Ana səhifə

Defra pvsd funded research final report


Yüklə 117.37 Kb.
tarix24.06.2016
ölçüsü117.37 Kb.

5th Sugar Beet DUS Test Centre meeting 11 January 2006 (Harvest year results 2005)

Agenda item 5



DRAFT NIAB

DEFRA PVSD FUNDED RESEARCH – FINAL REPORT


PROJECT TITLE:

Simulation of Sugar Beet DUS Testing Based on the Scheme in Operation on Denmark

PROJECT START DATE:

01/04/04


PROJECT END DATE:

31/03/05


NIAB PROJECT LEADER:

John LAW (BSS)


PROJECT COSTS:

£12.5k


_____
SUMMARY OF PROJECT ACTIVITIES AND RESULTS:
Simulation of Sugar Beet DUS Testing, Based on the Scheme in Operation in Denmark


  1. Introduction

UK Sugar Beet (SBT) Distinctness, Uniformity and Stability (DUS) testing is currently based on two components. Firstly, four ‘core’ leaf characteristics are recorded (20 plants/plot) for varieties in a single trial, from which a set of derived characteristics are generated. Secondly, variety performance data are collated from the full set of VCU trials, with these data ‘imported’ into the DUS data set. The current decision rule is 2@5% (i.e. the within year criteria). Many varieties are not distinguishable after 2 years of DUS testing under the current arrangement, but most are cleared (i.e. are D, U and S) after 3 years. However, there are some exceptions in certain years, when good varieties in terms of agronomic performance cannot proceed and be commercialised, due to a lack of DU and S.
Work over a number of years has already been conducted at NIAB, looking at improving the efficiency of the SBT DUS test, with a view to being able to clear more varieties at the 2-year stage. The COYD x% decision rule has not thus far been implemented, as the ‘leaf’ and ‘VCU’ character sets (above) have a different variance structure and finding an appropriate value of ‘x’ to equate to the existing scheme has proved elusive. The use of COY is however under annual review.
There have been claims made that it is ‘easier’ to achieve D, U and S of sugar beet varieties in Denmark. It is known that the Danes have applied a different approach to SBT DUS testing (summarized below #). The purpose of this project was to learn more in detail about the Danish approach and to simulate as far as possible this approach, using historical UK data.
#In Denmark we have only one location (Tystofte) for DUS-purposes (all agriculture species). For VCU of sugar beet we have 3 locations (Tystofte and two other locations). We keep all data from DUS and VCU separate e.g. sugar content are assessed in both DUS and VCU trials. For distinctness test COYD is used. In general the level of probability of 1% is applied, but if we see two or more not correlated characteristics with a probability between 1 and 5% the variety can be accepted to be distinct.

  1. Materials and Methods

To simulate the Danish approach, in terms of a dedicated trial used for DUS decision taking (i.e. both leaf and VCU characters from the same, single trial), leaf measurement data and combined characters from a single VCU trial were collated from the existing data sets. The ‘UK single site’ data were analysed and the results compared to those obtained using the current UK approach. (ie leaf data from a single trial 20 plants/plot; VCU characters from the mean all trials).


For the UK simulated Danish approach (UK_SDK) both COYD5% and COYD1% were computed and compared to the UK 2 @5% rule.
Data were back-loaded from 1999-2004, giving five pairs of years (e.g. 1999-2000, …, 2003-2004) and four sets of 3-year decisions (1999-2001, …, 2002-2004).
The approved set of characteristics in the current UK DUS protocol were used in each year.

