S₁ PER SE RECURRENT SELECTION IN THREE SPRING CANOLA (BRASSICA NAPUS) POPULATIONS

J. D. Patel, M. Elhalwagy,  I. Falak, L. Tulsieram

Pioneer Hi-Bred International Inc., Canola Research Center, 12111 Mississauga Road,  RR#4, Georgetown, Ontario, Canada.  Email:  pateljd@phibred.com

ABSTRACT

Eight cycles of S₁ per se recurrent selection were applied in three spring canola (Brassica napus) populations, Pop-A, Pop-B and Pop-E.  The traits under selection were days to maturity, oil %, protein %, visual agronomic score and blackleg resistance.  All three populations were later maturing and lower in oil % compared to the checks at cycle 4 (C₄) and became earlier maturing and higher in oil % compared to the same checks at C₁₁.  Protein % declined slightly after eight cycles of recurrent selection. These populations also became superior for visual agronomic score and blackleg resistance. 

KEYWORDS:  oil %, protein %, heritability, blackleg, maturity

INTRODUCTION

Different methods of recurrent selection have been used in open pollinated crops to improve germplasm and populations. Although, many of the recurrent selection methods could be used in self pollinated crops,  their applications have been restricted mostly to open pollinated crops. In most Brassica napus breeding programs, very little emphasis is placed on population improvement and recurrent selection. Recent shifts, in spring and winter forms of Brassica napus, from open pollinated varieties to hybrids, has led many canola  breeders to look for heterotic source populations to extract inbred lines.  Well-defined heterotic pools are not available in spring or winter canola.  Many organizations are trying to create the heterotic pools for future inbred line development.

There are several forms of population improvement methods (Hallauer and Miranda, 1981).  A single population can be improved using intra-population methods such as S₁ per se, S₂ per se, mass selection, or half-sib selection.  However, in order to increase heterosis between two populations over time, inter-population selection such as reciprocal full-sib (FS) and reciprocal half-sib (HS) recurrent selection should be used.  Thompson and Hughes (1986) described a few methods of population improvement.  Recurrent selection has been reported in oilseed turnip rape, B. campestris (Downey and Rakaow, 1987).  Patel et al. (1991) reported four cycles of recurrent selection in three spring B. napus populations and observed drastic improvements in C₃ FS over C₀ FS for yield, maturity and oil %.

The objective of this investigation was to apply S₁ per se recurrent selection in the same three populations to improve maturity, quality traits (oil%, protein%), disease resistance and agronomic performance.

MATERIALS AND METHODS

The synthesis of cycle zero (C₀) of the three populations is as described by Patel et al. (1991). Two populations (Pop-A and Pop-B) were subjected to four cycles (C₀ to C₃) of full-sib (FS) reciprocal recurrent selection. In the third population (Pop-E), two cycles of S₁ per se were completed (Patel et al 1991).  Pop-A, Pop-B and Pop-E were all subjected to an additional eight cycles of S₁ per se recurrent selection from 1991 to 1998, as specified in Table 1.  Each cycle was completed in one year, following the same three steps: (1) S₁ production in the greenhouse during winter, (2) S₁ evaluation in field during summer, and (3) resynthesis in the greenhouse during fall. In each cycle, the S₁ evaluation was carried out in the replicated nursery row experiments planted at two locations in Ontario and W. Canada.  The observations such as days to mature, visual agronomic score (1=poor, 9=excellent), internal blackleg score (1=poor, 9=excellent) were recorded in each replicate at a minimum of one location.  At harvest, 15-20 g of seed were collected from each replicate at each site and were analyzed for oil %, protein % using Near Infrared Spectroscopy. During each year, approximately 10 check varieties were included in each trial while evaluating the S₁s.  Three checks, Westar, Profit and Legend were common in each cycle. Combined analysis of variance was performed for the traits recorded at two locations while simple ANOVA was performed for the rest of the traits.  For each trait in each cycle, population mean (m) and phenotypic standard deviation (s) were estimated by calculating the average and standard deviation of all the S₁ lines.  Since the three checks (Westar, Profit and Legend) were common in each cycle, the differences between the mean of each S₁ line (X_i) and the mean of check (X_C) was expressed in terms of population standard deviation units and used in plotting the frequency distribution charts (Fig 1 & 2).

