Organic Outdoor TomatoesRegional screening and breeding in Germany

Many thanks to the Rut- and Klaus-Bahlsen Foundation, the German Federal Organic Farming Scheme and our colleagues on farm and in garden!

Results

After two years of screening the share of commer-cial varieties was 9.1%, and 31.8% were obtained from genebanks. 59.1% of the most interesting va-rieties originated from NGOs and individual seed savers. During four years of screening and regional tests the top ten varieties were identified. Included are wild-, cocktail-, salad-, processing-, and beef-steak-tomatoes (see pictures). The high heritability of Phytophthora-infections on leaves and fruits, and for yield (74-96%) confirmed the suitability of the experimental design (Table 1). For all traits the varieties were the most important variance component. Site specific adaptation, i.e. variety x location interaction, was of minor impor-tance or absent. For fruit infections variety x year interactions were stronger than variety x location interactions.

Introduction

Tomatoes are the most important vegetable in Germany, but more than 90% of the consumption are imported. One of the major obstacles for increased production is the breakdown of outdoor tomato production. Increasing infections with late blight (Phyto-phthora infestans) have almost eliminated commercial cropping and seriously impaired private production. The aim of the project during the years 2003-2006 was 1) to evaluate and recommend suitable open pollinated varieties for organic outdoor production and 2) develop a regional breeding program within organic horticulture.

Materials and methods

Regional evaluation was based on 3500 accessions. In close contact with genebanks, NGOs, seed trade, and private seed savers, 92 varieties were selected for comparative trials at three farms in central and northwestern Germany. The number of varieties was reduced during 2003 to 2006. Some additional varieties with superior performance in a screening at one farm were included. Two replications with two plants (2006: 3 x 2) were grown per location. Layout and maintenance of the plots favoured Phytophthora-infections.Each year a group of 10-12 varieties was tested at 25 to 35 locations. Backyard and market gardens, botanical and other public gardens, and research institutions participated.Suitable varieties identified in the screening were used to make crosses to combine the important traits earlyness, field resistance against late blight, high yield, and good fruit quality. The most advanced crosses were included in the screening 2005-2006.

Bernd Horneburg and Heiko C. BeckerDepartment of Crop Sciences, Georg-August-University, Von Siebold Str. 8, D-37075

Göttingen, Germany, bhorneb@gwdg.de

As an example for the regional breeding program a wide cross between the wild tomato Golden Currant (fruit weight 6 g) and the sa-lad tomato Matina is shown below. The relative value for late blight infections (mean of 2003-2004, Matina = 100) for Golden Currant was 56 for leaf infections, and 50 for fruit infections. Wild tomatoes yield poorly when grown at a stake. One typical way of production is shown below to the left. The important traits field resistance against late blight and long harvest period of the wild tomato were suc-cessfully incorporated in the F5- and F6-generations (Table 2) and resulted in the highest yield in the screening 2006.

Rote Murmel

Golden Currant

Cerise rot

Matina

Celsior Cerise gelb Paprikaförmige

OutlookThe methodology for a regional breeding program is developed and needs to be put into practice! We would appreciate co-operation with other regions facing similar problems.

Table 2: Performance of the parent variety Matina and its cross with the wild tomato Golden Currant in the F4- and F5-generation. Mean values of the years 2005 and 2006 at three farms

Fruit

weight Yield until 15.9.

in g/plant Yield until 15.10.

in g/plant Harvest period

in days Late blight leaf

infection 1) Late blight fruit

infection 1) Matina 56 934 1273 50 282 204 Golden Currant x Matina 32 1145 1791 60 223 162 1) Area under disease progressive curve

Table 1: Variance components for late blight infections and yield at three farms Variance components Years Number

of varieties

Varieties Varieties x

Locations

Varieties x

Years

Varieties x Locations x Years

Herita-bility

Leaf infection 1) 2003-2004 44 1051** 0 266.8** 1053** 86.61 2003-2005 22 1588** 57.02 224.6** 605.0** 94.31 2003-2006 10 1772** 146.8* 76.85 629.5 2) 96.11 2005-2006 17 1993** 206.3* 49.32 373.8 2) 95.82 Fruit infection 1) 2003-2004 44 969.8** 0 354.9** 1291** 81.74 2003-2005 22 1512** 74.31 653.9** 708.5** 85.86 2003-2006 10 784.7** 84.41+ 429.4** 481.1 2) 84.18 2005-2006 17 1284** 77.12 734.1** 483.4 2) 74.15 Yield per plant in g 2005-2006 17 73574** 0 0 186407 2) 79.04 0 indicates negative estimates 1) Area under disease progressive curve 2) The estimate includes a part of the error and was not tested for significance +, *, ** significant at the 0.10, 0.05, 0.01 probability level

X →

Golden Currant x

Matina F5

Golden Currant

Celsior Resi Cerise gelbDeBeraoRote Zora

Paprikaförmige