Family and parent evaluation for sugarcane yield in early stage breeding populations in South Africa

Loading...
Thumbnail Image
Date
2016-11
Authors
Mbuma, Ntombokulunga Wedy
Journal Title
Journal ISSN
Volume Title
Publisher
University of the Free State
Abstract
English: Family selection provides the potential to improve gains for quantitative traits with low heritability such as cane yield at early selection stages. Family evaluation data can also be used to identify superior parents for use in future crossing. The objectives of the study were to compare family with individual genotype selection for cane yield components; to identify and determine the proportions of elite families for sugarcane yield; to determine the optimum family selection rate and identify ideal trait combinations among the elite families; to identify superior parents using family data and determine the proportion of superior parents within populations in the Midlands breeding programmes. Family data on stalk number, stalk height and stalk diameter were collected from a sample of the first 20 genotypes per family plot. Family yield data were analysed using Statistical Analysis System (SAS) linear mixed models. Family variance was 1.2 to 5.0 times that of individual genotype variance indicating larger variability among families compared to individual progenies. Families produced larger broad-sense heritability estimates (25 to 90%) than individual genotypes (1.6 to 23.5%) suggesting that selection for superior families would be more accurate than for individual genotypes. Families produced higher predicted selection gains (9 to 59%) compared to individual genotypes (1 to 12%) which indicated that family selection would be more efficient. Results indicated highly significant differences (P<0.0001) for family and group effects for all traits. The highly significant differences (P<0.0001) observed for family within group effects for stalk diameter indicated large variability for families within the different cane yield groups. The humic soil populations produced a significantly (P<0.05) high proportion of elite families and the higher optimum selection rate (30%) compared to sandy soil trials (25%) suggested that humic soil populations contained more elite families compared to sandy soil populations. Using BLUP estimates, female (82H0397, 96H0259, 98B0460, 98S0290, 98H0590, 98S1362 and N52) and male (02S0639, 82H0397, 97B0272, 98B0460, 98B1889 and N52) parents produced progenies with significantly (P<0.05) high cane yield compared to the population mean in both humic and sandy soil populations. The identified elite parents could be used in building a core germplasm pool of genotypes that produce elite progenies. Humic soil populations contained a higher proportion of elite parents (30.6%) compared to sandy soil populations (22.5%) which is probably due to longer cycles of recurrent breeding and selection in humic soil breeding programmes. The low proportions of elite parents for both humic and sandy soil populations obtained suggest that intensive parent evaluation and development is required.
Afrikaans: Familie-seleksie het die potensiaal om vordering vir kwantitatiewe eienskappe met lae oorerflikhede, soos rietopbrengs in ‘n vroeë seleksiestadium, te bevorder. Familie-evaluasie data kan ook gebruik word om beter ouers te identifiseer wat gebruik kan word in toekomstige kruisings. Die doelwitte van hierdie studie was om familie- en individuele genotipe seleksie met mekaar te vergelyk ten op sigte van rietopbrengs-komponente; om die proporsies “elite” families te evaleueer en te bepaal vir rietopbrengs; om die optimale familie-seleksie tempo te bepaal en om die ideale eienskapkombinasies tussen families te identifiseer; om beter ouers te identifiseer deur gebruik te maak van familie-data en om die proporsie beter ouers binne die Midlandse telingsprogramme te identifiseer. Familie-data in terme van aantal stele, steelhoogte en steeldeursnee is versamel vanaf die eerste 20 genotipes per familie-plot. Familie-opbrengs data is ontleed deur gebruik te maak van Statistiese Analise Stelsel (SAS) se lineêre gemengde modelle. Familie-variansie was 1.2 tot 5.0 keer dié van individuele genotipe variansie en dit toon aan dat groter veranderlikheid tussen families voorgekom het as tussen individuele nageslagte. Families het groter breë sin oorerflikheidsskattings (25 tot 90%) gelewer in vergelyking met individuele genotipes (16 tot 23.5%) en dit toon aan dat seleksie vir beter families meer akkuraat sal wees as vir individuele genotipes. Families het hoër verwagte seleksie-vorderings (9 tot 59%) gelewer in vergelyking met individuele genotipes (1 tot 12%) wat aanngetoon het dat familie-seleksie meer effektief sal wees. Resultate het aangetoon dat daar hoogs betekenisvolle verskille (P<0.0001) vir familie- en groepeffekte vir alle eienskappe voorgekom het. Die hoogse betekenisvolle verskille (P<0.0001) vir familie- en groepeffekte vir steeldeursnee het aangetoon dat daar groot veranderlikheid tussen families binne die verskillende rietopbrengsgroepe was. Die humusgrond-populasies het ‘n betekinsvolle (P<0.05) hoër proporsie “elite” families geproduseer en die hoër optimum seleksietempo (30%) invergelyking met sandgrond proewe (25%) het voorgestel dat die humusgrond populasies meer “elite” families vervat het as die sandgrond populasies. Deur gebruik te maak van BLUP skattings het vroulike (82H0397, 96H0259, 98B0460, 98S0290, 98H0590, 98S1362 en N52) en manlike (02S0639, 82H0397, 97B0272, 98B0460, 98B1889 en N52) ouers nageslagte geproduseer wat betekenisvol (P<0.05) hoë rietopbrengste in beide humus- en sandgrond populasies gelewer het. Die geïdentifiseerde “elite” ouers kan gebruik word om ‘n kern kiemplasmapoel van genotipes te bou wat weer “elite” nageslagte kan produseer. Humusgrond populasies het ‘n hoër proporsie “elite” ouers (30.6%) vervat in vergelyking met sandgrond populasies (22.5%) wat waarskynlik toegeskryf kan word aan die langer seleksiesiklusse van herhalende teling en seleksie in humusgrond telingsprogramme. Die lae proporsies “elite” ouers wat voogekom het vir beide humus- en sandgrond populasies stel voor dat intensiewe ouerevaluasie en ontwikkeling benodig word.
Description
Keywords
Sugarcane, Family, Individual genotype, Parents, Yield, Predicted selection gains, Dissertation (M.Sc. ((Plant Sciences: Plant Breeding))--University of the Free State))--University of the Free State, 2016
Citation