RESULTS AND DISCUSSION

The analysis of variance for each cycle in each population showed mean squares due to genotypes to be significant for most of the traits.  The population mean (m), standard deviation (s), mean of the selected lines (X_S), and mean of the checks (X_C) are presented in Table 2.  For days to maturity, (m-X_C) and (m-X_C)/s comparisons between the cycles showed that the effect of recurrent selection was more dramatic in Pop-B and Pop-E compared to Pop-A (Table 1 and Fig 1 & 2).  On average, Pop-B and Pop-E lines were respectively 2.0 and 0.6 days later maturing than the checks at C₄.  These lines became 2.0 and 2.8 days earlier, respectively, than the checks at C₁₁ or C₉.  The selected S₁ lines in the advanced cycles were respectively 1.9, 2.6, and 3.0 days earlier than the checks in Pop-A, Pop-B and Pop-E.  S₁ lines in each cycle were visually scored for their yield potential.  The S1 lines looked inferior compared to the mean of the checks in C₄ in both Pop-A and B, which improved drastically in C₁₁ (Fig 1).  The improvement in Pop-E was not so dramatic (Table 2, Fig  2).  The m-X_C comparisons between different cycles for oil suggested that at early cycle, S₁ lines in Pop-A, Pop-B and Pop-E were respectively 1.1%, 1.1% and 0.1% lower than the check mean.  These became respectively 2.0%, 2.8% and 2.7% higher than the same check mean (Table 2 and Fig 1 & 2).  The selected S₁ lines in the advanced cycles were more than 3.0% higher in oil than mean of the checks.  For all three populations, (m-X_C) and (m-X_C)/s comparisons for protein % showed that in the advanced cycles, protein % went down slightly (Table 2, Fig 1 & 2).  For blackleg resistance, Pop-A made more progress compared to Pop-B as is evidenced by (m-X_C)/s comparisons between the cycles.  This is because Pop-B was very strong for blackleg resistance at C₄.

Table (1): Summary of eleven cycles of recurrent selection

Type of recurrent selection	Pop	Cycle	Year	Rep	Loc	# of Trials	Plot type	Total no  of  progenies	Avg. # of selected progenies
FS	AxB	C0-C3	1987-90	2	2	4	Yield	96	25
S1 per se	A, B	C4-C11	1991-98	2	2	1	Row	150	30
S1 per se	E	C2-C9	1991-98	2	2	1	Row	150	30

The present study showed that broad sense heritability estimates were high for quality traits like oil % and protein %, moderate for days to maturity, and low for visual agronomic score.  The progress from recurrent selection was generally high for all the traits except for protein because selection pressure for protein % was not as high as other traits, as shown by (X_S-m )/ s comparisons. The heritability estimates are not available for blackleg score since the data collection was done from a single replicate only. During every cycle (C₀ to C₁₁) in each population, while selfing the plants in the greenhouse (S₀ to S₁), severe blackleg pressure was applied and susceptible plants were discarded.  This explains why blackleg resistance in advanced cycle has improved so dramatically in all populations.

