Thesis (MScAgric)--Stellenbosch University, 2011. / ENGLISH ABSTRACT: Glyphosate remains an important herbicide in weed control. This is due to several positive attributes it has including sytemicity, wide spectrum of weed control and environmental friendliness. Its efficacy and lack of residual activity are therefore important to ensure adequate weed control without imposing hazards to the environment. Despite these favourable attributes for weed control glyphosate has its shortcomings. Evolution of resistance to glyphosate has been a major concern from 1996. However, there are other factors that reduce the maximum potential of glyphosate. Any factors that reduce glyphosate efficacy may result in the target plant being subjected to non-lethal concentrations of glyphosate. This in turn may predispose the plants to developing herbicide resistance. Some factors that may influence efficacy of glyphosate, and therefore be possibly selecting for resistance were investigated in this study. Although glyphosate is a postemergence herbicide, its efficacy is not exempt from the effect of soil and nutrients in which the weeds occur. The possibility of this occurring was investigated in a greenhouse study on ryegrass (Lolium spp.) In this study ryegrass was grown in three soils: pure sand (SS), soil from pasture paddock (PS) and soil from crop field (CS). The soils varied in nutrient composition and, although all were classified as sand, they had varying proportions of sand, loam and clay. This investigation consisted of four experiments. The first experiment was investigating the effect of growing a susceptible commercial ryegrass cultivar on PS, CS and SS soils on the efficacy of glyphosate (360 g a.i. L-1 formulation) applied at five glyphosate application rates (GAR). The GARs were 0 (0x), 67.5 (1/8x), 135 (1/4x), 270 (1/2x) and 540 (1x) g a.i. ha-1. The second experiment investigated the effect of growing a susceptible commercial ryegrass cultivar and a glyphosate resistant ryegrass biotype on PS and CS soils on the efficacy of glyphosate. The application rates were 0 (0x), 270 (1/2x), 540 (1x), 1080 (2x) and 2160 (4x) g a.i. ha-1. The third and the fourth experiments were similar to the first experiment except: The latter investigated the role of nutrient content of irrigation water (pure water or balanced nutrient solution) and; the former investigated the effect of soil activity (by covering the soil surface with cotton at the time of spraying) of glyphosate with regard to the role it plays on efficacy of glyphosate. Our findings showed that: i) soil affects the efficacy of glyphosate with more control (19% survivors) found in the PS soil compared to 50% and 62% survivors in CS and SS soils respectively, this effect may be dependent upon the species resistance as; ii) the effect in the resistant ryegrass biotype was reversed with about 95% of survivors in the PS soil compared to about 78% in CS soil; iii) efficacy of glyphosate is influenced by the soil nutrient status and the nutrient content of the irrigation water. This was shown by decrease in the control of ryegrass (100% survivors) grown in SS soil when fed with pure water compared to 45% when nutrient fed. In PS soil there was no significant effect. This was probably due to inherently higher nutrient content of the PS soil; and iv) glyphosate efficacy is influenced by the amount of glyphosate reaching the soil (absorbed through the roots). This was shown in PS soil where 1/8x GAR resulted in 93.3% survivors in covered soil compared to 60% in uncovered soil. A similar trend was also observed at 1/4x GAR. An opposite effect was shown in SS soil with 0% and 40% survival at 1/4x GAR in the covered and uncovered soil respectively. Glyphosate has been hailed as an environmentally friendly herbicide as it rapidly degrades in soil and it sorbs on metals embedded in soil matrix. However, reports in the literature have showed reduction in crop yield due to soil glyphosate residues. In these studies, glyphosate phytotoxicity was found to be dependent on certain soil characteristics and nutrient content. Following this, a greenhouse study was conducted to assess the phytotoxic activity of glyphosate on a susceptible commercial ryegrass cultivar grown in PS, CS and SS soils. Glyphosate was applied at 0 (G1), 540 (G2) and 3240 (G3) g a.i. ha-1. Ryegrass seedlings of comparable size were transplanted into the soil at intervals of two hours, three weeks and four weeks after glyphosate application referred to as TAS1, TAS2 and TAS3 respectively. Evidence of soil glyphosate activity was shown by the decrease in percentage survival with the application of glyphosate. This was significant in the SS soil where about 60% and 48% survival in G1 and G2 GAR respectively was observed compared to about 100% in the untreated control when transplanted three weeks after glyphosate application. The decrease in percentage survival was time mediated with significant effect of G2 GAR shown at TAS 1 whereas at G3 GAR the effect was significant at TAS1 and TAS 2. At TAS 3 there was no effect at all GARs. Similar trends were observed with dry mass and shoot length. Trace metals required for normal plant growth have been implicated in the reduction of glyphosate efficacy. This follows glyphosate’s original development as a metal chelator. Glyphosate-trace metal antagonism has recently sparked interest following co-application in glyphosate resistant soybeans. Molybdenum (Mo), an anion, may play a role at the physiological level on the antagonism of glyphosate. A greenhouse assay was carried out where seedlings grown from seeds (of susceptible commercial ryegrass cultivar (S biotype) and glyphosate resistant biotype (R biotype)) were grown with nutrient solutions containing 0x, 1x and 2x molybdenum (Mo) concentrations where 1x is 0.05 mg L-1 Mo. Glyphosate was applied at 0 (0x), 135 (1/4x), 270 (1/2x), 540 (1x) and 1040 (2x) g a.i. ha-1 rates. In the R biotype applying 2x Mo resulted in 0% survival in the R biotype at 1x GAR compared to 50% and 90% survival at the same GAR with 0x and 1x Mo. In terms of dry mass and shoot length the results did not show any conclusive trends. / AFRIKAANSE OPSOMMING: Glifosaat is ‘n baie belangrike onkruiddoder wat in verskeie onkruidbeheerstelsels gebruik word. Dit is as gevolg van verskeie positiewe eienskappe waaroor dit besit onder andere sistemiese werking, wye spektrum van onkruidbeheer en omgewingsvriendelike werking. Die onkruiddoder se effektiwiteit en gebrek aan residuele aktiwiteit is dus belangrik om bevredigende onkruidbeheer te bewerkstellig sonder om skade aan die omgewing te veroorsaak. Ten spyte van hierdie voordelige