“…From a biotic perspective, among-ecotype differentiation ( Cortés et al, 2012a , b , 2013 ; Blair et al, 2016 ), intrapopulation divergence ( Cortés et al, 2011 ; Blair et al, 2012 , 2018 ; Kelleher et al, 2012 ), and within-family variation ( Galeano et al, 2012 ; Blair et al, 2013 ) could encourage or coerce adaptation. Population’s functioning, abundance, distribution, and diversity, as predicted from controlled experiments ( Way and Oren, 2010 ; Elmendorf et al, 2012 ; Wolkovich et al, 2012 ; Andresen et al, 2016 ; Becklin et al, 2017 ; Singh et al, 2017 ), experimental evolution ( Tenaillon et al, 2012 ; Mallard et al, 2018 ; Pfenninger and Foucault, 2020 ), biological monitoring ( Walther et al, 2002 ; Franks et al, 2013 ; Wipf et al, 2013 ; Reichstein et al, 2014 ; Hällfors et al, 2020 ), and shifts observed in the fossil record ( Alsos et al, 2009 ; Willis and MacDonald, 2011 ; Lyons et al, 2016 ; Bruelheide et al, 2018 ), can feed back on climate change ( Pearson et al, 2013 ) and so be considered as drivers themselves. Regardless of the exact nature and extent of the data type, both GP and ML may offer suitable scenarios to merge diverse, and even conflicting, data sources in order to pinpoint emergent properties ( Street et al, 2011 ) out of a complex system, as is thermal genomic adaptation.…”