Variation in cone and seed characters in clonal seed orchards of <Emphasis Type="Italic">Pinus sylvestris</Emphasis>

Cone and seed characters were observed on top, middle and bottom portions of tree crown in 3 ramets/clone in three Turkish Pinus sylvestris seed orchards. Broad sense heritability (clonal repeatability), and correlations among characters were estimated. Around one quarter of the seed production occurred in the top portion, half in the middle and one quarter in the bottom portion of the crown for all orchards. The percentage of filled seeds varied little with the crown position, indicating more or less similar levels of selfing in the bottom of the crown as in the top. The seed weight was typically 11 mg. Differences were found for studied cone and seed characters among orchards and crown positions. Variation among grafts within clone was higher than among clones for most characters. The heritability was on average below 0.5 (e.g., cone diameter, number of filled seed per cone) and rarely rose above that (e.g., cone form, length/diameter; percentage of filled seed) for any individual characters. The coefficient of variation within clones was often higher than among clones. Thus, non-genetic factors seem often more important for the variation in performance of grafts than their genetic constitution. Cone form (length/diameter) was the character where the clone influence was the strongest. Cone number and cone dry weight showed significant correlations with seed characters (numbers of total and filled seed, percentage and weight of filled seeds). Significant correlation was found between seed characters.     

Growth characters and number of strobili in clonal seed orchards of Pinus sylvestris

Observations were made on six grafts for each of 25 clones in three Scots pine (Pinus sylvestris) seed orchards in Turkey. The characters studied were number of female and male strobili, height below and above the longest branch, total height; diameter at base and breast height, crown diameter, and number of branches. Variation, broad-sense heritability (H ²) and correlations between characters were estimated. Variation among clones was lower than among grafts within clone for all characters. The genetic variation for number of strobili varied between 0 and 17% of total variation, while that for growth characters values varied between 2 and 13%. The number of female strobili appeared more variable among trees than the number of male strobili. H ² was not consistently high for any character or seed orchard. The number of strobili increased with the size of the tree, but not dramatically. Correlations between measures of tree size (both on clone level and individual graft level) and the number of strobili were in the magnitude r ≈ 0.3. Diameter at breast height seems a reasonable predictor for number of strobili.     

Optimum lifetime for Swedish <Emphasis Type="Italic">Picea abies</Emphasis> seed orchards

The optimum lifetime of Picea abies seed orchards in Sweden was calculated using a model which considers changes in seed yield over time, annual progress of genetic gain, establishment, management and seed costs, and both the monetary and genetic value of the seeds produced. The longer the seed orchard is active, the more genetically outdated the produced seeds will become, and thus their value will decline with time. A main scenario was constructed using the best available estimates for Swedish Norway spruce and was compared to various alternatives reflecting likely variations. The optimal lifetimes in all of these scenarios was in the order of 40 years. Sensitivity analyses showed that use of orchards for slightly more or less than their optimal lifetimes does not lead to large losses. A more effective tree breeding program could slightly reduce the optimal lifetimes. The optimal lifetime depends strongly on the relationship between monetary value and genetic gain. Thus, in cases where genetic progress is sufficiently rapid, the optimal lifetime may be less than 30 years.     

Genetic thinning of clonal seed orchards using linear deployment may improve both gain and diversity

The linear deployment algorithm defines a straight line relationship between the number of ramets and the breeding value of a given clone. However, when used to determine thinning in clonal seed orchards, there is an upper limit, determined by the number of ramets in the seed orchard prior to thinning. Linear deployment thinning maximises genetic gain and effective clone number whilst minimising thinning intensity, thus both production and genetic diversity are optimised. Here two applications of linear deployment to Swedish Norway spruce seed orchards are described.