eienskappe het glifosaat ook tekortkominge. Vanaf 1996 is ontwikkeling van weerstand teen glifosaat in onkruide ‘n groot bron van kommer. Daar is egter ook ander faktore wat die maksimum potensiaal van glifosaat strem. Enige faktore wat die effektiwiteit van glifosaat strem mag veroorsaak dat die teikenplant aan subletale dosisse van glifosaat blootgestel word. Dit mag weer daartoe lei dat sulke plante blootgestel word aan die ontwikkeling van weerstand. Sommige faktore wat die effektiwiteit van glifosaat mag strem en dus moontlik kan lei tot seleksie vir weerstand is in hierdie studie ondersoek. Alhoewel glifosaat ‘n na-opkoms middel is kan die effektiwiteit moontlik beïnvloed word deur grondfaktore en nutriënte. Hierdie moontlikheid is ondersoek in ‘n glashuisstudie waarin raaigras (Lolium spp.) gebruik is. In hierdie studie is raaigras geplant in drie verskillende grondsoorte nl. suiwer sand (SS), grond vanaf ‘n weidingskamp (PS) en grond vanaf ‘n gewasland (CS). Die gronde het gevarieer in terme van nutriëntinhoud en alhoewel al drie gronde as sand geklassifiseer is, was daar tog verskille in fisiese eienskappe. Hierdie afdeling het uit vier eksperimente bestaan. In die eerste eksperiment is die invloed van verskillende gronde (PS, CS en SS) waarin die kommersiële raaigras kultivar geplant was op die effektiwiteit van glifosaat (360 g a.b. L-1 formulasie) teen vyf verskillende dosisse (GAR) ondersoek. Die dosisse was 0 (0x), 67.5 (1/8x), 135 (1/4x), 270 (1/2x) and 540 (1x) g a.b. ha-1. Die tweede eksperiment het die invloed van twee grondsoorte (PS en CS) waarin ‘n vatbare kommersiële raaigras kultivar en ‘n glifosaat weerstandbiedende raaigras biotipe geplant is, se invloed op die effektiwiteit van glifosaat ondersoek. Die dosisse was 0 (0x), 270 (1/2x), 540 (1x), 1080 (2x) and 2160 (4x) g a.b. ha-1. Die derde en vierde eksperimente was soortgelyk aan die eerste eksperiment behalwe dat die derde eksperiment die invloed van voeding (suiwer gedistileerde water teenoor ‘n gebalanseerde voedingsoplossing) saam met besproeiing ondersoek het. In die vierde eksperiment is die grondwerking van glifosaat wat as blaarbespuiting toegedien is ondersoek deurdat sommige potte wat gespuit is se oppervlakte met ‘n laag watte bedek is tydens die spuitproses en dadelik na spuit verwyder is teenoor die ander behandeling waar die grondoppervlakte nie bedek is nie. Die resultate het getoon dat i) grondtipe die effektiwiteit van glifosaat beïnvloed met beter beheer (19% oorlewing) in die PS grond vergeleke met 50% en 62% oorlewing in die CS en SS grondtipes respektiewelik. Hierdie effek kan moontlik beïnvloed word deur die weerstandsvlak van spesies omdat ii) die effek in die weerstandbiedende biotipe omgekeer is met ongeveer 95% oorlewing in the PS grondtipe vergeleke met 78% in die CS grondtipe; iii) effektiwiteit van glifosaat is beïnvloed deur die voedingstatus van die grond en die besproeiingswater. Dit word aangedui deur die afname in beheer van raaigras (100% oorlewing) wat in SS grond gegroei het en met suiwer gedistilleerde water besproei is vergeleke met 45% oorlewing in dieselfde grond wanneer met ‘n gebalanseerde voedingsoplossing besproei is. Plante wat in PS grond gegroei het het geen betekenisvolle verskille tussen die besproeiingsbehandelings getoon in hulle reaksie op glifosaattoediening nie, waarskynlik as gevolg van die inherente hoër nutriëntinhoud van die grond en iv) glifosaat effektiwiteit word beïnvloed deur die hoeveelheid glifosaat wat die grond bereik en deur die wortels opgeneem word. Dit is bewys in plante wat in PS grond gegroei het waar 93.3% plante oorleef het waar die grondoppervlakte bedek was teenoor 60% oorlewendes waar die grondoppervlakte nie bedek was nie indien glifosaat teen 1/8x toegedien is. ‘n Soortgelyke tendens is by die 1/4x dosis waargeneem. ‘n Teenoorgestelde effek is in SS grond waargeneem waar die oorlewingspersentasie in bedekte en onbedekte grond by 1/4x glifosaatdosis 0% en 40% onderskeidelik was. Glifosaat is aanvanklik aangeprys as ‘n omgewingsvriendelike onkruiddoder omdat dit vinnig in grond afgebreek word en omdat dit geadsorbeer word aan metale in die grondmatriks. In teenstelling hiermee is daar egter verslae in die literatuur wat dui daarop dat glifosaatresidue in die grond gewasopbrengste kan verlaag. In die gemelde studies is gevind dat fitotoksisiteit van glifosaat residue afhang van grondeienskappe en grondvrugbaarheid. Na aanleiding hiervan is ‘n glashuisstudie uitgevoer waarin die fitotoksisiteit van residuele glifosaat op ‘n kommersiële raaigraskultivar wat in PS, CS en SS gronde groei, ondersoek is. Glifosaat is op die grond in potte toegedien teen 0 (G1), 540 (G2) and 3240 (G3) g a.b. ha-1. Raaigras saailinge is daarna in die potte ingeplant twee ure, drie weke en vier weke nadat die glifosaat toegedien is. Bewys van grondaktiwiteit van glifosaat is gelewer deur die vermindering in persentasie oorlewing van die saailinge met toediening van glifosaat. Die vermindering in oorlewing was betekenisvol in die SS grond waar ongeveer 60% en 48% oorlewing van saailinge was by G2 en G3 dosisse onderskeidelik teenoor 100% oorlewing in die onbehandelde kontrole. Die afname in persentasie oorlewing is deur tyd beïnvloed deurdat die G2 dosis slegs by die twee ure behandeling betekenisvolle verlaging in oorlewing veroorsaak het terwyl die G3 dosis by die twee ure sowel as die drie weke behandeling betekenisvolle verlagings veroorsaak het. Vier weke na toediening was daar geen effek van glifosaat op die saailinge in enige van die gronde gewees nie. Soortgelyke tendense is waargeneem by die droëmassa en lengte data. Spoorelemente wat noodsaaklik is vir normale plantgroeiprosesse is al geïmpliseer in verlaging van glifosaat effektiwiteit. Dit is waarskynlik omdat glifosaat oorspronklik ontwikkel is as ‘n metaal cheleerder. Glifosaat-spoorelement antagonisme was onlangs in die nuus nadat glifosaat saam met sulke elemente toegedien is op glifosaat weerstandbiedende sojabone. Molibdeen (Mo), ’n anioon, mag ’n rol op fisiologiese vlak speel in doie effektiwiteit van glifosaat. ‘n Glashuisstudie is uitgevoer waarin saailinge van ‘n glifosaat vatbare kommersiële raaigras kultivar en ‘n glifosaat weerstandbiedende raaigras biotipe besproei is met voedingsmengsels wat 0x, 1x en 2x Mo bevat waar 1x 0.05 mg L-1 Mo is. Glifosaat is op die plante toegedien teen 0 (0x), 135 (1/4x), 270 (1/2x), 540 (1x) and 1040 (2x) g a.b. ha-1 dosisse. Die 2x Mo toediening het gelei tot 0% oorlewing in die R biotipe by 1x GAR vergeleke met 50% en 90% oorlewing by dieselfde GAR met 0x en 1x Mo. In terme van droëmateriaal en lengtegroei was daar geen konkrete tendense nie.

