19
BLACK SEA PHYTOPLANKTON DATA QUALITY- PROBLEMS AND PROGRESS Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No 40, P.O.Box 152, Bulgaria 2HCMR P.O. BOX 712, Anavissos 19013, Greece 3P.P Shirshov IO-RAS, 36, Nahimovski prospect, Moscow, Russia, 117997 4NIMRD “Grigore Antipa”, Mamaia bul., No 300, Constanta 3 , RO-900581 5IBSS,2 Nakhimov Ave. Sevastopol, 99011 Ukraine Black sea OUTLOOK Conference, Odessa, 01-04 October , 2011

Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

Embed Size (px)

Citation preview

Page 1: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

BLACK SEA PHYTOPLANKTON DATA QUALITY- PROBLEMS AND PROGRESS

Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5

1Institute of Oceanology-BAS, Parvi Mai str., No 40, P.O.Box 152, Bulgaria 2HCMR P.O. BOX 712, Anavissos 19013, Greece 3P.P Shirshov IO-RAS, 36, Nahimovski prospect, Moscow, Russia, 117997 4NIMRD “Grigore Antipa”, Mamaia bul., No 300, Constanta 3 , RO-900581 5IBSS,2 Nakhimov Ave. Sevastopol, 99011 Ukraine

Black sea OUTLOOK Conference, Odessa, 01-04 October , 2011

Page 2: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

The quality of biological data has gained a recognition as an essential part of international monitoring programmes, in response to the demand for strategic environmental evaluations such as the EU WFD, the MSFD etc. Informed decisions for environmental sound management can be made only on the basis of reliable data, and therefore certain level of data quality should be achieved to assure accuracy and precision of all measurement systems.

The structural characteristics of phytoplankton communities bear valuable information about the evolution of microalgal communities and the trajectories of shifts under multiple environmental factors, including anthropogenic impacts. Details of phytoplankton analytical procedures are essential to compare data produced by different analysts either during a long-term monitoring programs in one area or between different areas in order to evaluate statistically significant long-term trends or spatial differences.

Page 3: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

OBJECTIVES

I - present results from the intercalibration excercise assess the degree of comparability/differences in phytoplankton

and chlorophyll a data analysis among 4 Black sea laboratories and where possible to make recommendations for further

improvement and harmonization of research methodology in the Black Sea .

II present actions taken towards solutions

The expectation is that the results will assist regional assessments based on combined data sets

“Quality control of phytoplankton counting and biovolume estimates—an easy task or a Gordian knot?” E.Rott et all., 2007

Page 4: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

MATERIALS AND METHODSSamples from two stations (coastal and open sea) were collected and distributed for laboratory analysis in 3 replicates for each partner (SESAME intercalibation cruise-Apri’2009)

Map of sampling stations

Phytoplankton attributes: Phytoplankton Total abundance [cells/l] Phytoplankton Total biomass (wet weight) –

[mg/m3] Phytoplankton common Taxonomic classes

(Bacillariophyceae, Dinophyceae, Prymnesiophyceae, Cryptophyceae and Small flagellates ).

Chlorophyll a measurements

Participants

IBSS-Sebastopol - UkRNIMRD – ROIO-BAS – BGIO-RAS - RU

Statistical analysis Robust statistical treatment (ANOVA test, Tukey test and a Lavene statistic) was applied in order to check the homogeneity of the variance between the groups. Stock plots were designed based on averages and standard deviation and the the Bray Curtis similarity index Common statistics employed during phytoplankton ring tests (average ±1 stdev and CV < 20 %).

