Genetic variability for yield and quality characteristics in South African pumpkin (Cucurbita maxima Duch.)

Abstract

English: The objective of this study was to study the inheritance of agronomic and quality characteristics in South African pumpkin germplasm and to identify characteristics that could be improved through hybrid breeding. 2. Six parental lines were crossed. The parents with the 15 F1-crosses were planted in Greytown as a randomized block design with six replications. Three of the replications were used for destructive measurements at the flowering stage, while the other three were used to evaluate yield and quality characteristics at a mature harvesting stage. 3. Significant phenotypic differences were found among the parental lines and their F1-offsprings for all the characteristics evaluated. C was the superior parent in nine of the 19 characteristics measured. For yield, fruit number, LPL, LBL and MBT parents in general tend to have a lower ranking than the crosses. The opposite was noticed for ST where the pure breeding parents performed better. 4. Specific combining abilities were calculated to be significant for all the characteristics, indicating that general combining ability predicted performances deviated from the observed performances for some of the crosses. 5. The calculated percentage variation due to GCA indicated that for most of the characteristics mostly additive variance influenced the expression of the characteristic. For LPL, yield, fruit mass, fruit number, ST, TT and MBT a large percentage (>25%) of the variation were influenced by non-additive gene action. Very high broad sense heritabilities were calculated for all the traits. The lowest value was calculated for OBES (0.81). Very high narrow sense heritabilities were calculated for W, ML, TL, W:ML, SHT, BT, density, OW:L, OTT and OST. Heritabilities for fruit mass, LBL, ST, TT and OBES were calculated to be high. Moderate heritabilities were calculated for yield, fruit number, LPL and MBT. 7. Ovary and leaf morphology showed strong phenotypic and genotypic correlations with fruit characteristics expressed at a mature stage. OW:L and OST had respectively genetic correlations of 0.87 and 0.91 with SHT. OTT had a genetic correlation of 0.90 with density. Density had negative correlations with LPL (-0.94) and LBL (-0.85). 8. The highest direct response to selection was calculated for fruit mass, which showed a 122.88% improvement on the population mean. Direct selection response for density was calculated to be 98.74%. For most characteristics it was concluded that gain through indirect selection could be as successful as with direct selection. 9. Positive mid-parent heterosis and high parent heterosis were calculated for all characteristics except ST that showed only negative HPH. Yield (-11.30 to 88.21 %), fruit mass (-58.90 to 94.05%), fruit number (-57.12 to 11.56%) and density (-41.64 to 93.81 %) were identified to express the highest heterosis in the trial.

Afrikaans: Die doel van die studie was om die oorerwing van agronomiese en kwaliteitseieskappe in Suid-Afrikaanse pampoenkiemplasma te bestudeer. 2. Ses ouerlyne is in In dialleelontwerp gekruis. Die ses ouers asook die 15 F1-kruisings is in Greytown geplant. Die proef is geplant as In gerandomiseerde blokontwerp met ses herhalings. Drie herhalings is vir destruktiewe waarnemings tydens blomstadium gebruik. Die oorblywende drie herhalings is gebruik vir die evaluering van opbrengsen kwaliteitseienskappe na oes, nadat die vrugte volwassenheid bereik het. 3. Vir beide die ouerlyne en F1-kruisings is betekenisvolle verskille bereken vir gemideldes van alle eieskappe. C was die beste ouer vir nege van die 19 eienskappe wat geëvalueer is. Vir opbrengs, vrugaantal, LPL, LBL en MBT het die ouers oor die algemeen swakker presteer as die kruisings. Die teenoorgestelde is waargeneem vir ST waar die suiwertelende ouers beter presteer het. 4. Die berekende spesifieke kombineervermoë was betekenisvol vir alle eienskappe. Dit wil sê die voorspelde voorkomste het afgewyk van die waargeneemde voorkomste. 5. Die berekende persentasie variasie as gevolg van algemene kombineervermoë het aangedui dat die meeste van die eienskappe se uitdrukking hoofsaaklik deur additiewe variansie beïnvloed word. Vir LPL, opbrengs, vrugmassa, vrugaantal, ST, TT en MBT is In groot persentasie (>25%) van die variasie veroorsaak deur nie-additiewe geenwerking. Baie hoë waardes vir erflikheid in die breë sin is vir alle eienskappe bereken. Die laagste waarde is vir OBES (0.81) bereken. Baie hoë waardes vir erflikheid in die nou sin is vir W, ML, TL, W:ML, SHT, BT digtheid, OW:L, OTT en OST bereken. Hoë erfbaarhede vir vrugmassa, LBL, ST, TT en OBES is ook bereken. Erfbaarhede vir opbrengs, vrugaantal, LPL en MBT is bereken as matig. 7. Vrugbeginsel- en blaarmorfologie het sterk fenotipiese en genotipiese korrelasies getoon met vrugeienskappe wat op "n volwasse stadium uitgedruk word. OW:L en OST het afsonderlik genetiese korrelasies van 0.87 en 0.91 met SHT getoon. OTT het "n genetiese korrelasie van 0.90 met digtheid gehad. Digtheid was negatief gekorreleerd met LPL (-0.94) en LBL (-0.85). 8. Die hoogste direkte seleksie respons is bereken vir vrugmassa, wat "n 122.88% verbetering op die populasie gemiddeld tot gevolg gehad het. Direkte seleksie respons vir digtheid was 98.74%. Vir meeste eienskappe kon afgelei word dat verbetering deur indirekte seleksie net so suksesvol kan wees as met direkte seleksie. 9. Positiewe mid-ouer en beter-ouer heterose is bereken vir alle eienskappe behalwe vir ST wat net negatiewe beter-ouer heterose getoon het. Opbrengs (-11.30-88.21 %), vrugmassa (-58.90-94.05%), vrugaantal (57.12-11.56%) en digtheid (-41.64-93.81 %) het die meeste heterose in die proef getoon.