Introduction

Eucalyptus bosistoana F. Muell. is a eucalypt species native to the Australian east coast and eastern high country. As old-growth timber it has been used in general construction, heavy engineering structures, sleepers, bridges, wharves, flooring, decking, cladding, cross-arms, posts, poles and piles \cite{Bootle2005,Poynton1979} it a good candidate for high value timber products in New Zealand as it shows high stiffness, class 1 durability \cite{Australia2005}, exhibits good growth rates, along with reasonable drought and frost resistance \cite{Poynton1979,Altaner} used for short rotation timber crops often exhibit excessive growth-strain production causing problems in the production of saw logs \cite{Blackburn2014,Valencia2010,Yang2002}
Recently E. bosistoana has been investigated for its potential as a high value commercial timber species within the dry-land areas of New Zealand \cite{Altaner,Apiolaza2011,Davies2017} establish an effective initial breeding population, a 4032 tree breeding trial was established to conduct very early screening (at age 21 months) of growth and wood properties; growth-strain, under-bark stem diameter, wood basic density, volumetric shrinkage, tree height and acoustic velocity. Growth-strain is rarely (if ever) incorporated in breeding trials as, until recently, with the development of a fast and cheap method \cite{Chauhan2010,Davies2017} it has been too expensive and time consuming to assess. For background information and a review of growth-strain see \citet{Alméras2016}. During the period between 2009 and 2012, seeds were collected from all known natural populations. ‘Plus’ trees for growth, form and health were selected within these populations. As a result, the genetic material contained in the trial represents all known populations of the species.
Starting from a wild population is not a common occurrence within modern tree breeding, and presents a unique challenge; how can the genetic variation be quickly and efficiently distilled down to only the trees which are likely to be useful in a full-length breeding cycle? Using early selection methods provides fast and effective ways to screen out individuals which will prove detrimental to the breeding program, without the expenditure of growing all individuals to a mature state. As a consequence, less gain is expected; however, the time and money invested in the trial is substantially lower. Very early section methods have been studied and used in breeding trials previously \cite{Apiolaza2011,Apiolaza2013,Aggarwal2013}.