Page 5: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

ParticipantSample concentration

Counting chamber Type of microscope

Volume of subsample

No cells counted per sample

IO-BAS Decantation UtermolSegwick Rafter Utermol

Nicon inverted+ immage analysis 1 ml 400 cells

 NIMRD-  Decantation Utermol  Utermol  Inverted 0.1 ml  

IO-RUSDecantation/inverse filtration

Nogott’s up to 0.1 mlNauman’s chamber 1 or 5 ml Light compound 1 ml

IBSS Decantation

Nauman’s chamber 1 ml0.05 ml Light compound 0.1 ml

Inventory of in-house lab methods

Partner Extraction Sample preparation InstrumentEquations reference

IO-BAS 90% acetone24 hours extraction 7000 rpm cuvette 1cm L Spectrophotometer

Jeffrey and Humphrey (1975)

NIMRD 90% acetone24 hours extraction 4000 rpm cuvette 1cm L Spectrophotometer

SCOR UNESCO (1968)

IO-RAS 90% acetone   Fluorimeter  

Phytoplankton counts

Chlorophyll a measurements

Page 6: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

RESULTS

Stock plot of the total phytoplankton abundance [cells/l - average and stdev] by partners

Analysis of variance:

SourceDF

Sum of squares

Mean squares F Pr > F

Model 2 0,028 0,014 31,233 0,004

Error 4 0,002 0,000Corrected Total 6 0,030      

The statistical treatment of total abundance data show significant differences between the Ukrainian-Bulgarian results and between the Ukrainian - Russian results, while the difference between Bulgarian and Russian data were not significant

Countries / Tukey (HSD) / Analysis of the differences between the categories with a confidence interval of 95%:

ContrastDifference

Standardized difference

Critical value

Pr >

DiffSignificant

UKR vs. RUS 0,151 7,748 3,564

0,003 Yes

UKR vs. BLG 0,118 5,548 3,564

0,011 Yes

BLG vs. RUS 0,032 1,671 3,564

0,321 No

Tukey's d critical value: 5,04

Similarity cluster matrix of total phytoplankton abundance [cells/l - square root transformation] by labs

Hierarchical clustering showed similarity between Bulgarian and Russian data > 85% while between Ukrainian - Bulgarian and Ukrainian - Russian was > 75% )

Phytoplankton abundance

Page 7: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

Countries / Tukey (HSD) / Analysis of the differences between the categories with a confidence interval of 95%:

ContrastDifferen

ceStandardized

differenceCritical value

Pr > Diff

Significant

BLG vs. UKR 1,417 5,530 3,564

0,011 Yes

BLG vs. RUS 0,199 0,852 3,564

0,695 No

RUS vs. UKR 1,217 5,206 3,564

0,014 Yes

Tukey's d critical value: 5,04

Countries / Tukey (HSD) / Analysis of the differences between the categories with a confidence interval of 95%:

ContrastDifference

Standardized difference

Critical value

Pr > Diff

Significant

UKR vs. RUS 0,762 7,987 3,564

0,003 Yes

UKR vs. BLG 0,639 6,113 3,564

0,008 Yes

BLG vs. RUS 0,123 1,291 3,564

0,471 No

Tukey's d critical value: 5,04

Countries / Tukey (HSD) / Analysis of the differences between the categories with a confidence interval of 95%:

ContrastDiffere

nceStandardized

differenceCritical value

Pr > Diff

Significant

BLG vs. UKR 1,417 5,530 3,564

0,011 Yes

BLG vs. RUS 0,199 0,852 3,564

0,695 No

RUS vs. UKR 1,217 5,206 3,564

0,014 Yes

Tukey's d critical value: 5,04

Phytoplankton abundance

Page 8: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

Based on testing the reproducibility of the in-house analysis (replicates) and employing the CV < 20% assumption for the total numerical abundance the results reveal a good reproducibility of the in-house replicates and very close results between the different labs, with the exception of Ukraine, where the difference was between 25-30%

