AOSA/SCST committees and regional groups
are studying pure seed determination in the
genus Artemisia (Asteraceae family)


Results of two studies:

1.  2004 Artemisia purity referee
2.  2004 Photo referee results

Artemisia
comparison

2004 Artemisia tridentata Referee

Application of AOSA and ISTA  pure seed definitions and the subsequent effects on viability, pure live seed, and pure seed units per gram for 2 lots of Artemisia tridentata.

Referee purpose: 
To compare ISTA and AOSA definitions for Artemisia pure seed. 
To assess consistency of Artemisia pure seed determination among labs.
To assess consistency of Artemisia PLS determination among labs.

Seed units of Artemisia are achenes.   Floral structures may or may not be attached.  In the seed trade, lots of Artemisia tridentata often have very low purity with a wide range of achene size and condition.   Purity methods vary among labs, causing frustration and confusion for those trying to compare results.  The main point of disagreement is the application of the pure seed definition.   This referee was conducted to compare the application of  ISTA and AOSA pure seed definitions and their subsequent effects on viability, pure live seed, and pure seed units per gram. 

Materials and Methods: 

Two lots of Artemisia tridentata seed were obtained and divided by the Wyoming Seed Laboratory.  Two 1/3 purity weights (0.25g.) were prepared for each lot.  It was hoped that smaller sample sizes would increase participation.   Ransom Seed Laboratory assigned and mailed the samples.  NCGRP compiled and evaluated the data.  Twelve analysts completed sample 1.  14 analysts completed sample 2.

SAS, PROC GLM procedure used to analyze data. 

Pure seed definitions:

AOSA

2.7 Kind or cultivar considered pure seed. - …
a. Immature or shriveled seed units, and seed units that are cracked or otherwise damaged.
b. Insect-damaged seeds, provided that the damage is entirely internal, or that the opening in the seed coat is not sufficiently large to allow the size of the remaining mass of tissue to be readily determined.
c. Seeds that have started to germinate
d. Seed units with nematode galls, fungus bodies (i.e. ergot, smut, etc.) and spongy or corky caryopses, which are entirely enclosed within the seed unit.  Refer to sections 2.10 a (1) and c (1). 
 
Artemisia  PSU 27:
Intact achene whether or not a seed is present.
Piece of broken achene larger than one-half of the original size, unless no seed is present.
Seed, with or without pericarp/seed coat.
Piece of broken seed, with or without pericarp/seed coat, larger than one-half the original size.

ISTA

Pure seed definition:
3.2.1.1.  The following structures (even if immature, undersized, shriveled, diseased or germinated, providing they can be definitely identified as of that species) unless transformed into visible fungal sclerotia (see 3.5.2.A.5.B. for exceptions when the uniform blowing method is used), smut balls or nematode galls:
            1. Intact seed units (= commonly found dispersal units i.e. achenes and similar fruits, schizocarps, florets etc) as defined for each genus or species in the Pure Seed Definitions in the current Annexe 3.2.1.A. 
            2.  Pieces of seed units larger than one-half their original size.
 
Annexe 3
Artemisia   PSD 1:
Achene, unless it is obvious that no seed is present.
Piece of achene larger than one-half the original size, unless it is obvious that no seed is present. 
Seed with the pericarp/testa partially or entirely removed. 
Piece of seed larger than one-half the original size, with the pericarp/testa partially or entirely removed.
 
Results:
   Using a 5% probability of error, statistically significant differences between ISTA and AOSA results were only found in PSU results for sample 2.  AOSA and ISTA results were not significantly different for any other test on these two samples.
   Using a 10% probability of error, statistically significant differences between ISTA and AOSA results were found for sample1 viability and PSU results. In sample 2, differences were significant between ISTA and AOSA for purity, viability, and PSU/gram results.   No statistically significant differences between AOSA and ISTA methods were observed for PLS results for either sample. 
   It is possible that insufficient sample sizes were tested to examine this issue adequately.  However, larger sample sizes would likely reduce participation.  It is unclear whether the wording for ISTA pure seed would improve consistency of pure seed determination. 