3. Results
In addition to having a dedicated DUS trial, the Danes use hypocotyl colour as a routine characteristic in sugar beet DUS testing, and find it very powerful at discrimination despite mixed hypocotyl reactions (heterogeneity) in some varieties. Attempts to replicate the Danish results in the UK, using seed of the reference varieties supplied by the Danes, were unconvincing. Even the reference varieties showed levels of heterogeneity that would make uniformity assessments difficult, and not all of the colour classes (states of expression) were clearly distinguishable. The reasons for this are not evident, but more work would be required before hypocotyl colour could be used with confidence in the UK.
Discussions took place with Danish colleagues, followed up with a field visit in May 2004 when the hypocotyl reactions were being recorded. This was a very useful visit as it allowed a definitive description of the Danish system to be established. The major differences between the systems are as below, with a summary of the results from the simulated UK data:


  • The Danish system is based around a single DUS trial, usually on the same site, with a reference seed sample being used in each sowing i.e. treating sugar beet as they do for most of the species they test for DUS.




  • The UK DUS testing system, in contrast to the Danish one outlined above, is operationally an adjunct of VCU testing with annual “bulks” used in the trials. VCU characteristics provide one part of the character set used for DUS – based on the over-trial means and the statistical standards based on the variety by site standard errors. Four “raw” leaf DUS measurements (and subsequent combined leaf characters) are taken from a single VCU trial – selected at random each year – where the discrimination is based on the variety by replicate error.




  • With the application of a single seed submission and constant field environment for DUS, the Danes have effectively eliminated two of the major sources of variation for DUS testing. In the UK we have seed, site and season variability; the Danes have only season. Of course it will depend on how consistently uniform is the Danes single site. The very fact that it is used year upon year and is located relatively close to the experimental station would suggest that at the very least the Danes have a good level of understanding of the site used for DUS testing of sugar beet.




  • UK data were back loaded with only the VCU data from the same single trial as the leaf measures. Note that there will still be annual variation due to different sites being used.




  • Depending on how that individual site behaved the errors may be lower or higher than the over-trial errors used in the current system.




  • Data from 1999 onwards have been used, with diploid and triploid varieties separated as usual. Five two-year simulated tests and four three-year simulated tests were performed with the full UK character set (“VCU set” + “leaf measurement set”).




  • In comparing the single UK site with the mean over all sites we used the current 2 @5% rule as the base line. The tables below show the percentages of varieties assessed as distinct (D) by the current system, but not distinct (ND) by the simulated system based on COYD.




  • Diploid varieties showed a reduction in D from 13-50% in the two-year cases and 33-89% reduction in D for the three-year cases. Triploids showed a 0-40% reduction for two-year cases and 55-100% reduction in three-year cases.




Result of multiple site 2@5% compared with













2000

2 yearTrip

2 yearDip













COYD @1%

%age

%age













% agreement

75

50













% disagreement

25

50













%D to ND

25

50


































2001

2 yearTrip

2 yearDip




2001

3 yearTrip

3 yearDip

COYD @1%

%age

%age




COYD @1%

%age

%age

% agreement

81

75




% agreement

44

67

% disagreement

19

25




% disagreement

55

33

%D to ND

19

25




%D to ND

55

33






















2002

2 yearTrip

2 yearDip




2002

3 yearTrip

3 yearDip

COYD @1%

%age

%age




COYD @1%

%age

%age

% agreement

100

87




% agreement

20

20

% disagreement

0

13




% disagreement

80

80

%D to ND

0

13




%D to ND

80

80






















2003

2 yearTrip

2 yearDip




2003

3 yearTrip

3 yearDip

COYD @1%

%age

%age




COYD @1%

%age

%age

% agreement

100

72




% agreement

38

11

% disagreement

0

28




% disagreement

62

89

%D to ND

0

28




%D to ND

62

89





















2004

2 yearTrip

2 yearDip




2004

3 yearTrip

3 yearDip

COYD @1%

%age

%age




COYD @1%

%age

%age

% agreement

30

76




% agreement

0

17

% disagreement

70

24




% disagreement

100

83

%D to ND

70

24




%D to ND

100

83



  • An example of the detailed summary results (Appendix 1) expand on the origins of the observed disagreements.




  • The Danes can also establish D after two years by application of a second criterion if the results show a variety non-distinctness after application of COYD at 1%. This second criterion is COYD2% in two or more ‘un-correlated’ characters. The definition and significance of this un-correlated status are far from clear.




