Thesis (MScAgric)--University of Stellenbosch, 2001. / ENGLISH ABSTRACT: The aim of this study was to determine the effect of a grass herbicide, a broadleaf herbicide with some grass control capabilities, method of tillage (tine and no-tillage) at planting of wheat as well as different medic cultivars on the regeneration, dry matter (OM) production and quality of a medic pasture. The trial was conducted at Langgewens experimental farm in the Swartland wheat producing area. Nine medic cultivars of three different species were evaluated after being sprayed with either haloxyfop-R-methyl (HAL) ester or imazamox (IMI) and subjected to either a tine tillage or a no tillage treatment at planting of wheat. Soil samples were taken during January 2000 to determine the size of the medic and weed seedbank as well as the degree of dormancy in the medic seeds, while OM samples were taken throughout the growing season to determine the OM production of the different medic cultivars and weed species. OM samples taken during October 1998 on the same pasture, were used to determine the crude protein (CP) and neutral detergent fibre (NOF) content of the pasture. The samples were subjected to in vitro digestion and the digestibility of pasture CP (OCP), NOF(ONOF) and DM (DOM)were determined. Results showed that seedling establishment differed between cultivars used, herbicide treatments applied as well as the crop stage in the rotation. The cultivars produced more seedlings where IMI was applied compared to HAL as well as where the area consisted of two year pasture compared to one year pasture (1998) and one year wheat (1999). After a year of pasture and a year of wheat, cultivars Sephi and Paraggio produced the most seedlings, while Caliph and Orion produced the least. Caliph however, showed a very high degree of seed dormancy while Orion's low seedling establishment was due to its sensitivity to the IMI herbicide used. Little difference was found between the nine cultivars early in the season (July - August) with regard to cumulative OM production, except for Orion, whose growth was severely damaged by the IMI treatment. At the end of the growing season (October), the cultivar Caliph's cumulative OM production (2010.1 kg/ha) was significantly higher than all the other cultivars, except for Parabinga (1053. 4 kg/ha). Oifferent pasture samples, of which the botanical composition was known, was analysed for CP, NOF, OOM, OCP and ONOF. There was no significant difference in pasture composition during 1998 but variation in the pasture composition did however cause the IMI treatment, compared to the HAL treatment, to have a lower ONOFand OOMcontent. A modelling procedure was used to predict the pasture quality parameters (CP, NOF, OOM,OCP and ONOF) from the pasture composition (medic hay, medic pods, grassy and broadleaf weeds). This prediction of CP, NOF, OOM, ONOF and OCP from the pasture components had a relative low accuracy (49 -74.1 %) and a further refinement of this model for possible use on farms in order to improve grazing management and animal production is advised. In conclusion it could be said that broadleaf weed control caused a definite increase in medic seed and OMproduction, but Orion should not be used with an IMI herbicide. All the cultivars, except for Orion, produced enough seedlings up to the second year to ensure sustainability of the medic pasture. All the cultivars, except for Orion, produced a sufficient amount of OM early in the growing season. Caliph however, produced by far the most OM later in the growing season. A reduction of broadleaf weeds and medic pods will increase the digestibility of NOFand OMand therefore increase the quality of the pasture. Pods however are an important part of summer forage and the aim should therefore rather be to reduce the number of broadleaf weeds in the pasture. / AFRIKAANSE OPSOMMING: Die doel van hierdie studie was om die effek van 'n gras en breëblaar onkruiddoder (wat sekere grasse beheer), metode van