2004 Artemisia referee answers to questions:

Germination test conditions:
     All respondents:  20C, 21 days, TB
     11 respondents mentioned light,   (4 respondents: 8 hrs,  1 respondent: 11 hrs, 1 respondent: 16 hrs.)

Separation techniques:
·          screens, microscope, bottom light
·          sieves and hand
·          I used the general blower to make the hand separation a little easier.  Did not use any kind of blowing point.  Used diaphanoscope some with ISTA samples only.
·          Screening (size 0.686) first then hand separation with microscope
·          Over Hoffman sieve cups (range 3-7) lost less than blower, BW, sieve cups, Blower, hand separation
·          Hand separation only, stereoscope
·          Hand separation using magnifying light, then scoping to check
·          Hand separation with 7x eye lens
·          Hand separation
·          Blower, hand separation
·          Hand separation only under a dissecting microscope
·          Hand separation using a dissecting microscope
 

Acknowledgements:

Many thanks to the Wyoming Seed Laboratory and Ransom Seed Laboratory.
Julie Laufmann (NCGRP) and Mark West (USDA/ARS Northern Plains Area Office)  assisted with statistical analysis.   Thanks very much to all the participants.  Your time and effort was greatly appreciated. 
 
Annette Miller, USDA/ARS NCGRP, 1111 South Mason St., Fort Collins, CO 80521-4500
970 495 3240, almiller@lamar.colostate.edu

2004 Artemisia referee data
Gray cells indicate the range of results within one standard deviation of the
mean.  For a 5% error rate (Pr > F = 0.05 or less) only PSU's for sample 2
showed a significant difference between AOSA and ISTA methods.  For a
10% error rate (Pr > F = 0.10 or less) significant differences between AOSA
and ISTA methods were seen in Purity results for sample 2, Viability results
for samples 1 and 2, and PSU values for samples 1 and 2.  No significant
differences between methods were evident for the PLS values of either sample.
Purity








Lab #   ISTA 1   Lab #   AOSA 1
Lab #   ISTA 2   Lab #   AOSA 2
24 12.55
25 10.74
25 20.53
25 29.06
11 12.78
26 11.68
18 26.98
27 29.25
26 13.23
13 12.77
26 27.31
18 29.37
16 13.42
27 12.92
11 27.63
26 31.25
12 13.84
24 14.01
27 29.24
1 31.47
21 13.91
10 14.09
4 30.53
12 32.02
median 13.94
21 14.09
mean 30.95
24 33.00
13 13.98
median 14.09
16 31.94
median 33.00
27 14.02
mean 14.32
median 32.00
21 33.19
mean 14.25
16 14.75
12 33.44
10 33.67
10 14.29
11 14.86
6 33.61
13 33.71
25 14.78
12 15.08
1 33.88
mean 33.93
4 15.97
6 17.76
10 34.76
4 35.80
6 18.33
4 19.11
21 34.80
11 38.45
 


 
13 35.00
16 39.11
N = 12
N= 12
24 36.07
6 43.30
SD = 1.56
SD = 2.33
 

 
 


 
N = 14
N = 14
Pr > F = 0.9386
SD = 4.28
SD = 4.08
not significant
Pr > F =  0.0707
         
if 5%, then not significant






if 10%, then differences significant
Viability








Lab #   ISTA 1   Lab #   AOSA 1
Lab #   ISTA 2   Lab #   AOSA 2
16 3
16 3
16 56
4 42
26 8
6 8
6 57
24 45
24 10
13 8
24 59
12 46
13 11
4 9
12 61
1 51
6 13
11 9
10 61
26 52
21 13
21 9
4 63
10 56
mean 13.5
median 10
mean 64
11 57
median 14
mean 10.3
26 66
median 57
25 15
25 11
median 67
18 57
27 15
26 11
27 68
mean 58
4 16
12 12
13 69
6 63
11 17
10 13
25 69
13 63
10 18
24 13
1 71
27 63
12 24
27 18
11 73
16 66
 