  • The majority of UK applicants could be separated after 2 years using 2 x COYD 5% and 2 x COYD 2% characters, irrespective of the level of correlation. See example below.

COYD @ 5% FOR TWO CHARACTERS - Second Part of Danish Method

COMPARISONS BETWEEN 901 KWS 0125 AND 861 HI0015

T VALUES +VE IF KWS 0125 > HI0015
SIG LEVELS COYD T VALUES F3 CRIT

YEARS T PROB% SIG YEARS

0 1 0 1

10 PETW - - -0.74 46.83 NS -0.79 -0.13 0.27 NS COY



11 PETL -5 + -1.77 9.11 NS -2.11 0.40 6.22 * COY

12 TOTL -1 + -2.04 5.50 NS -2.66 0.26 5.69 * COY

13 LWDT - + -0.21 83.75 NS -0.97 0.69 1.32 NS COY

14 leafleng -5 - -1.49 15.07 NS -2.31 -0.15 1.74 NS COY

15 leafshap - - -1.59 12.77 NS -1.64 -1.12 0.21 NS COY

16 leafsize - + -1.11 28.16 NS -1.79 0.29 2.17 NS COY

17 petshape - + -1.10 28.65 NS -1.89 0.52 3.80 NS COY

18 TL/LL - + -0.04 97.15 NS -0.64 0.30 0.19 NS COY

19 TL/PW -1 + -1.39 18.12 NS -2.60 0.44 3.98 NS COY

44 LW/PW - + 0.28 77.99 NS -0.30 0.72 0.33 NS COY

55 AMMEQSUG + - 0.06 95.11 NS 0.35 -0.42 0.09 NS COY

5 AMMEQBT + - -0.24 81.34 NS 0.01 -0.86 0.06 NS COY

90 AMMGSUG + - 0.06 95.06 NS 0.35 -0.41 0.09 NS COY

6 ROOTYLD + +1 2.39 2.68 * 1.11 2.92 0.78 NS COY

45 CROWNHT - - -0.68 50.42 NS -0.03 -0.72 0.39 NS COY

73 FLBLIS + + 0.48 63.55 NS 0.87 0.19 0.12 NS COY

74 FLCOL + + 0.48 63.60 NS 0.99 0.08 0.19 NS COY

4 KMEQBT - -1 -2.32 3.11 * -1.66 -4.10 0.34 NS COY

22 KMEQSUG - -2 -1.34 19.42 NS -0.80 -2.52 0.26 NS COY

71 LFHABIT +1 - 0.98 33.89 NS 3.31 -1.63 5.23 * COY

72 LFWAVE + + 0.71 48.31 NS 0.77 0.73 0.07 NS COY

3 NAMEQBT + +5 0.80 43.45 NS 0.64 2.03 0.05 NS COY

33 NAMEQSUG + +5 0.86 39.94 NS 0.76 2.18 0.05 NS COY

1 SUG% -1 -2 -1.84 8.05 NS -2.58 -2.57 0.02 NS COY

66 SUGYLD + +2 1.73 9.90 NS 0.60 2.40 0.79 NS COY

8 TOPSIZE -1 - -2.30 3.25 * -3.65 -1.95 0.57 NS COY

2 TOTIMPSU - - -0.37 71.77 NS -0.05 -1.08 0.10 NS COY


  1. Conclusions

The clear differences in operational approach to sugar beet testing in UK and Denmark are responsible for the perception that it is “easier” for varieties to be DU and S in Denmark.


The Danes have a system where they can use hypocotyl colouration to good effect, based on an optimised growing system and very experienced, dedicated and trained recorders. It is generally their most discriminating single character. They do accept that they have within-variety variation, and have developed statistical approaches to cater for these effects. The UK scored the same variety material and found great difficulty in establishing a reliable assessment protocol. Of particular concern was the level of variability in the standard variety control colours (see pictures in appendix 2 and 3).


  • The use of “un-correlated characters” as a second distinctness criterion in the Danish approach further allows most candidates to be distinguished after 2-years of test. The definition of “un-correlated” is not straightforward and would need to be clarified before it could be used in a UK protocol.




