Area-based breeding zones to minimize maladaptation

A new breeding zone delineation scheme identifies for a given number of zones the zone-boundary placement that minimizes regional maladaptation in breeding programs. First, an adaptive map is created by using conventional genetic test data. Then, the large array of predicted adaptive values is subjected to cluster analysis, which assigns each grid cell of the region to one of a predetermined number of clusters (breeding zones) such that the sum of the squared distances between each cellʹs adaptive value and its cluster mean is minimized. This approach minimizes the average adaptive distance between the origin of a breeding programʹs selected trees and planting locations throughout the region of focus. The procedure is illustrated by the use of adaptive values of 69 interior spruce (Picea engelmannii Parry ex Engelm. × Picea glauca (Moench) Voss) open-pollinated families (sources) from southeast British Columbia, Canada. Adaptive values of each 1.5 km × 1.5 km grid cell in the 80 000-km2 region were predicted using a geneco logical model (R2 = 0.64), and the values were subjected to cluster analysis to identify breeding zone boundaries that were then mapped using a geographic information system. Regardless of the number of zones created, a regional maladaptation index was consistently smaller when zones were devised with area-based cluster (ABC) analysis than when zones were created by dividing the region into bands of equal elevational or adaptive-value widths. Application of the ABC procedure should assist in identifying the optimum breeding-zone alignment for a given number of zones.