bewerking tydens die saai van koring asook nege verskillende medic kultivars op die regenerasie, droë materiaal produksie en kwaliteit van medic weidings te bepaal. Die proef is gedoen op Langgewens proefplaas wat geleë is in die Swartland koring produserende gebied. Nege medic kultivars is geëvalueer nadat die weiding met of haloxyfop-R-metiel ester (HAL) of imazamox (IMI) onkruiddoders gespuit is en onderwerp is aan of 'n vlak tand of geen bewerking tydens die saai van koring. Grondmonsters is geneem in Januarie 2000 om die grootte van die medic en onkruid saadbank asook om die graad van dormansie in die verskillende medic kultivars se sade te bepaal. Droë materiaal monsters is gedurende die 2000 groeiseisoen geneem om die droë materiaal produksie van die verskillende medic kultivars asook onkruid spesies te bepaal. Droë materiaal monsters is gedurende Oktober 1998 geneem en gebruik om die ruproteïn (CP) en neutraaloplosbare vesel (NDF) inhoud van die weiding te bepaal. Die monsters is in vitro verteer en die verteerbaarheid van CP (OCP), NDF (ONOF) en droë materiaal (DOM) is bepaal. Resultate wys dat saailing vestiging verskil tussen die verskillende kultivars wat gebruik is, verskillende onkruiddoder behandelings asook die stadium van die weidings/koring. Die kultivars het meer geproduseer waar die weiding met IMI behandel is in vergelyking met waar HAL toegedien is, asook waar koring nog nie gesaai is nie. Na 'n jaar van weiding en 'n jaar van koring, het die kultivars Sephi en Paraggio die meeste saailinge, en Caliph en Orion die minste saailinge gehad. Caliph het egter 'n hoë graad van dormansie in sy saad getoon, terwyl die swak vestiging van Orion die gevolg is van die kultivar se hoë sensitiwiteit teenoor IMI. Min verskil is gevind tussen die nege kultivars, vroeg in die groei seisoen (Julie - Augustus), wat kumulatiewe droë materiaal produksie betref, behalwe vir Orion wat erg beskadig is deur die IMI behandeling. Aan die einde van die groeiseisoen (Oktober 2000) was die kumulatiewe droë materiaal produksie van die kultivar Caliph (2010.1 kg/ha) betekenisvol hoër as al die ander kultivars behalwe vir Parabinga (1053.4 kg/ha). Weidingsmonsters, waarvan die botaniese samestelling bekend was, is ontleed vir CP, NDF, DDM, DCP en DNDF. Daar is geen betekenisvolle verskille gevind in die botaniese samestelling van die weidingmonsters geneem in 1998 nie,maar die variasie in botaniese samestelling het veroorsaak dat IMI in vergelyking met HAL 'n laer DNDF and DDM inhoud het. 'n Model is opgestel wat die weidingskomponente (medic hooi, medic peule, gras en breëblaar onkruide) gebruik om die kwaliteits parameters (CP, NDF, DDM, DCP en DNDF) van die weiding te skat. Hierdie skatting van CP, NDF, DDM, DCP en DNDF deur van die weidingskomponente gebruik te maak het 'n relatiewe lae akuraatheid gehad (49 -74.1 %) en verdere verfyning van hierdie model vir moontlike gebruik op plase, ten einde weidings bestuur en diere produksie te verbeter, word voorgestel. Die gevolgtrekking kan gemaak word dat breëblaar onkruidbeheer 'n definitiewe verbetering in die medic saailing en droë materiaal produksie van die medies te weeg gebring het, maar die kultivar Orion behoort nie saam met 'n IMI gebruik te word nie. AI die getoetste kultivars, behalwe Orion, het voldoende saad oorlewing vertoon tot en met die tweede jaar van die rotasie om lewensvatbaarheid van die sisteem te verseker en alle kultivars, behalwe Orion, het voldoende droë materiaal produseer vroeg in die groeiseisoen. Caliph het egter laat in die groeiseisoen by verre die meeste droë materiaal geproduseer. 'n Vermindering in die hoeveelheid breëblaar onkruide en peule in die weiding sal tot 'n verhoging in die verteerbaarheid van NDF en DM lei en dus 'n verhoging in die kwaliteit van die weiding tot gevolg hê. Peule is egter 'n belangrike bron van voedsel aan weidende diere gedurende droë somermaande en die verbetering van weidings moet eerder gedoen word deur te poog om breëblaar onkruide te beheer.