  • To replicate the Danish system of single seed submission and fixed, designated DUS trial in the UK would be practically possible, but would inevitably entail increased costs. The existing UK DUS system has developed as a pragmatic and generally low cost approach to the realities of sugar beet breeding, variety testing, seed production and commercialisation. To move away from this would require the agreement of all stakeholders. The increased costs of a single seed submission/single trial system might be off-set to a degree, by for instance reducing the characters recorded for DUS purposes. However, applicants would have to accept that there may be an increase in the risk that their particular variety would not be D, U and S.




  • At the moment, although there are instances where highly performing varieties fail because of DUS problems, these are relatively rare and most varieties are cleared within three years. Unfortunately, there is no prospect of a “no-cost + efficiency” improvement that can be applied to the current UK sugar beet philosophy based on the Danish experience.



  1. Further Work

Contact should be retained with DUS sugar beet experts in Denmark and in other UPOV member states and further afield through the normal working processes, conferences and UPOV meetings. The development of new characteristics remains an important objective.

The application of the 2 @ 5% criterion versus COYDx% is an ongoing exercise in the UK that will be reported formally from time to time.


In addition, the discrimination power of the existing character set is reviewed annually as part of the ISO procedures applied to the DUS testing processes at NIAB generally.
Appendix 1.
Example Criterion Comparison

TWO YEAR DIPLOID – YEARS 2003 + 2004

Multiple Sites – Current NIAB Method
SUMMARY FOR 2 X 5% CRITERION
CANDIDATE VAR 1058 1059 1070 1071 1072 1073 1079 1088 1091

432 ROBERTA D D D D D D D D D

901 CINDERELLA D ND D D D D D D D

847 DOMINIKA D D ND ND ND ND D D ND

917 GANDALF D ND D ND D D D D ND

800 LATOYA* D D D D D D D D D

849 DORENA D ND D ND D ND D D ND

875 RAYO D ND D D D D D D D

951 ANEMONA D ND ND ND D D D D ND

974 SALVADOR D D D D D D D D ND

978 ASPECT ND ND ND ND D ND ND D ND

980 RADAR D ND ND ND D D D D ND

996 BANDIT D D D D D D D D D

997 RASKAL D ND ND ND ND ND D D ND

1001 PERNILLA D D D D D D D D D

1003 JUSTINA D D D D D D D D D

1009 DELFINA D D D ND D ND D D D

1011 LEXION D ND ND ND D ND ND D ND

1012 MIRIAM D D D D ND D D D D

1024 BEVERLY ND ND D ND D ND D D ND

1027 ACAPULCO D ND ND ND D D D D ND

1032 HI0037 D D D D D D D D D

1035 BUXOM D ND ND ND ND D D D ND

1040 BOBCAT D ND ND ND D ND ND D ND

1050 HARRY ND D D D D D D D ND

1052 MARS ND D D D D D ND D ND

1058 LION0313 - ND ND ND D ND D D ND

1059 LION0311 ND - D ND D ND D D ND

1070 FAME ND D - ND D ND ND D ND

1071 KINGSTON ND ND ND - ND ND D ND ND

1072 PALACE D D D ND - D D ND D

1073 MENTOR * ND ND ND ND D - ND ND ND

1079 ACE D D ND D D ND - D D

1088 3R64 D D D ND ND ND D - D

1091 Stru2303 ND ND ND ND D ND D D -

1095 D0203 D ND ND D D D D D ND

1097 S2363 ND D D D D D D D ND

1099 S2361 D D D ND D D ND D ND

1106 HI0362 D D D D ND ND ND D ND

1107 HI0361 D ND ND ND D ND ND D ND

1108 HI0358 D D D D ND ND D ND D

1110 HI0349 D D ND ND ND ND ND ND ND

1112 HI0367 D D D ND D ND D D ND

NO OF ND VARS 10 19 17 24 9 20 10 5 27


DISTINCTNESS ND ND ND ND ND ND ND ND ND
CANDIDATE VAR 1058 1059 1070 1071 1072 1073 1079 1088 1091
SUMMARY FOR 2 X 5% CRITERION
CANDIDATE VAR 1095 1097 1099 1106 1107 1108 1110 1112