Amphibian populations are in mass decline on a global scale. Various explanations have been considered, including harmful effects from exposure to toxicants. Using Xenopus laevis adults and tadpoles, potential sublethal effects of atrazine, a herbicide, were investigated in this thesis. I also investigated the toxicity of an organic herbicide compared this with the toxicity of a synthetic herbicide, using LC50 values. Whether X. laevis adult frogs could communicate chemically was tested experimentally. The results suggest that adult female X. laevis communicate chemically, but there was no evidence that male individuals did so. For testing tadpoles I used a kin-preference assay. An encouraging trend for kin preference was evident, for both an outbred and an inbred line. Tadpoles changed their behaviour after exposure to l0μg/L of atrazine for 24 hr. Kin preferences in the control tests were reversed after exposure. A hypothesis of altruistic kin avoidance was suggested by these results. However, when individuals were isolated and then exposed, these individuals had more pronounced preference for kin compared to controls. X. laevis tadpoles exposed to Organic Interceptor (organic herbicide) had a LC50 that was more than 7000 times lower than the 20% recommended dose, whereas Roundup Renews' (a synthetic herbicide) LC50 was around 8 times lower than the 1% recommended dose. This research adds to evidence that toxicants have a negative impact on amphibian populations, and suggests that more research needs to be conducted to identify other sublethal effects of toxicants and to clarify the implications these effects might have for the amphibian populations in nature.

