Seeds collected from 10 Dalmatian pyrethrum (Tanacetum cinerariifolium /Trevir./ Sch. Bip. Asteraceae) populations naturally occurring along the Adriatic coast and islands, were used to establish a common garden with the aim of estimating the variation of pyrethrin compounds between and within populations. A recently optimized matrix solid phase dispersion (MSPD) method was used for the extraction of six pyrethrin compounds from dry flower heads of 200 Dalmatian pyrethrum individuals. Separation and quantification of pyrethrin compounds were performed by high performance liquid chromatography with diode array detector (HPLC-DAD). High variability in pyrethrin content and composition was observed within and between populations. Total pyrethrin content in individual samples varied from 0.10% to 1.35% of flower dry weight. On average, the lowest total pyrethrin content was observed in the population of Pelješac (0.22%) and the highest in the island populations of Mali Lošinj and Zlarin (0.87%). The population of Mali Lošinj had the most favourable pyrethrin profile, including the lowest variability in total pyrethrin content, the highest pyrethrin I (PI) content (up to 60.47%) and PI/PII ratio (up to 5.88), as well as the lowest pyrethrin II (PII) content. In contrast, the populations from Biokovo and Pelješac represented pyrethrin profiles with the lowest content of total pyrethrin, PI and PI/PII ratio. The Biokovo population also had the highest PII content (43.18% on average), while the Pelješac population had the highest jasmolin I (JI) and jasmolin II (JII) content. Four different chemotypes were determined by cluster analysis. Chemotype P1 is the most promising for future breeding programs and is characterized by the highest total pyrethrin content, PI content and PI/PII ratio, as important measures of insecticidal activity. Correlation analysis of each pyrethrin compound with 25 environmental variables and Principal Component Analysis (PCA) were carried out and the importance of optimum temperature and precipitation on pyrethrin quantity and quality was suggested. Total pyrethrin content was significantly correlated with six temperature and five precipitation parameters. Pyrethrin I, PII and cinerin II (CII) were significantly correlated mainly with temperature parameters; JII and cinerin I (CI) with precipitation parameters, while JI was correlated with temperature range and precipitation parameters. The mountainous Biokovo population was an outlier in the PCA analysis, mainly due to the extremely low values of temperature parameters at a higher altitude. Precipitation variables separated the central Adriatic islands inhabiting drier habitats from populations inhabiting habitats with greater precipitation. The high-resolution data from this study provide the opportunity for individual-based selection for breeding plants with the best pyrethrin profiles, as well as for increased adaptation to broader climatic conditions.
Read full abstract