Niche modelling software can be used to assess the probability of detecting a population of a plant species at a certain location. In this study, we used the distribution of the wild relatives of lettuce (Lactuca spp.) to investigate the applicability of Maxent species distribution models for collecting missions. Geographic origin data of genebank and herbarium specimens and climatic data of the origin locations were used as input. For Lactuca saligna, we varied the input data by omitting the specimens from different parts of the known distribution area to assess the robustness of the predicted distributions. Furthermore, we examined the accuracy of the modelling by comparing the predicted probabilities of population presence against recent expedition data for the endemic Lactuca georgica and the cosmopolitan Lactuca serriola. We found Maxent to be quite robust in its predictions, although its usefulness was higher for endemic taxa than for more widespread species. The exclusion of occurrence data from the perceived range margins of the species can result in important information about local adaptation to distinct climatic conditions. We discuss the potential for enhanced use of Maxent in germplasm collecting planning.

Genetic variation in Lactuca serriola, the closest wild relative of cultivated lettuce, was studied across Europe from the Czech Republic to the United Kingdom, using three molecular marker systems, simple sequence repeat (SSR, microsatellites), AFLP and nucleotide-binding site (NBS) profiling. The ‘functional’ marker system NBS profiling, targeting disease resistance genes of the NBS/LRR family, did not show marked differences in genetic diversity parameters to the other systems. The autogamy of the species resulted in low observed heterozygosity and high population differentiation. Intra-population variation ranged from complete homogeneity to nearly complete heterogeneity. The highest genetic diversity was found in central Europe. The SSR results were compared to SSR variation screened earlier in the lettuce collection of the Centre for Genetic Resources, the Netherlands (CGN). In the UK, practically only a single SSR genotype was found. This genotype together with a few other common SSR genotypes comprised a large part of the plants sampled on the continent. Among the ten most frequent SSR genotypes observed, eight were already present in the CGN collection. Overall, the CGN collection appears to already have a fair representation of genetic variation from NW Europe. The results are discussed in relation to sampling strategies for improving genebank collections of crop wild relatives.

Mean δ15N of whole-plants of lettuce and barley varied by [les ]3‰ when given a chemically and isotopically uniform N source. This variation was related to the presence, absence and species of arbuscular mycorrhizal (AM) fungi and to external N concentration. A highly AM-susceptible plant (lettuce) responded to treatments differently than a less susceptible one (barley). The largest change in whole plant δ15N was related to the experimental combination most likely to be found in field conditions: species of fungus interacting with varying external concentrations of N.

The mechanisms underlying observed variations of plant δ15N are not understood, nor can they be approached directly using δ15N. However, descriptive data, such as presented here, are important to the development of a mechanistic model, because they suggest relationships for future research, using inter alia15N-enriched tracers. They also confirm that plant N sources, cannot be identified using plant δ15N, even when the type of mycorrhizal association (endo vs. ecto) is known.