Lab AverageN [cells/l]

stdev CV%

BLG 3520319 167824.6 4.8

RUS 3268008 178086.8 5.4

UKR 4620032 160602.3 3.5

RO 3253500

all 3665465 648061.9 17.7

BG/RO 3386909 188669.4 5.6

BG/RO/RUS 3347276 150035.3 4.5

Average abundance [cells/l] stdev and CV [%] by partners

Average BAC [cells/l]

stdev CV%

RU 2318659 261722 11.3BG 2942261 26535 7.0RO 3253500UK2 3787862 294579 7.8All 3075570 613651 20.0

LAB Dinophyceae average {cells/l]

stdev CV%

Prymnesiophyceae Average [cells/l]

stdev CV% Small flagellates Average [cells/l]

stdev CV%

RU 7803 1173 15.0 46913 10611 22.6 753600 421620 55.9BG 9913 2342 23.6 53026 5840 11.0 347713 142143 40.9RO 20600 67350 600UK 54495 9180 16.8 118422 18608 15.7 65790 55825 84.9All 23203 21601 93.1 71428 32479 45.5 291926 342668 117.4

The comparison of phytoplankton abundance results by taxonomic classes reveal compliance to the 20% CV only for Bacillariophyceae, while for the other classes the differences among the participating labs are substantial, especially critical for the small flagellates, where even the in-house results show inconsistencies for all partners

Phytoplankton abundance

Page 9: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

Stock plot of the total phytoplankton biomass [mg/m3 - average and stdev] by partners

Countries / Tukey (HSD) / Analysis of the differences between the categories with a confidence interval of 95%:

ContrastDifferen

ceStandardized

differenceCritical value

Pr > Diff

Significant

UKR vs. RUS 0,192 3,319 3,564

0,062 No

UKR vs. BLG 0,115 1,809 3,564

0,278 No

BLG vs. RUS 0,078 1,338 3,564

0,450 No

Tukey's d critical value: 5,04

Similar to the total abundance the hierarchical clustering of total biomass showed similarity between Bulgarian and Russian data > 85% while between Ukrainian - Bulgarian and Ukrainian - Russian was > 75% )

Phytoplankton total biomass

0

500

1000

1500

2000

2500

3000

3500

RU BG RO UK

B[m

g/m

3]

Phytoplankton biomass

Page 10: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

  Average B [mg/m3] stdev CV%RU 1935.736 372 19.2BG 2284.548 103 4.5RO 2861.63    UK 2974.873 4 0.16All 2514.197 489 19.5

LAB Bacillariophyceae Av B [mg/m3]

stdev CV% Dino-phyceae Av B [mg/m3

stdev CV% Prymnesiophyceae Av B [mg/m3

stdev CV% Small flagellates Av B [mg/m3]

stdev CV%

RU 1620.711 420 25.9 111.800 5 4.5 9.098 3 28.0 58.872

22 37.1

BG 2007.946 150 7.5 88.630 12 13.9 11.775 1 8.5 93.118

38 40.9

RO 2607.660 228.330 9.770 0.190

UKR 2614.258 86 3.3 297.024 95 32.0 52.264 15 27.8 4.263 4 84.7

All 2212.644 486 22.0 181.446 98 54.2 20.727 21 101.6 39.111 45 114.7

Similar to the phytoplankton abundance at the level of taxonomic classes the differences among the participating labs are substantial, especially critical for Prymnesiophyceae and small flagellates, where even the in-house results show inconsistencies for all partners. Albeit the good agreement between the data among some of the labs this is not systematic for all the taxonomic groups

Phytoplankton biomass

Page 11: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

Bacillariophyceae

0

10000

20000

30000

40000

50000

60000

0 2 4 6 8 10species

Bio

vo

lum

e

[mkm

3]

304-RU

304-BG

304-RO

304-UK

Dinophyceae

0

20000

40000

60000

80000

100000

120000

0 2 4 6 8 10species

Bio

vo

lum

e

[mkm

3]

304-RU

304-BG

304-RO

304-UK

Bacillariophyceae

1.Cerataulina pelagica 2.Chaetoceros socialis3.Chaetoceros curvisetus 4. Nitzschia tenuirostris 5. Proboscia alata 6. Pseudo-nitzschia p-delicatissima 7. Skeletonema costatum8.Thalassionema nitzschioidesDinophyceae

