It has already been noted that when the tree breeding work began in the 1950s, a wide range of species was considered. Broadleaved species were included and early plus tree selection resulted in the identification of 130 beech, 172 oak, 33 ash and 86 birch individuals. Whilst none of these species received the same level of attention as the conifers, some progeny tests and untested clonal seed orchards were established.
Beech plus trees were selected predominantly in large estates in the beech growing areas of southern England (the South Downs and Cotswolds). Two progeny tests involving 38 families were established in 1953 and it was noted that significant differences between families were not evident until the 15th year after planting. Consistently superior families were, however, observed thereafter and the same observations were made in a later trial of 10 families planted in 1964 on two sites, and in a trial of up to 49 families planted at a wider range of three sites in 1974. Nevertheless, the progeny testing programme was not sufficiently systematic to enable the construction of breeding populations. It was recognised that gains of around 15% would bring little benefit in a species with an average yield class of six and that breeding would be hampered by a strong negative correlation between vigour and stem form. Five untested clonal seed orchards were established between 1956 and 1960, each based on 10 clones; records indicate that only very limited seed collections were ever made from these. A comprehensive clone bank comprising nearly all selections was established at Alice Holt in 1964.
A recent opportunity arose to revive work on beech at the provenance level through participation in an international series of trials. Two new field experiments have recently been established in southern Britain in which sources throughout the European range of the species can be compared. This should provide an opportunity to investigate the growth potential of European material from more southerly latitudes.
Work on oak was more limited, involving the planting of a small progeny trial in 1954 in the Forest of Dean. An untested seed orchard of ash was established and extended over the period 1956-66 but only one seed collection is recorded. Work on hardwood species was curtailed in the major rationalisation in the mid-1960s which concentrated future efforts on a limited number of conifers. It was not until the trend towards increased hardwood planting became established in the mid 1980s that improvement work was revived. This led to the establishment of work in the former Physiology Branch at Alice Holt in which the selection of oak clones on the basis of physiological characteristics was carried out. Early in this project it was realised that a reliable vegetative propagation system for oak would be a prerequisite of clonal selection.
Interest in oak had developed on a European basis by this time and the opportunity was taken to participate in a series of provenance trials with other European partners. These trials, supported by IUFRO were established in 1990 and 1992. As well as planting the material on up to four sites in each year, seed from registered British seed stands was included in the provenances under test. As these experiments approach 10 years from planting, they will provide information on the most productive sources and inform the current assertions that native British sources are most desirable. Experiments were also established at this time to progeny-test random selections of individuals from registered seed stands of both pedunculate and sessile oak; these will give good estimates of the heritability of important traits for the first time in a hardwood species.
Broadleaves on former agricultural land
In the early 1990s interest developed in the use of broadleaves on former agricultural land in lowland areas. Ash and sycamore were considered to be the species with the greatest potential in these sites and a series of provenance trials was established on appropriate sites in close collaboration with agricultural colleges (Cahalan et al., 1995; Cundall et al., 1998). Early indications are that superior sources from both Britain and continental Europe exist and that, although sycamore is a non-native species, there is a positive correlation in Britain between the time of flushing of a source and its latitude.
The early selections of birch plus-trees were made across a number of species and crosses were explored between these trees and others selected in Scandinavian breeding programmes during the 1950-60s at Alice Holt. However, real interest in the improvement of birch then subsided until the 1990s during which it was revived, reflecting renewed recognition of the potential of the species in Scotland and northern England. Concentrating mainly on Betula pendula, a major study of variation in native populations of silver birch has been established in a series of five provenance tests with a sixth in Ireland. These trials will be the first comprehensive set of experiments in which adaptive variation can be studied in a representative range of populations of a native species. The work results from cooperation with Edinburgh University and should provide clear information on the extent of adaptive variation between populations. This could lead to regional zonation and seed transfer rules for birch and could provide the basis for setting up untested clonal seed orchards based within individual zones.
The recent increase in interest in broadleaved species has led to only limited programmes of plus-tree selection by Tree Improvement Branch. However, coordinated work of this type is supported by the cooperative efforts of the British Hardwood Improvement Programme, an association of landowners, universities and scientific institutions, which is concentrating effort on ash, oak, cherry, walnut, birch and sweet chestnut. Improved material is likely to emerge from seedling seed orchards or as vegetatively propagated clones or mixtures from this programme, and Tree Improvement Branch will remain a major participant as this cooperative work develops. Work on the improvement of broadleaved species is important, but (with the exception of birch) the gains and speed of achievement is unlikely to be as high as in conifers. Cooperative work among a range of interests may emerge as an effective way of carrying breeding forward.