and plant height, stem diameter, number of ramifications, number of reproductive structures and seed production per plant. Only a small proportion of seeds (about 1%) were shed prior to soybean harvest. Combine harvesters collected more than 90% of capsules, but between 7% and 40% of the seeds were returned to the field. Seed viability was unaffected by passing through the machine. The patterns of seed dispersal varied depending on the design of the combine harvester. Two models shed seeds between 0 m and 21 m from their source, but another shed seeds between 0 m and 98 m. Calculations, based in life history parameters, showed that weed seed production would increase more rapidly if the seeds were dispersed during crop harvesting than if they arc not, even when the return of seeds to the soil by the combine is not large.     

The effects of seed mix and management on the abundance of desirable and pernicious unsown species in arable buffer strip communities

Buffer strips are refuges for a variety of plants providing resources, such as pollen, nectar and seeds, for higher trophic levels, including invertebrates, mammals and birds. Margins can also harbour plant species that are potentially injurious to the adjacent arable crop (undesirable species). Sowing perennial species in non‐cropped buffer strips can reduce weed incidence, but limits the abundance of annuals with the potential to support wider biodiversity (desirable species). We investigated the responses of unsown plant species present in buffer strips established with three different seed mixes managed annually with three contrasting management regimes (cutting, sward scarification and selective graminicide). Sward scarification had the strongest influence on the unsown desirable (e.g. Sonchus spp.) and unsown pernicious (e.g. Elytrigia repens) species, and was generally associated with higher cover values of these species. However, abundances of several desirable weed species, in particular Poa annua, were not promoted by scarification. The treatments of cutting and graminicide tended to have negative impacts on the unsown species, except for Cirsium vulgare, which increased with graminicide application. Differences in unsown species cover between seed mixes were minimal, although the grass‐only mix was more susceptible to establishment by C. vulgare and Galium aparine than the two grass and forb mixes. Annual scarification can enable desirable annuals and sown perennials to co‐exist, however, this practice can also promote pernicious species, and so is unlikely to be widely adopted as a management tool in its current form.     

A survey of weed seed contamination of rice paddy in Cambodia

Cambodia has experienced a rapid shift from transplanted to hand broadcast seeded rice, with a consequent increase in seeding rates from 25–30 to 100–200 kg ha^?1. To reduce costs, farmers keep their own seed for sowing with the risk of greater weed seed contamination of the sowing seed. A survey of weed seed contamination in harvested rice paddy was conducted in two provinces of Cambodia (Battambang and Takeo) at the end of the wet season in 2016. Farmers were interviewed about rice‐seeding practices, and a total of 110 farmers' fresh paddy samples were inspected for weed seed contamination from the two provinces. Sowing seed samples collected from 28 seed producer lots and 71 samples of farmer‐kept seed were also analysed for weed seed contamination. In both provinces, the majority of farmers kept their own seed or bought seed from a neighbour. Farm‐kept seed for sowing accounted for 88% of sown seed in Battambang and 89% in Takeo. Seeds of 41 different weed species from 13 plant families were found in the farmers' freshly harvested paddy samples. Overall, farmers managed to reduce the number of weed propagules by 60% and seed producers by 95%. There was no significant difference between farmer‐kept seed and seed producer/seed company seed for the total number of weed seeds present. When shown photos, farmers' rankings of the 10 most common weed species found in freshly harvested paddy did not closely correspond to the actual weed seed frequency in the paddy. When farmers were asked to rank the frequency of weeds in their fields without the option to choose from a list, they ranked the weeds differently. Farmers ranked Ischaemum rugosum, Echinochloa spp. and Fimbristylis miliacea as the three most frequent weed species in their fields. The most frequent weeds in harvested paddy, apart from weedy rice, were I. rugosum and Melochia corchorifolia. Farmers did not rank M. corchorifolia as a frequently occurring weed, and most farmers could not recognise M. corchorifolia from photographs. The priority for improved seed hygiene is to place the emphasis on assisting farmers to further improve their seed purification techniques and to caution them to inspect seed before purchasing from neighbours, seed producers and seed companies in the absence of the implementation of seed certification regulation.