 
21 73
21 66
N = 12
N = 12
18 79
25 75
SD = 5.33
SD = 3.65
N = 14
N = 14
 


 
SD = 7.73
SD = 9.86
Pr > F = 0.0955
Pr > F = 0.0885
if 5%, then not significant
if 5%, then not significant
if 10%, then differences significant
if 10%, then differences significant











Pure Live Seed (PLS)




Lab #   ISTA 1   Lab #   AOSA 1
Lab #   ISTA 2   Lab #   AOSA 2
16 0.40
16 0.44
11 10.17
12 14.72
26 1.06
21 1.00
25 14.17
4 14.78
24 1.26
13 1.02
16 17.89
24 14.85
13 1.54
25 1.18
26 18.02
1 16.00
mean 1.96
26 1.28
mean 19.11
26 16.25
21 2.00
11 1.34
4 19.16
18 16.74
27 2.10
median 1.54
6 19.16
27 18.56
median 2.13
4 1.74
27 19.72
median 18.60
11 2.17
12 1.81
median 20.00
10 18.85
25 2.22
24 1.82
12 20.40
mean 19.92
6 2.38
10 1.83
10 21.20
13 21.24
4 2.54
27 2.34
24 21.28
25 21.80
10 2.57
mean 2.50
18 21.31
11 22.00
12 3.32
6 14.21
1 24.00
21 22.00
 


 
13 24.15
16 25.81
N = 12
N = 12
21 25.00
6 27.28
SD = 0.78
SD = 3.72
 


 
 


 
N = 14
N = 14
Pr > F = 0.6293
SD = 3.85
SD = 4.16
not significant
 


 






Pr > F = 0.6004






not significant











Pure Seed Units (PSU) per gram

Lab #   ISTA 1   Lab #   AOSA 1
Lab #   ISTA 2   Lab #   AOSA 2
11 709
25 620
25 526
25 774
24 736
26 664
18 701
21 843
21 776
27 750
11 723
27 876
27 800
13 783
26 765
18 908
16 808
21 829
27 818
13 952
13 824
24 855
21 898
26 1019
median 827
median 894
1 909
16 1057
26 831
mean 920
mean 909
mean 1065
12 841
10 934
median 912
median 1065
mean 845
11 985
4 915
1 1073
10 857
16 1035
12 958
10 1118
25 875
12 1178
6 980
11 1175
4 925
4 1180
13 983
6 1203
6 1158
6 1238
24 1088
24 1239
 


 
10 1126
12 1282
N = 12
N = 12
16 1185
4 1556
SD = 114
SD = 206
N = 14
N = 14
 


 
SD = 185
SD = 212
Pr > F = 0.2774
Pr > F = 0.0498
not significant
significant at 5% rate








2004 Artemisia photo referee
referee


Artemisia photo referee 2004 

Photos taken and results compiled by Annette Miller. 

31-33 respondents (not everyone responded for every photo)

Respondents were asked to classify each photo numbered above as Pure or Inert according to the AOSA definition and then again according to the ISTA definition.  

AOSA

2.7 Kind or cultivar considered pure seed. - …
a. Immature or shriveled seed units, and seed units that are cracked or otherwise damaged.
b. Insect-damaged seeds, provided that the damage is entirely internal, or that the opening in the seed coat is not sufficiently large to allow the size of the remaining mass of tissue to be readily determined.
c. Seeds that have started to germinate
d. Seed units with nematode galls, fungus bodies (i.e. ergot, smut, etc.) and spongy or corky caryopses, which are entirely enclosed within the seed unit.  Refer to sections 2.10 a (1) and c (1). 
 
