Range-wide Differential Adaptation and Genomic Vulnerability in Critically Endangered Asian Rosewoods

In the billion-dollar global illegal wildlife trade, rosewoods have been the world’s most trafficked wild product since 2005[1][1]. Dalbergia cochinchinensis and D. oliveri are the most sought-after rosewoods in the Greater Mekong Subregion[2][2]. They are exposed to significant genetic risks and the lack of knowledge on their adaptability limits the effectiveness of conservation efforts. Here we present genome assemblies and range-wide genomic scans of adaptive variation, together with predictions of genomic vulnerability to climate change. Adaptive genomic variation was differentially associated with temperature and precipitation-related variables between the species, although their natural ranges overlap. The findings are consistent with differences in pioneering ability and in drought tolerance[3][3]. We predict their genomic offsets will increase over time and with increasing carbon emission pathway but at a faster pace in D. cochinchinensis than in D. oliveri . These results and the distinct gene-environment association in the eastern coastal edge suggest species-specific conservation actions: germplasm representation across the range in D. cochinchinensis and focused on vulnerability hotspots in D. oliveri . We translated our genomic models into a seed source matching application, seedeR , to rapidly inform restoration efforts. Our ecological genomic research uncovering contrasting selection forces acting in sympatric rosewoods is of relevance to conserving tropical trees globally and combating risks from climate change. Significant statement In the billion-dollar global illegal wildlife trade, rosewoods have been the world’s most trafficked wild product since 2005, with Dalbergia cochinchinensis and D. oliveri being the most sought-after and endangered species in Southeast Asia. Emerging efforts for their restoration have lacked a suitable evidence base on adaptability and adaptive potential. We integrated range-wide genomic data and climate models to detect the differential adaptation between D. cochinchinensis and D. oliveri in relevance to temperature- and precipitation-related variables and projected their vulnerability until 2100. We highlighted the stronger local adaptation in the coastal edge of the species ranges suggesting conservation priority. We developed genomic resources including chromosome-level genome assemblies and a web-based application seedeR for genomic model-enabled assisted migration and restoration. ### Competing Interest Statement The authors have declared no competing interest. [1]: #ref-1 [2]: #ref-2 [3]: #ref-3