Ceratium fusus -1Gyrodinium fusius -2Heterocapsa triquetra -3Prorocentrum compressum -4Prorocentrum micans -5Protoperidinium bipes -6Protoperidinium granii -7Scrippsiella trochoidea -8

PrymnesiophyceaeEmiliania huxleyiSmall flagellates

For the common species biovolume comparative analysis reveal the differences are substantial, for the most abundant species such as Pseudo-nitzschia delicatissima and Emiliania huxley the biovolume varies more than twice (202-409 mkm3 and 145- 268 mkm3 respectively ) for some species the differences exceeding 3 fold

 Volume of subsample No species

RUS 1 57BG 1 59RO 0.1 39UKR 0.1 31

As expected the analyzed sub-sample volume is important for the species diversity (no of species) recorded in the samples

Page 12: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

The results of chlorophyll a measurements reveal good in-house reproducibility for BG and RO and higher than 10% difference for RU lab - Table 18. The difference between the BG and RO data is within the (range average ±s 1stdev)

Chlorophyll a - st. 301 - RO, RU, BG

0

2

4

6

8

10

12

0 1 2 3 4 5

chl.

a [m

kg/l]

BG

Ro

Rus

Chlorophyll a - st.304

0.00

1.002.00

3.00

4.00

5.006.00

7.00

8.00

0 1 2 3 4 5

chl.

a [

mk

g/l

]

BG

Ro

Rus

Station BG RO RUSS-BG01-05 (M301) 7.53 8.08 7.53

S-BG01-05 (M301) 7.13 7.85 6.11

S-BG01-05 (M301) 7.97 10.19 7.92

average 7.54 8.70 7.19

stdev 0.42 1.29 0.95

CV% 5.6 14.8 13.3

S-BG01-13 0.61 0.68 0.46

S-BG01-13 0.61 0.72 0.59

S-BG01-13 0.61 0.68 0.38

average 0.61 0.69 0.48

stdev 0.00 0.02 0.15

CV% 0.00 3.06 30.41

S-BG01-08 (M304) 6.33 6.24 4.18

S-BG01-08 (M304) 6.32 6.29 3.82

S-BG01-08 (M304) 6.86 6.84 4.71

average 6.50 6.46 4.24

stdev 0.31 0.33 0.45

CV% 4.75 5.16 10.50

Chlorophyll a [mg/m3]

Page 13: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

• For the total phytoplankton abundance the results between Bulgaria, Russia and Romania are comparable (insignificant differences) while the difference with Ukrainian lab is between 25-30% • For the total phytoplankton biomass there is a good agreement between Romania and Ukraine, about 20% (acceptable) difference between Bulgaria and all other labs and a 30% difference between Russia, Romania and Ukraine.• Both for the phytoplankton abundance and biomass at the level of taxonomic classes the differences are substantial especially for Prymnesiophyceae and Small flagellates. If taxonomic classes data would be used as one data set they should be treated with caution.• The difference of chl. a results between BG and RO is about 10% and the data sets could be comparable. The BG and RO measurements are between 1.3-1.5 higher than the RUS results.

The results give ground for the following recommendations:A phytoplankton check-list with unified geometric shapes for the different species is essential to avoid differences in the species biovolume estimation that reflect the final biomass resultsAnalysis of at least 1 ml counting chamber is highly recommended to better detect species diversity In order to avoid taxonomic miss-match WoRMS taxonomy is mandatory

CONCLUSSIONS and RECCOMENDATIONS

Page 14: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

BLACK SEA COMMISSIONUP-GRADE BS SCENE PROJECT +

Phytoplankton expert group

Phytoplankton Workshop, Istanbul 21-23 June 2010

Country Organisation  Expert Name Contact detailsBulgria IO-BAS-BG Assoc. Prof. Snejana Moncheva [email protected]

Romania NIMRD – ,RO Dr. Laura Boichenko [email protected]

Russia IO P.P.Shirshov, RAS, Dr. Alexander Mikaelyan [email protected]