Two varieties of sideoats grama (Vaughn and NM-28) were planted in late summer of 1980. The seedlings were treated with a variety of broadleaf weed control treatments at 1-, 3-, 5-leaf, and tillering stages of grass seedling growth to determine the effectiveness of these treatments for controlling weeds and the appropriate stages of grass seedling growth to apply them and consequently on grass establishment, density, forage production and plant height. Four months after seeding, NM-28 stands were more dense, produced more forage and the plants were taller than Vaughn. Treatments of dicamba at 0.28 and 0.56 kg/ha were the most effective herbicidal treatments in controlling weeds and resulted in highest stand densities and yields when applied at the 1- and 3-leaf seedling growth. Both treatments gave over 90% control of weeds at the earlier stages but control decreased at later stages of treatment. Lower control resulted when 2,4-D at 0.56 and 1.12 kg/ha were applied. Even though hand weeding treatment resulted weed-free plots during the first four months of seedling development, it resulted in less stand densities and yields when compared to herbicidal treatments. One year after treatment, Vaughn established into taller stands and produced higher yields than NM-28. Dicamba 0.28 kg/ha and 2,4-D 0.56 kg/ha effectively aided in better stand density and yield. Treatments applied at earlier stages of growth resulted in better establishment of both varieties. Dicamba at 0.28 kg/ha was the most effective herbicidal treatment for both varieties. However, hand weeding evaluated one year after seeding, effectively increased both accessions' yield and was the most effective treatment. Greenhouse and laboratory experiments were conducted to determine the uptake and distribution of ('14)C-2,4-D in both sideoats grama accessions and palmer amaranth seedling plant parts at 1, 4, 8 and 24h after treatment. Over all harvest times, 25.3, 2.8 and 3.1% activity were recovered in Vaughn treated leaf, tillers and crown, respectively. That was compared to 33.9, .68 and 6.8% in the same NM-28 plant parts, respectively. Higher percentages of activity were recovered in palmer amaranth plant parts; 50.5% in treated leaf, 8.5% above treated leaf and 5.4% below.

