Latex yield and growth are the key complex traits in commercial rubber production. The present study is the first to report genome-wide association mapping of latex yield and girth, for 170 Amazonian accessions grown in a suboptimal area characterized by limited rainfall and a lengthy dry season. Targeted sequence enrichment to capture gene transcripts generated 14,155 high quality filtered single nucleotide polymorphisms (SNPs) of which 94.3% resided in coding regions. The rapid decay of linkage disequilibrium over physical and genetic distance found in the accessions was comparable to those previously reported for several outcrossing species. A mixed linear model detected three significant SNPs in three candidate genes involved in plant adaptation to drought stress, individually explaining 12.7-15.7% of the phenotypic variance. The SNPs identified in the study will help to extend understanding, and to support genetic improvement of rubber trees grown in drought-affected regions.

No abstract

Drought shapes the distribution and survival of trees even in tropical wet forests, and the wood and leaf trait spectra are used to understand drought adaptations. However, trait variation may result from ontogenetic adjustment or be related to tree size and not reflect evolutionary adaptations. Intraspecific variation in adaptations to drought can be an important factor in a species’ distribution and response to climate change, but excluding potentially confounding factors and proving adaptive evolution are challenging. Provenance trials can identify hereditary variability. We analysed wood and leaf traits in rubber (Hevea brasiliensis) tree clones from 15 locations in the Amazon basin that were planted in northern Thailand, controlled for tree size, tested for genetic relatedness and the phylogenetic signal in traits and compared trait variations with the climate at the location of origin. Correlations between traits and tree size were low. Intraspecific trait variation was similar to relationships in published among‐species comparisons, and correlations among wood traits and correlations among leaf traits were stronger than between wood and leaf traits. Genotype explained 30%–70% of the trait variation, and traits differed in how much of this variation was controlled by location or the relatedness among clones. There was no correlation with mean temperature or total annual rainfall. However, rainfall in the driest quarter (19–199 mm) was strongly related to leaf mass per area, carbon isotopic composition and area‐based nitrogen content (r2 = 0.54–0.70) and weaker to wood traits (vessel density and vessel lumen fraction). Trees from locations with a stronger dry season also had higher growth rates in Thailand. All traits correlating with climate showed a significant phylogenetic signal. We found no evidence of increased drought tolerance, but the trait spectrum and higher growth in trees from drier locations suggest that deciduous rubber trees have adapted via drought avoidance rather than tolerance. Our study also underlines the importance of looking at a suite of traits rather than individual ones to understand adaptive strategies. A http://onlinelibrary.wiley.com/doi/10.1111/1365-2435.13203/suppinfo is available for this article.

Jatropha curcas L. (physic nut) is native to Central America and now naturalized widely in many tropical and subtropical areas. Microsatellite markers were isolated and characterized. Eleven out of 55 markers showed polymorphisms, and the allelic variation was investigated using 26 accessions of J. curcas collected from several provinces in Thailand. Each marker showed 2 to 5 alleles and the average polymorphic information content (PIC) was 0.49. Thirty four markers (62 %) were also successfully amplified in J. integerrima, J. gossypifolia and J. podagrica.Additional key words: genetic diversity, microsatellite-enriched genomic library, physic nut.

Pará rubber (Hevea brasiliensis) is grown primarily for latex. Some clones (timber clones) are frequently planted for wood production. Rubberwood, from old rubber trees, is an important economic byproduct. In general, the quality of the wood depends on multiple factors, including chemical and mechanical properties. This study investigates the effects of clone variety, height (1.3 and 6.0 m above ground level), and the variation between clone and height on chemical and mechanical properties of seven Pará rubber clones (RRIT 251, RRIM 600, RRI-CH-35-59, RRI-CH-35-650, RRI-CH-35-1397, RRI-CH-35-1757, and RRI-CH-35-2086). Thirteen-year old trees were sampled selecting one tree per clone. Results showed that height had a significant effect on the chemical components of the wood. Furthermore, the interaction between clone and height also had an effect on the chemical composition of cellulose, hemicelluloses, lignin, and extractives. In contrast, the mechanical properties of rubberwood depended a lot on the clone variety. The chemical and mechanical properties of rubberwood were significantly correlated among the seven clones studied. RRI-CH-35-1757, RRI-CH-35-2086, and RRIM 600 had the highest values for holocellulose as the main chemical components and mechanical strength.