Table (2): Effect of eight cycles of S1 per se recurrent selection on different parameters
Parameter	Days to mature		Visual Score		Oil %		Protein %		Blackleg
Pop-A	C4	C11	C4	C11	C4	C11	C4	C11	C4	C9
Mean of S1 (m)	81.7	84.8	6.2	6.0	45.5	44.9	25.1	25.8	6.0	7.7
Std. Dev. (s)	1.3	1.8	0.7	1.0	0.8	1.4	0.6	0.8	1.7	1.0
Mean Sel S1 (XS)	81.9	84.3	6.3	6.6	46.2	46.1	25.1	25.8	6.2	8.2
Mean of Chk (XC)	82.0	86.2	6.9	4.4	46.6	42.9	25.5	27.4	5.1	3.9
(XS-m)/s	0.1	-0.3	0.1	0.6	0.9	0.8	0.0	0.0	0.1	0.5
(XS-XC)/s	-0.2	-1.0	-0.6	2.1	-0.3	2.2	-0.4	-1.9	1.1	4.2
(m-XC)/s	-0.2	-0.8	-0.9	1.6	-1.3	1.4	-0.8	-1.8	0.5	3.7
Heritability %	43.9	55.8	40.4	46.2	78.3	80.2	66.6	54.3	-	75.8
Pop-B	C4	C11	C4	C11	C4	C11	C4	C11	C4	C9
Mean of S1 (m)	82.2	87.1	6.5	6.1	45.3	45.4	25.5	27.0	7.5	8.1
Std. Dev. (s)	1.2	1.6	0.6	0.9	1.5	1.3	0.9	0.7	1.0	1.3
Mean Sel S1 (XS)	81.9	86.6	6.7	6.4	46.4	46.2	25.3	27.0	7.6	8.5
Mean of Chk (XC)	80.2	89.2	6.6	3.2	46.4	42.6	25.5	28.5	3.4	2.7
(XS-m)/s	-0.2	-0.3	0.3	0.2	0.7	0.6	-0.3	-0.1	0.1	0.3
(XS-XC)/s	1.8	-1.6	0.1	3.5	-0.1	2.8	-0.2	-2.2	4.3	4.5
(m-XC)/s	1.8	-1.3	-0.2	3.2	-0.8	2.1	0.0	-2.2	4.0	4.2
Heritability %	47.1	58.9	18.8	42.1	92.0	79.5	86.2	53.6	-	-
Pop-E	C2	C9	C2	C9	C2	C9	C2	C9	C2	C7
Mean of S1 (m)	79.6	85.2	6.9	7.6	46.2	45.6	25.2	26.8	-	8.4
Std. Dev. (s)	1.4	1.8	0.5	0.5	0.6	1.5	0.5	0.8	-	0.8
Mean Sel S1 (XS)	79.4	84.9	6.7	8.0	46.6	46.7	25.7	26.6	-	8.4
Mean of Chk (XC)	79.0	88.0	6.6	7.0	46.3	42.9	25.6	28.2	-	3.9
(XS-m)/s	-0.1	-0.2	-0.3	0.8	0.8	0.8	0.8	-0.3	-	0.1
(XS-XC)/s	0.3	-1.8	0.1	2.2	0.5	2.6	0.0	-1.9	-	6.0
(m-XC)/s	0.4	-1.6	0.5	1.4	-0.3	1.8	-0.8	-1.6	-	5.8
Heritability %	54.4	56.8	12.4	40.9	48.4	80.6	54.2	61.9	-	-

The Four cycles of reciprocal full-sib recurrent selection was quite effective in improving S₀ x S₀ full-sib performance for yield, maturity and oil % (Patel et al 1991).  Since the Canadian canola variety registration system imposed extremely strict quality and disease standards, it was decided to switch from full-sib recurrent selection to S₁ per se recurrent selection in 1991.  S₁ per se recurrent selection has been quite effective in improving the performance of the lines for the important traits.  From each cycle, the selected S₁ lines were used in a pedigree breeding program.  Some of Pioneer Hi-Bred’s commercial canola cultivars have originated directly or indirectly from the various cycles of recurrent selection populations.

REFERENCES

DOWNEY, R.K., and RAKOW, G.F.W. 1987. Rapeseed and Mustard. In: Principles f Cultivar Development. W.E. Fehr (ed.). McMillan, New York. pp. 437-486.

HALLAUER, A.R. and MIRANDA, J.B. 1981. Quantitative Genetics in maize breeding. Iowa State University Press, Ames, pp. 359-361.

THOMPSON, K.F. and HUGES, W.G. 1986. Breeding  and varieties. In: Oilseed Rape. D.H. Scarisbrick and R.W. Daniels (eds.) Collins, London, pp. 32-82.

PATEL, J.D., ELHALWAGY, M., CHARNE, D.G., and GRANT, I. 1991. Intra and interpopulation improvement in spring Brassica napus. Proc. 8^th International Rapeseed Congress. Saskatoon. Vol 1, pp A-01.