Two experiments were conducted to determine the efficacy and safety of soil applied broadcast preemergent herbicides including Nortron, Prowl, Prefar, and Prefar + Prowl combinations applied at planting to fall seeded dry bulb onions. Crop stand reductions ranging from 10 to 33 percent resulted from Prowl 3.3EC use rates of 1.2 and 2.4 pt/acre. Prowl 3.3EC applied at 0.6 pt/acre, Prefar 4E applied at 4 and 6 qt/acre, Nortron 4SC applied at 2 pt/acre, and the combination of 0.6 pt Prowl 3.3 EC plus Prefar 4E were safe on fall seeded onions. Prefar plus Prowl combinations also provided winter weed control equal to or better than the standard 14 lb/acre Dacthal 75WP.

The herbicides used in lettuce have changed little in more than 30 years. Poast was registered for grass control in the 1980's although preemergent applications of Kerb, Balan and Prefar have been the principal herbicides used in lettuce production since the mid 1960's. Balan was changed from a 1.5 lb./gal. Emulsifiable concentrate to a 60% dry flowable formulation in the mid 80's, Kerb has always been a 50% wettable powder and Prefar is still a 4 lb./gal. emulsifiable concentrate. Growers are constantly changing cultural practices to improve production or to become more efficient. The change in one cultural practice can, and often does, effect other cultural practices. The use of sprinklers to establish lettuce has become increasingly widespread in the Yuma area over the past 20 years. Kerb and Prefar can be mechanically incorporated into shaped beds although both are commonly incorporated with irrigation water. The change in irrigation practices during stand establishment from furrow irrigation to sprinklers has effected the performance of both Kerb and Prefar. Balan is normally disced into the soil prior to bed formation and is not as effected by irrigation practices during stand establishment. Four tests are presented in this paper that help explain the effect of irrigation practices on the performance of Kerb and Prefar.

Onions treated with bromoxynil (Buctril7) or oxyfluorfen (Goal7) at the time when the first true leaf was emerging were not injured. No significant onion crop stand reduction occurred from any of the postemergence (POST) treatments. Onion height was not affected by any of the POST treatments through the season. A single application of Goal or Buctril offered up to 7 WAT of very good weed control with excellent crop safety. Onions treated at the typical 2-leaf stage of growth with Buctril or Goal exhibited no significant crop injury. Delayed and reduced control of knotweed (Polygonum aviculare) could have contributed to the decreased onion yield in the herbicide treated onions compared to the handweeded check. Onions in the untreated check were significantly reduced compared to Goal treated onions or the handweeded check.

At 4 weeks after treatment (WAT), all preemergence (PREE) treatments were completely safe on cantaloupes. At 1 WAT of postemergence (POST) applications, marginally acceptable melon injury (11 to 19%) was observed. At 6 WAT, crop injury increased significantly for both halosulfuron and bentazon. Halosulfuron (POST) following bensulide (PREE) caused minimal crop injury. The pigweeds were marginally controlled when POST treatments followed PREE herbicides. Tumble pigweed (Amaranthus albus) was more difficult to control than prostrate pigweed (A. blitoides). Halosulfuron gave good control of nutsedge (Cyperus rotundus) at 6 WAT.

The addition of an adjuvant, Agridex, to halosulfuron or Basagran7 (bentazon) did not increase crop injury significantly compared to treatments without Agridex. The addition of Agridex to halosulfuron slightly improved morningglory (Ipomoea hederacea) control compared to without the use of an adjuvant. Bentazon at 1.0 lb/A plus Agridex gave very good morningglory control at 92%. At 2 WAT on 20 Aug, cantaloupe injury decreased for halosulfuron and bentazon treatments. A second application of halosulfuron at 0.05 lb/A did not cause additional crop injury.

Most herbicides applied alone preemergence (PREE) caused minimal crop injury (<10%) when furrow irrigated. Pendimethalin (Prowl) applied at 0.5 lb AI/A plus bensulide (Prefar) at 6.0 lb AI/A under sprinkler irrigation in Tolleson caused crop injury that was highly unacceptable and the crop stand was severely reduced. Onion yields were significantly reduced for the Prowl plus Prefar treatments. Prowl at 0.25 or 0.5 lb AI/A alone provided very good (>90%) weed control of all weeds. Combination treatments of Prowl plus other herbicides provided very good weed control but did not offer enhanced control of weeds already controlled by Prowl alone. The combinations of ethofumesate (Nortron) with metolachlor (Dual) or dimethenamid (Frontier) gave improved weed control compared to when either was applied alone.

Pendimethalin (Prowl7) applied preemergence (PREE) at 0.25 to 0.50 lb AI/A caused no observable injury and did not affect yields of onions that were furrow irrigated . Prowl applied PREE at 0.50 lb AI/A caused significant crop stand and yield reduction compared to lower rates or the untreated check under sprinkler irrigation. Prowl applied preplant incorporated (PPI) at rates ranging from 0.25 to 0.75 lb AI/A did not significantly injure onions or cause a significant yield reduction. Combination treatments of Prowl plus bensulide (Prefar7) applied PREE did not cause any measurable crop height or stand reduction compared to the standard treatment or untreated check. Prowl at 0.25 lb AI/A plus Prefar at 4.0 lb AI/A adequately controlled cheeseweed, yellow sweetclover, sowthistle, and London rocket.