432 ROBERTA D D D D D D D D

901 CINDERELLA D D D D D D D D

847 DOMINIKA D ND D ND ND D D D

917 GANDALF D D D D ND D D ND

800 LATOYA* D D D D D D D D

849 DORENA D D D D ND D D D

875 RAYO D D ND D D D D D

951 ANEMONA ND D D ND ND D ND D

974 SALVADOR D D D D ND ND D ND

978 ASPECT D ND ND D ND D D D

980 RADAR ND D ND ND ND D ND ND

996 BANDIT D D D D D D D D

997 RASKAL D D D ND ND ND ND ND

1001 PERNILLA D D D D D D D D

1003 JUSTINA D D D D D D D D

1009 DELFINA D D D ND D D D ND

1011 LEXION D D ND ND ND D D ND

1012 MIRIAM D D D D D ND D D

1024 BEVERLY ND D D ND ND D ND ND

1027 ACAPULCO D D ND ND ND D ND ND

1032 HI0037 D D D D D D D D

1035 BUXOM D ND D ND D ND ND ND

1040 BOBCAT ND ND ND ND ND D D ND

1050 HARRY ND ND ND D D D D D

1052 MARS ND D ND D D D D D

1058 LION0313 D ND D D D D D D

1059 LION0311 ND D D D ND D D D

1070 FAME ND D D D ND D ND D

1071 KINGSTON D D ND D ND D ND ND

1072 PALACE D D D ND D ND ND D

1073 MENTOR * D D D ND ND ND ND ND

1079 ACE D D ND ND ND D ND D

1088 3R64 D D D D D ND ND D

1091 Stru2303 ND ND ND ND ND D ND ND

1095 D0203 - ND ND D ND D D D

1097 S2363 ND - ND D D D D D

1099 S2361 ND ND - D ND D D D

1106 HI0362 D D D - ND D ND ND

1107 HI0361 ND D ND ND - D D ND

1108 HI0358 D D D D D - D D

1110 HI0349 D D D ND D D - ND

1112 HI0367 D D D ND ND D ND -

NO OF ND VARS 12 9 14 17 22 7 15 16


DISTINCTNESS ND ND ND ND ND ND ND ND
CANDIDATE VAR 1095 1097 1099 1106 1107 1108 1110 1112

Single Site – First Part of the Danish Method

SUMMARY FOR COYD CRITERION AT 1.0% LEVEL


CANDIDATE VAR 1058 1059 1070 1071 1072 1073 1079 1088 1091

432 ROBERTA D D D D D D D D D

901 CINDERELLA D ND ND ND D ND D ND D

847 DOMINIKA D ND ND ND ND ND ND D ND

917 GANDALF D ND D ND D D D D ND

800 LATOYA* D D D D D D D D D

849 DORENA D D D ND D D D D D

875 RAYO D ND D D D D D D D

951 ANEMONA ND ND ND ND ND ND D D ND

974 SALVADOR D ND D ND ND ND D D ND

978 ASPECT ND D ND ND ND ND ND D ND

980 RADAR ND ND ND ND D ND D D ND

996 BANDIT D D D D D D D D D

997 RASKAL D ND ND ND ND ND ND ND ND

1001 PERNILLA D D D D D D D D D

1003 JUSTINA D D D D D D D D D

1009 DELFINA D D D ND ND ND D ND D

1011 LEXION D ND ND ND ND ND D D ND

1012 MIRIAM D D D D D D D D D

1024 BEVERLY D ND ND ND D ND D D ND

1027 ACAPULCO D ND ND ND ND ND D ND ND

1032 HI0037 D D D D D D D D D

1035 BUXOM D ND ND ND ND ND ND ND ND

1040 BOBCAT ND ND ND ND D ND ND D ND

1050 HARRY ND ND D D D D ND D ND

1052 MARS ND D D D D D ND D ND

