Seedling quality and short-term field performance of three Amazonian forest species as affected by site conditions

High quality seedlings are essential for the successful establishment of forest stands. Seedling quality can be assessed by the morphological attributes (e.g., height, diameter, dry mass, Dickson's quality index) measured in the nursery phase. The defining, producing, and handling of seedlings can be based on specific characteristics suited to a site, as proposed by the Target Plant Concept. However, a target plant concept is an important research gap for Amazonian tree species. Here, we explore the associations between morphological attributes of seedlings in the nursery and determine the combined effect of the seedling quality attributes and site preparation methods on one-year field performance to define a range for nondestructive morphological attributes that can be associated with successful establishment for Amazonian commercial forest species - Bertholletia excelsa, Dipteryx odorata and Tachigali vulgaris. We measured morphological attributes in the nursery and analyzed the correlations to detect better predictors of destructive attributes. Then, we related these initial morphological attributes to survival and growth one year after planting in two site conditions: manual holing preparation and mechanical site preparation (subsoiling plus harrowing). We represent growth using different metrics: one-year size, absolute and relative growth rates in height and root collar diameter. Site conditions were assessed by soil physical properties. For all species, root collar diameter was a good predictor of destructive attributes that evaluated seedling quality, such as Dickson's quality index. Site preparation methods resulted in different site qualities. Mechanical site preparation improved the total porosity and reduced the bulk density and resistance to penetration. Survival was not affected by initial attributes or site preparation methods. The initial attributes were poor predictors of field growth for D. odorata and T. vulgaris. The field performance of B. excelsa seedlings was affected by initial attributes, site, and the combination of both. The definition of a range for operational attributes according to site conditions is true only for B. excelsa. These results are important to help fill research gaps related to technical procedures to establish large-scale reforestation projects using Amazonian tree species.