The degradation of imazapyr, fiumetsulam and thifensulfuron applied at 500, 40 and . M) g active ingredient (a.i,) ha"', respectively, to si!t loam soil was sttidied under laboratory and field conditions. Herbicide residues were analysed by a lentil {Lem culinaris L.) bioassay. Results showed ihat temperature had a significant effect on herbicide degradation. Vk'hereas the impact of soil organic matter ami pH were less well defined. Half-lives for iniazapyr, flumetsulam and thifensulfuron in soil samples from the 0-5 cm layer (64% organic carbon) at 15 "C were 125, 88 and 5-4 days, respectively, and 69, 30 and .^-9 days at 30 °C. ]n soil sampled from the 15-20 cm layer (3-5% organic carbon) half-lives were 155, 70 and 6-4 days, respectively, at 15 "C and 77, 24 and 4-8 days at 30 °C. A field experiment investigated the degradation and teaching of each herbicide under two precipitation regimes (natural precipitation (208 mm), and nattjral precipitation pJus 75 mm irrigation (283 mm)] over 4 months to a soil depth of 25 cm. Thifensiitfuron degraded rapidiy, whereas residues of flumetsulam and imazapyr leached below 25 cm in both the low-and highprecipitasion treatments after 4 months. Significant itnai^apyr residues were still present in the soil to 25 cm depth after 3 months. A miilt:component model for herbicide dissipation was developed and evaluated using data from the laboratory and fieid experiments. lntrottuction

A survey of weeds and weed management practices was conducted during 1997/98 in apple orchards and vineyards of Canterbury and Nelson/Marlborough regions of New Zealand. According to growers, mallow (Malva spp.) ranked as the most widespread weed in apple orchards of both regions. In vineyards, mallow was the most common weed in Nelson/Marlborough and the second most common weed in Canterbury. Other dominant weeds in Nelson/Marlborough included grasses, black nightshade (Solarium nigrum), fathen (Chenopodium album), and redroot (Amaranthus retroflexus). In Canterbury, grasses, clovers, fathen, and thistles were common. In almost all orchards knockdown herbicides were used but the use of residual herbicides varied. The total number of herbicide applications/year has increased in both regions over the past 10 years, but the proportional use of residual herbicides has declined. Some other weed management practices are used in orchards but a sound integration of these is commonly lacking.

Summary Three possible sources of field contamination by weed seeds were studied during 1983–86 in two areas of Fars Province, Iran. These sources were crop seed, irrigation water and sheep manure. Manure was found to be the most important, adding almost 10 million seeds ha−1 at each application. Farmer‐saved seed of wheat, the main crop of the area, added an average of 182,000 weed seeds ha−1. Irrigation water added no more than 120 seeds ha−1, but was able to carry seeds over long distances without affecting viability.

Field trials were undertaken over three years to determine the effect of tine weeding in organic pea (Pisum sativum) crops in Canterbury The tine weeding treatments were performed either as single or repeated passes at different times of crop growth Significant weed control and yield increase were achieved with a single pass of tine weeding either at preemergence or at the 2leaf stage of the crop On some occasions two passes of tine were required to control new weeds Late postemergence tine weeding did not have a positive effect on pea yield

Ripgut brome (Bromus diandrus), prairie grass (B. willdenowii) and soft brome (B. hordeaceus) are becoming serious weed problems in arable farms of New Zealand. This paper reports results from three years of field experiments evaluating chemical treatments for brome control in cereal crops. Ripgut brome was shown to be very competitive, with moderate infestations reducing grain yields by 25-30%. A strong relationship (R 2 =0.96) was found between seedling numbers of ripgut brome and final grain yield. The study found several promising chemical treatments that offer a good degree of control. These include preemergence applications of cyanazine + terbuthylazine, chlorsulfuron + terbuthylazine, and metribuzin. The same herbicides applied postemergence were not as effective for brome grass control.

Summary: The effect of degree of water stress in Avena sativa on diclofop‐methyl efficacy was assessed. Within 24 h of applying stress by adding mannitol to the root medium (0 to 12.5%), the rate of leaf extension of the youngest leaves (leaves 3 and 4) decreased with increasing mannitol concentration. Without water stress, application of diclofop‐methyl had little effect on extension rate of leaves 3 and 4 during the first 4 days after spraying. Subsequently, it caused a significant decrease in the extension rate of leaf 4 with more pronounced effects on later leaves. Diclofop‐methyl had little effect on leaf extension rate of plants given mannitol. Shoot dry weight at harvest for unsprayed plants decreased with increased mannitol concentration and for sprayed plants it was greater without mannitol than with mannitol (all levels). However, at mannitol concentrations greater than 4%, shoot dry weight for sprayed and unsprayed plants was not significantly different. Sprayed plants exposed to 2–8.5% mannitol produced seed heads but those at zero mannitol did not. When the root medium of all treatments was flushed of mannitol one week after spraying, then main‐tained without mannitol, shoot dry weight at harvest for unsprayed plants decreased slightly with increased mannitol concentration applied initially. However, shoot dry weight for sprayed plants increased with increased mannitol concentration. Without mannitol two weeks after spraying, chlorophyll concentrations of leaves 3 and 4 were greater and water saturation deficit (WSD) values were lower for unsprayed plants than sprayed plants but there were no differences at 6.2% mannitol. It is proposed that tolerance to diclofop‐methyl by A. sativa as a result of water stress, is primarily due to a decreased rate of leaf expansion resulting in lower demand for membrane synthesis and less strain on membranes damaged by the herbicide.

