Predicting seasonal production of eragrostis curvula

Abstract

The potential for mathematical prediction of the yield and quality of forage crops in fodder production planning is still untapped. This paper presents a multiple regression equation accounting for 88% of the variation that occurred in some 270 individual annual yields of Eragrostis curvula. The data originated from nine sites in the eastern high summer‐rainfall regions in South Africa. The significant variables were level of applied nitrogen, rainfall, age of pasture and a combination of soil factors.

The “Richards” function was used to derive the seasonal distribution of growth rates of E. curvula at four of the nine sites. A procedure for predicting the corresponding production curve for a given farm site is proposed, based on the expected total yeild and a selected standard curve.

This technique shows promise and it will be applied to other pasture species.

Uittreksel

In die beplanning van voerproduksie is die potensiaal vir wiskundige voorspelling van die opbrengs en kwaliteit van voergewasse nog nie gebruik nie. Dié referaat bied ‘n veelvoudige regressie funksie aan, wat 88% van die variasie in meer as 270 individuele jaarlikse opbrengste van Eragrostis curvula verklaar. Die gegewens het by nege terreine in die ooste‐like hoë somer‐reënval gebied van Suid‐Afrika ontstaan. Die beduidende veranderlikes is peil van toegediende stikstof reënval, ouderdom van die weiding, en ‘n kombinasie van grondfaktore.

Die “Richards” funksie is gebruik om voorlopige syfers vir die seisoenale verspreiding van die groeitempo van curvula by vier van die terreine aan te bied. ‘n Prosedure om die ooreenstemmende produksie kurwe vir ‘n gegewe plaas te voorspel, is voorgestel, gebaseer op die verwagte totale opbrengs en ‘n uitgesoekte standaard kurwe.

Die metode lyk belowend en ander voergewasse sal soortgelyks ontleed word.