Forest productivity and tree diversity relationships depend on ecological context within mid-Atlantic and Appalachian forests (USA)

► We investigate productivity-tree diversity relationships in different ecoregions. ► In mid-Atlantic coastal regions (USA) the most productive stands were least diverse. ► In Appalachian mountain regions the most productive stands were most diverse. ► We explore explanations of patterns including ecological, climatic, and land use. Factors influencing the relationship between ecosystem productivity and biological diversity form the basis of much ecological theory. An understanding of how productivity–diversity relationships are influenced by scale of observation and unique attributes of ecoregions may provide important insights to aid conservation planning for carbon retention and biodiversity. Here we use publically available datasets to investigate patterns of productivity–diversity and explore potential factors influencing these patterns in forests located in five mid-Atlantic and Appalachian states. We used a geographic information system (GIS) to overlay multiple publically available datasets including remotely sensed estimates of productivity from MODIS and tree diversity estimated from the forest inventory analysis (FIA) database. We evaluated productivity–diversity relationships using two scales of observation (among and within ecoregions). We also determined if productivity–diversity relationships might be related to region-wide patterns in land use, and if the relationships varied by forest type, land management zones, and along gradients of mean productivity and diversity. Productivity–diversity relationships depended on scale and varied among ecoregions, and land use was correlated with both productivity and diversity. Mountainous ecoregions were characterized by positive productivity–diversity relationships, whereas coastal ecoregions were characterized by negative productivity–diversity relationships. Forest types and management zones that were on average less productive and more diverse exhibited positive productivity–diversity relationships across stands (the most productive stands were more diverse). In contrast, ecoregions and forest types that were on average more productive and less diverse exhibited negative productivity–diversity relationships (the most productive stands were less diverse). In conclusion, regional and local ecological and anthropogenic factors likely influence productivity–diversity relationships and these relationships appear to change along gradients of productivity and diversity.