1058 LION0313 - ND D ND D ND ND D ND

1059 LION0311 ND - D ND D ND D D ND

1070 FAME D D - ND ND ND ND D ND

1071 KINGSTON ND ND ND - ND ND ND ND ND

1072 PALACE D D ND ND - ND D ND ND

1073 MENTOR * ND ND ND ND ND - ND ND ND

1079 ACE ND D ND ND D ND - D D

1088 3R64 D D D ND ND ND D - ND

1091 Stru2303 ND ND ND ND ND ND D ND -

1095 D0203 ND ND ND ND D D D D ND

1097 S2363 ND ND ND D D D ND D ND

1099 S2361 ND ND ND ND D ND ND D ND

1106 HI0362 D ND ND ND ND ND ND D ND

1107 HI0361 D ND ND ND D ND D D ND

1108 HI0358 D D D ND ND ND D D D

1110 HI0349 ND ND ND ND ND ND ND ND ND

1112 HI0367 D D ND ND ND ND D ND ND

NO OF ND VARS 15 24 23 30 18 27 15 11 28


DISTINCTNESS ND ND ND ND ND ND ND ND ND
CANDIDATE VAR 1058 1059 1070 1071 1072 1073 1079 1088 1091
SUMMARY FOR COYD CRITERION AT 1.0% LEVEL
CANDIDATE VAR 1095 1097 1099 1106 1107 1108 1110 1112

432 ROBERTA D D D D D D D D

901 CINDERELLA D ND D D ND D ND ND

847 DOMINIKA D ND D ND ND ND ND ND

917 GANDALF D D D D ND ND D ND

800 LATOYA* D D D D D D D D

849 DORENA D D D D D ND ND ND

875 RAYO ND D ND D D D D D

951 ANEMONA ND D ND ND ND D ND ND

974 SALVADOR D D D ND ND D D ND

978 ASPECT ND ND ND ND ND D D ND

980 RADAR ND D ND ND ND D ND ND

996 BANDIT D D D D D ND D ND

997 RASKAL D D D ND ND ND ND ND

1001 PERNILLA D D D D D D D D

1003 JUSTINA D D D D D D D D

1009 DELFINA D D D D D D D ND

1011 LEXION D D D ND ND D ND ND

1012 MIRIAM D D D D D ND D ND

1024 BEVERLY ND ND ND ND ND D ND ND

1027 ACAPULCO ND ND D ND ND D ND ND

1032 HI0037 D D D D D D D D

1035 BUXOM ND ND D ND ND ND ND ND

1040 BOBCAT ND ND ND ND ND D ND ND

1050 HARRY D ND ND ND D D D D

1052 MARS ND D ND D D D D D

1058 LION0313 ND ND ND D D D ND D

1059 LION0311 ND ND ND ND ND D ND D

1070 FAME ND ND ND ND ND D ND ND

1071 KINGSTON ND D ND ND ND ND ND ND

1072 PALACE D D D ND D ND ND ND

1073 MENTOR * D D ND ND ND ND ND ND

1079 ACE D ND ND ND D D ND D

1088 3R64 D D D D D D ND ND

1091 Stru2303 ND ND ND ND ND D ND ND

1095 D0203 - D ND ND ND D D D

1097 S2363 D - ND ND D D D D

1099 S2361 ND ND - ND ND D D ND

1106 HI0362 ND ND ND - ND ND ND ND

1107 HI0361 ND D ND ND - D ND ND

1108 HI0358 D D D ND D - ND D

1110 HI0349 D D D ND ND ND - ND

1112 HI0367 D D ND ND ND D ND -

NO OF ND VARS 17 15 20 26 23 12 24 27


DISTINCTNESS ND ND ND ND ND ND ND ND
CANDIDATE VAR 1095 1097 1099 1106 1107 1108 1110 1112


Page of 30/04/04


Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©atelim.com 2016
rəhbərliyinə müraciət