What influences the long-term development of mixtures in British forests?

Abstract

Six experiments were established between 1955 and 1962 in different parts of northern and western Britain which used replicated randomized block designs to compare the performance of two species 50:50 mixtures with pure stands of the component species. The species involved were variously lodgepole pine (Pinus contorta Dougl.), Japanese larch (Larix kaempferi Lamb. Carr.), Scots pine (Pinus sylvestris L.), silver birch (Betula pendula Roth.), Sitka spruce (Picea sitchensis Bong. Carr.) and western hemlock (Tsuga heterophylla Raf. Sarg.). The first four species are light demanding, while Sitka spruce is of intermediate shade tolerance and western hemlock is very shade tolerant: only Scots pine and silver birch are native to Great Britain. In three experiments (Bickley, Ceannacroc, Hambleton), the mixtures were of two light-demanding species, while at the other three sites, the mixture tested contained species of different shade tolerance. The experiments were followed for around 50 years, similar to a full rotation of even-aged conifer stands in Britain. Five experiments showed a tendency for one species to dominate in mixture, possibly reflecting differences in the shade tolerance or other functional traits of the component species. In the three experiments, the basal area of the mixtures at the last assessment was significantly higher than predicted based on the performance of the pure stands (i.e. the mixture ‘overyielded’). In two of these cases, the mixture had had a higher basal area than found in the more productive pure stand indicating ‘transgressive overyielding’. Significant basal area differences were generally more evident at the later assessment date. The exception was in a Scots pine: western hemlock mixture where greater overyielding at the earlier date indicated a nursing (‘facilitation’) effect. In the remaining experiments, the performance of the mixture conformed to predictions from the growth of the component species in pure stands. Taken overall, the results suggest that functional traits can be used to interpret the performance of mixtures but prediction of the outcome will require better understanding of the interplay between species and site characteristics plus the influence of silvicultural interventions.