Artemisia  PSU 27:
Intact achene whether or not a seed is present.
Piece of broken achene larger than one-half of the original size, unless no seed is present.
Seed, with or without pericarp/seed coat.
Piece of broken seed, with or without pericarp/seed coat, larger than one-half the original size.
___________________________________________________________________________________

ISTA  Pure seed definition:
3.2.1.1.  The following structures (even if immature, undersized, shriveled, diseased or germinated, providing they can be definitely identified as of that species) unless transformed into visible fungal sclerotia (see 3.5.2.A.5.B. for exceptions when the uniform blowing method is used), smut balls or nematode galls:
            1. Intact seed units (= commonly found dispersal units i.e. achenes and similar fruits, schizocarps, florets etc) as defined for each genus or species in the Pure Seed Definitions in the current Annexe 3.2.1.A. 
            2.  Pieces of seed units larger than one-half their original size.
 
Annexe 3
Artemisia   PSD 1:
Achene, unless it is obvious that no seed is present.
Piece of achene larger than one-half the original size, unless it is obvious that no seed is present. 
Seed with the pericarp/testa partially or entirely removed. 
Piece of seed larger than one-half the original size, with the pericarp/testa partially or entirely removed.
 

Higher agreement was achieved more often using the AOSA definition than the ISTA definition. Green and yellow boxes are full agreement.  Orange boxes indicate reduced agreement.

IMAGE #

AOSA

ISTA

 

AOSA

ISTA

1

33 P

3 I
29 P

 

P 100%

I 9%
P 91%

2

33 P

5 I
27 P

 

P 100%

I 16%
P 84%

3

2 I
31P

15 I
17 P

 

I 6%
P 94%

I 47%
P 53%

4

2 I
31P

12 I
20 P

 

I 6%
P 94%

I 37%
P 63%

5

7 I
26 P

18 I
13 P

 

I 21%
P 79%

I 58%
P 42%

6

33 P

2 I
30 P

 

P 100%

I 6%
P 94%

7

12 I
21 P

15 I
17 P

 

I 36%
P 54%

I 47%
P 53%

8

3 I
30 P

10 I
22 P

 

I 9%
P 91%

I 31%
P 69%

9

33 P

4 I
27 P

 

P 100%

I 13%
P 87%

10

31 I
2 P

28 I
3 P

 

I 94%
P 6%

I 90%
P 10%

11

33 P

2 I
29 P

 

P 100%

I 6%
P 94%

12

33 P

8 I
24 P

 

P 100%

I 25%
P 75%

13

31 I
2 P

32 I

 

I 94%
P 6%

P 100%

14

33 P

9 I
23 P

 

P 100%

I 28%
P 72%

15

2 I
30 P

6 I
26 P

 

I 6%
P 94%

I 19%
P 81%

16

2 I
31 P

6 I
26 P

 

I 6%
P 94%

I 19%
P 81%

17

8 I
25P

19 I
12 P

 

I 24%
P 76%

I 61%
P 39%

18

3 I
30 P

4 I
28 P

 

I 9%
P 91%

I 13%
P 87%

19

33 P

32 P

 

P 100%

P 100%

20

12 I
21 P

23 I
9 P

 

I 36%
P 54%

I 72%
P 28%

21

33 P

1 I
31 P

 

P 100%

I 3%
P 97%


Some possible reasons for the results:
1.  The nature of a photo referee.
One answer sheet had the following comment:
"Difficult to evaluate under ISTA rules since you can't determine if a seed is present or not when looking at a photo."  The photo quality could have contributed to the evaluation difficulty.  It's difficult or impossible to see translucence in these photos.
2.  AOSA instructions leave less room for questionable judgements.  The intention of the wording is clearer.
3.  People are less familiar with the intention of the ISTA wording than they are with the AOSA definition.  Or, they may not have sufficient training to make a judgement that matches the intention of the wording for this pure seed unit (More of a problem with ISTA than AOSA).  One analyst told me on the phone that he did not think there was any difference between the two definitions.