Bread Wheat Landrace Database

Background

Collaborators

Scientists

The Project

FIGS Sets

Introduction

One of the sub-aims of the project that supported the development of this site was to construct FIGS subsets, of the combined VIR, ICARDA, AWCC wheat landrace collection, that are likely to have a high probability of containing variation for a number of economically important traits. The text below provides a description of how a number of these FIGS subsets were developed. Specific details about the individual accessions that make up these sets can be viewed from the Data Selection screen under the Databases tab of this web-site.

Method to derive the Core set

A representative core collection set was selected from all the landraces using a method that was similar to commonly used techniques to create core sets based on geo-graphic information. This core set was derived by first grouping all accessions from collections sites based on soil type and Agro Climatic Zones (ACZs were based on length of growing period classified as limited by moisture, temperature or both - in-house method).. Sub-groups within the ACZ-soil type groups were then formed based on the country of origin. The number of accessions within each of these country sub-groups determined acceptance into the core set. Where there were 1-20 accessions in a cluster one accession was randomly placed in the core set. Where there were 21 � 60 accessions in a cluster, 5% of the accessions were chosen for the core set at random. Where there were more than 61 accessions in the cluster an additional hierarchical cluster analysis, using site data, was performed with the number of clusters preset at 5% of the total number of accessions in the sub-group. One accession from the resulting clusters was randomly selected for the core set.

Distribution of the core set

Yellow dots represent collection sites - 700 sites, 750 accessions, 52 countries represented.

Method to derive the Drought set

The set was designed to identify those accessions most likely to provide genetic variation for traits that contribute to drought tolerance, resistance and/or avoidance. The following method was employed. Accessions from known irrigation sites were excluded. Next, accessions ACZs, with length of growing periods between 4 and 7 months, and from regions where moisture, or moisture and temperature, are limiting factors to growth were identified. The third selection criterion was for sites where the annual precipitation was in the range of 180-300mm. All remaining accessions were subjected to an hierarchical cluster analysis using all site agro-climatic parameters with a pre-set cluster number of 750. One accession from each of these clusters was randomly chosen to make up the FIGS set.

Distribution of the Drought set

Yellow dots represent collection sites. 350 sites � 750 accessions � 19 countries.

Method to derive the Salinity set

• A map was extracted from the digital Soil Map of the World (FAO-UNESCO, 1995) that shows the probability of the occurrence of saline soils (Fig. 1) after conversion of the original vector map to a 1-km resolution grid.

º Saline soils = Sholenchak soil order for soil units with saline phase (FAO, 1974)

º Definition is not compatible with Soil Taxonomy (USDA, 1999)

• A set of accessions 422 accessions, out of 17,000 was identified by overlaying the collection site coordinates onto the salinity surface (see figure). Accessions were chosen from those sites with a 40% or higher probability of encountering saline soils for laboratory screening. Most of accessions originated from Central Asia and the Indian subcontinent (Table)

Accessions per region from collection sites with a 40% or higher probability of being saline.

Europe	North Africa	West Asia	Central Asia	Asia	Total
6	66	23	198	129	422

Salinity surface with Salt set collection sites overlaid.

Darker blue represents higher probability of encountering saline soils.

Method to derive Powdery Mildew set

The construction of this set used a different approach than above. We sourced information on the collection sites of a set of 400 accession know to have resistance to Powdery Mildew. The agroclimatic information for these sites was combined with same information for sites from the combined landrace collection. The set was developed using the following steps:

• Excluded project set accessions from sites with a low precision for geographic coordinates.

• Excluded project set accessions from sites without agro-climatic data.

• Combined mildew set geographic information with the project set information.

• Retained project accessions from ACZs that contained 20 or more PM resistant accessions (PMr) collection sites. The following ACZs were represented H-C-W, H-K-W, PH-C-W, PH-K-M, PH-K-W, SA-C-W, SH-C-W, SH-K-W. This step retained over 8000 accessions.

• Performed 2 step cluster analysis using all agro-climatic data + ACZ + Length of Growing Period zones � leaving out lat long variables. Set clusters at 200.

• Chose only accessions from those clusters containing PMr accessions � this cut down the set to 4900 accessions.

• Performed Hierarchical Cluster Analysis on all available agro-climatic parameters with clusters set at 400.

• Only chose those accessions from clusters containing PMr accessions were retained - 3028 accessions.

• Performed Principal Components Analysis using all available continuous agro-climatic variables.

• In order to choose a set of material from the Project set �each principal coordinate x, y, z of the Project set were compared with the coordinates of the PMr set � case by case.

• Project accessions were chosen if their first 3 principal coordinates fell within a mathematical box surrounding the 3D coordinates of members of the PMr set. Thus an accession was included in 1 of three sets if all three of the x, y, z principal coordinates fell within 1% or 2% or 3% of the value of the Range for the x, y, z principal coordinates for the combined set (Project set + PMr set). Thus 3 different sets were chosen containing 663, 1323 and 1875 accessions respectively.

Acknowledgement: An important contribution to the development of the Powdery Mildew FIGS set was the kind contribution of Dr Harold Bockelman, who extracted data from the GRIN database (USDA-ARS, National Plant Germplasm System, Germplasm Resources Information Network, online http://www.ars-grin.gov/npgs).

Distribution of Powdery Mildew set

Method to derive Russian Wheat Aphid set

From a journal article by Baker. C.A. and Porter. R. entitled �Recent developments with Russian Wheat Aphid in wheat� we learn that the vast majority (89%) of RWA-resistance that has been identified to date was originally collected in areas of the world where the RAW is endemic, namely the Middle East and eastern European countries. The highest percentage of resistant accessions were those collected in Turkmenistan (23.1%), Kazakhstan (10.8%), and Armenia (9.9%); with surrounding counties having lower percentages. However, the vast majority of resistant accessions were collected in Afghanistan (107). Pakistan (58), and Iran (42).

Further literature and anecdotal evidence suggestions that RWA favor lower humidity and higher altitudes.

In light of the above the following steps were followed to construct the FIGS set:

� Accessions from the CAC, Black Sea, SW Asia, and East Mediterranean countries were chosen. 10566 accessions were identified, 58% of the total project set, representing 25 countries.

� Accessions from step one were filtered further using agro-climatic zones (ACZ) with low humidity, i.e. arid , semi-arid , and sub-humid ; this reduced the number of countries to 20 and cut down the set to 3338 accessions or 18.78% of the total number of Project accessions

The collections sites associated with the remaining accessions were split into 3 sub-regions and accessions were chosen within altitude classes as in the table below. This was done to ensure a broad geographic representation whilst still focusing on environments most likely to yield resistant genotypes.

Filtering by Region and Altitude

Roles	Altitude	Number of accessions	Number of sites
CAC + RUS	1200-2000m	1125 (6.33%)	521
Asian highlands (PKT, AFG, TUR , IND , IRN)	800-2000m
Mediterranean countries	800m

Distribution of Powdery Mildew set