Turkey ,inop,Turkey Mr. Fatih Sahin [email protected]

Ukraine IBSS, Sebastopol Mrs. Oleksandra Sergeeva [email protected]

Ukraine IBSS, Sebastopol Dr. Vladimir Vladimirov [email protected]

Ukraine IBSS, Sebastopol Dr. Yuliya Bryantseva [email protected]

Ukraine IBSS, Sebastopol Mr. Denis Slipetskyy [email protected]

Ukraine Botanical Institute, Kiev Dr. Alexander Krahmalnii [email protected]

Ukraine Odessa University, Odessa Mrs. Natalia Dereziuk [email protected]

Black Sea Commission Permanent Secretariat

Prof. Ahmet Kideys [email protected]

Black Sea Commission Permanent Secretariat

Dr. Violeta Velikova   [email protected]

Black Sea Commission Permanent Secretariat

Mr. Vladimir Myroshnychenko [email protected]

Black Sea Commission Permanent Secretariat

Ms. Nilufer Akpinar [email protected]

Page 15: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

CHECK LIST OF BLACK SEA PHYTOPLANKTON, REFERENCE BIOVOLUMES and DATA BASE

Software (Oleksandra Sergeyeva, Kseniia Skuratova IBSS, Sebastopol, Ukraine)The special software for creation of online marine species checklists was developed. This software is based on wiki engine and has special developed functions which make it easy to add, delete, move species and add any type of structured information in the form of patterns, which can be easily added by the checklist administrator on request of users.Each species has the corresponding page, where all information is placed either in form of predefined patterns or in the form of text, images, tables etc.

on-line: http://phyto.bss.ibss.org.ua

Biovolume, shapes etc.(Bryantsteva Y., IBSS, Sebastopol, Ukraine)Efforts have been made to create one reference list of biovolumes for Black Sea microalgae. Thus for each species in the checklist the appropriate suggested figure to calculate biovolume was attached. For detailed research of morphometric characteristics of the community the more precise figure is also suggested where possible.

on-line: http://phyto.bss.ibss.org.ua

Page 16: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

UPGRADE BLACK SEA SCENE

GA 226592, FP7, EC

GUIDELINES FOR QUALITY CONTROL OF BIOLOGICAL DATA- PHYTOPLANKTON

Snejana Moncheva November 2010

1 INTRODUCTION 4

2 THE QUALITY SYSTEM FOR BIOLOGICAL DATA - QA/QC 4

3 QA/QC FOR FIELD SAMPLING - SAMPLE PRESERVATION AND STORAGE 6

Equipment Sampling Protocol Sample Preservation Sub-Sampling – Validation of Homogenization

4 QUALITY ASSESSMENT FOR LABORATORY ANALYSIS 6

Taxonomy Cell Counts Biovolume/Biomass Estimation Re-Analysis Repeatability and Reproducibility Uncertainty Control Charts for Biological Measurements Training and Inter-Laboratory Comparability Testing

5 QUALITY ASSURANCE OF DATA REPORTING 12

Documentation Data management MetaData Reporting Form 6. DATA FLAGGING SYSTEM 14

7. CONCLUSIONS 15

8. REFERENCES 16

9. LIST OF PARTICIPANTS 17

Page 17: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

PHYTOPLANKTON MANUAL - updated

Page 18: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

THE WAY FORWARD

Finalise the check-list

Ph

yto

pla

nkt

on

exp

ert

gro

up

Finalise the automated system

Apply the QC/QA guidelines to real data sets

Finalise the data-base format

Conduct ring testsMaintain the BS Commission web

site operational !!!!

Page 19: Snejana Moncheva1, Maria Pantazi2, Larisa Pautova3, Laura Boichenko4, Dan Vassiliu4,Luydmila Mantzosh5 1Institute of Oceanology-BAS, Parvi Mai str., No

THANK YOU FOR THE ATTENTION

Participants in theSESAME intercalibration cruise