During this work the genetic variation within a cultivated lentil landrace wasexploited through mass and pedigree honeycomb selection. The aim was toinvestigate the efficiency of the honeycomb selection method in the management ofthe planting stock relative to its load in seed borne viruses, the ability to select forvirus-free lines and possibly isolate resistant/tolerant to viruses lines, as well as theprospects of sowing monogenotype or polygenotype lentil varieties under low inputconditions.A non-replicated (NR-0) honeycomb experiment was established in the region ofOrestiada, Greece, during the 2006-2007 season, at the low density of 1,2 plants/m2.Fifteen plants were selected for high grain yield to form as many 1st generation lines.Additionally, the application of five different intensities of selection led to five 1stgeneration populations. In 2007-08, progeny evaluation was conducted in an R-21experiment and three lines were isolated based on the genotype crop yield potential,which is defined by the progeny yield potential index and coefficient of homeostasis.From these superior three lines we selected the nine plants showing the highest cropyield potential (formula A) and the three plants with the highest absolute yield, whilethree more high yielding plants were selected from other lines with low crop yieldpotential, thus leading to 15 2nd generation lines. In addition, five different intensitiesof selection were applied to form an equal number of 1st generation populations. In2008-09, progeny evaluation of the 2nd generation lines was conducted in an R-21experiment. At the same time, the 1st generation selections were evaluated in densestand conditions according to the randomized complete-block (RCB) design at anearby field. Based on coefficient of yield within the ring 11 plants were selected thatled to 11 3rd generation lines and one new population consisting of 2nd generationsline seeds. In 2010-11, progeny evaluation of the 3rd generation selections wasconducted in an R-13 experiment, while nearby 2nd generation selections wereevaluated in dense stand conditions, in a RCB experiment.During the first year of experimentation, the severe infection by seed-borneviruses resulted in low yield as well as unsatisfactory results for the rest of theparameters calculated. Several plants in the field exhibited viral-disease-like symptoms and a vast number of them died. This didn’t allow the application of thebreeding selection criteria, so that the selection was only based on the absolute plantyield. ELISA tests detected only the PSbMV in the original population.The selection of symptomless plants from the original population in the absenceof competition proved an efficient method for improving the sanitary status of theplanting stock, since the 2nd generation populations yielded higher compared to theoriginal population by 16% in the absence of competition and 5,7% in farming densitycondition. Moreover, a significant positive linear correlation was found between thenumber of the selected plants and the number of them found positive for the BYMV.The 1st generation lines 8, 2 and 15 were selected in the absence of competitiondue to their high crop yield potential and were used to derive the 2nd generation lines.Lines with low coefficient of homeostasis were positive for virus presence. On theother hand, line 8 and the lines deriving from this had high coefficient of homeostasisand escaped virus infection. The coefficient of homeostasis seems to be an effectivecriterion for revealing a putative presence of seed borne virus infection in the linesunder evaluation. The importance of this criterion is moreover demonstrated by theselection of line 15 according to its value that later produced promising 2nd and 3rdgeneration lines.Four 2nd generation lines deriving from line 8 yielded significantly higher than theoriginal population in the absence of competition by 81 to 139%. The application ofthe ECYP as line selection criterion proved efficient for the identification of highyielding and free from virus infection genotypes. In contrast, lines selected only fortheir high yield regardless of their crop yield potential failed to exhibit stability ofperformance.The 3rd generation selections had low CV most probably due to the improvementof the planting material sanitary status, since seed-borne viruses were successfullyremoved and possibly tolerant/resistant to viral diseases genotypes resulted in lowpressure from virus infections. The crop yield potential recorded in this generationcompared to the progress observed in the previous generations indicates that geneticstability has been achieved.The lines which were distinguished due to crop yield potential (ECYP), respondedpositively to the selection and progeny behavior showed that the equation can be a powerful tool for creating inbred lines that are characterized by stability and highproductivity. Eventually, the application of the honeycomb method selection resultedin lines with satisfactory performance under farmer’s conditions, since three 2ndgeneration lines had higher yield compared to the original population by 22 to 24%.Under typical plant density conditions, the 1st generation heterogenouspopulations showed higher yield than the original population compared to the linesby 7,2 and 5,6% respectively. In contrast, the 2nd generation lines yielded higher thanthe populations by 9,4 and 5,7% compared to the original population respectively.Thus, the answer to the question whether monogenotype or polygenotype varietiesshould be recommended for cultivation requires further investigation.
Intracultivar selection for seed quality of soybeans in an ultra-low-density selection model (Honeycomb Selection Designs)