Avena sativa cv. Amuri fed either low or high nitrate was sprayed with diclofopmethyl ( 1 kg a.i. ha-I) at the three leaf stage. The short term effects of the herbicide on chlorophyll concentration of leaves (laminae) and short and long term effects on d.wt of the component plant parts were determined by comparison with unsprayed plants.For unsprayed and sprayed plants, total leaf d.wt approximately doubled during the first twelve days after commencing treatments. Growth was substantially greater at high nitrate than low nitrate. For unsprayed plants, the increase in total leaf d.wt was due primarily to growth of leaf 3 but for sprayed plants it was due to growth of leaves 1 and 2. Twelve days after commencing treatments, d.wt of leaves 1 and 2 was substantially greater for sprayed plants than for unsprayed plants given similar nitrate, while chlorophyll concentration was substantially less. Leaf 3 d.wt and chlorophyll concentration were substantially greater in unsprayed plants than in sprayed plants given similar nitrate. For unsprayed plants, values were greater at high nitrate than low nitrate, for sprayed plants the converse was the case.Forty nine days after commencing treatments, unsprayed plants had a greater total plant d.wt than sprayed plants given similar nitrate. Total plant d.wt for unsprayed plants was greater at high nitrate than low nitrate, the opposite was the case for sprayed plants. Unsprayed plants a t both nitrate levels and sprayed plants given low nitrate produced seed heads but sprayed plants given high nitrate did not. Diclofop-methyl at a rate of 0.3 kg a.i. ha-' stopped seed head production at high nitrate. Retention and uptake of diclofop-methyl were not significantly different at low and high nitrate. At 1 kg a.i. ha-l diclofop-methyl, plants switched from low to high nitrate at spraying showed damage similar to that shown by plants given high nitrate throughout. Addition of 200 vg GA into the leaf sheaths two days prior to spraying increased the efficacy of diclofop-methyl at low nitrate.It is proposed that increased efficiency of diclofop-methyl at high nitrate is due to increased leaf damage caused by a greater rate of leaf expansion.

Summary: Résumé: Zusammenfassung Wheat (Triticum aestivum L.) cultivars showed differential tolerance to chlorsulfuron. Cultivar Kotare showed no injury symptoms following foliar applications of chlorsulfuron at 15 or 60 g a.i. ha−1, while cultivars Rongotea and Lancer showed early damage in pot and field experiments at both rates of chlorsulfuron. Cultivars Abele and Jasper were intermediate in their response. The number of spikelets per ear was the only yield component affected by chlorsulfuron and was reduced in Lancer and Rongotea. Retention, uptake and transiocation of chlorsulfuron were not different between Kotare and Ron‐gotea. Within 48 h of application, Kotare metabolized 92–2% of [14C]chlorsulfuron, while Lancer and Rongotea metabolized only 43–5% and 63% of the herbicide, respectively. The concentration of chlorsulfuron in young tissues of Kotare, Lancer and Rongotea, 48 h after application was calculated as 1.2, 31.9 and 15.6 ng g−1 dry weight, respectively. It is concluded that differential rates of metabolism are the main reason for differences in sensitivity to chlorsulfuron between the wheat cultivars tested. Le mécanisme des différences de tolérance au chlorsulfuron entre variétés de blé Des variétés de blé (Triticum aestivum L.) ont montré des différences de tolérance au chlorsul‐furon. On n'a aucun symptôme de phytotoxicité sur la variété‘Kotare’ après une application foliaire de chlorsulfuron à 15 ou 60 g m.a. ha−1, alors que les variétés ‘Rongotea’ et ‘Lancer’ présentaient des symptômes précoces aux deux doses, dans des expériences en pot et au champ. La réponse des variété‘Abele’ et ‘Jasper’était intermédiaire. Le nombre de grains par épillet était la seule composante du rendement affectée parle chlorsulfuron et était réduit chez Lancer et Rongotea. La rétention, la pénétration et la migration de chlorsulfuron n'était pas différentes entre Kotare et Rongotea. Quarante huit heures après le traitement, Kotare avait metabolise 92,2% du [14C]chlorsulfuron, alors que Lancer et Rongotea ne métabolisaient que respectivement 43,5 et 63% de l'herbicide. La concentration de chlorsulfuron dans les tissus jeunes de Kotare, Lancer et Rongotea 48 h après la traitement a étéévaluée à respectivement 1,2,31,9 et 15,6 ng g−1 de matière sèche. Il est conclu que des différences de vitesse de métabolisation sont la raison principale des différences de sensibilité au chlorsulfuron observées chez les variétés de blé testées. Mechanismen der unterschiedlichen Reaktion von Weizensorten auf Chlorsulfuron‐Behandlun‐gen Es wurde eine unterschiedliche Toleranz bei Weizensorten (Triticum aestivum L.) gegenüber Chlorsulfuron‐Behandlungen beobachtet. Bei der Sorte ‘Kotare’ traten nach Behandlungen mit 15 oder 60 g AS ha keine Schadsymtome auf, während die Sorten ‘Rongotea’ und ‘Lancer’ sowohl in Topfals auch in Freilandversuchen bei beiden Dosen frühe Schädigungen erlitten; die Sorten ‘Abel’ und ‘Jasper’ reagierten mittelstark. Unter den Ertragskomponenten war nur die Zahl der Ährchen pro Ähre betroffen, sie war bei ‘...