Previous Page Table of Contents Next Page

Results to end of Nursery Stage


J. O'Driscoll
Research Branch, Forest and Wildlife Service, Ireland 1


The present-day range of Sitka spruce (P. sitchensis) extends from latitude 61°N in Alaska to latitude 39°N in California, a distance of 2 880 km. This range corresponds fairly closely to its pre-Wisconsin glaciation distribution. During the ice age the distribution north of latitude 48°N was covered with ice up to 900 m thick. Remnant populations were thought to have survived mainly on nunataks (3). The effect of the ice sheet on the southern arm of the distribution was to push it as far south as San Francisco. Remnant populations of this southerly advance still exist in Mendocino county, California, which is 50 km further south than the terminus of the present-day main distribution (1).

Throughout its present distribution it is confined mainly to a narrow coastal belt commonly known as the fog belt (Figure 1). Its limited extension inland is usually along river basins. In this situation it has been found up to 50 km from the sea in Oregon and Washington and up to 200 km in Alaska and British Columbia. At these extremes of its inland penetration it is found in association with white spruce (P. glauca). Its elevational range is also limited throughout its distribution, with the exception of Alaska where it is found up to 800 m. More usually it is found from sea level up to 300 m.

The general climate through the species range is a maritime one characterized by equable temperatures, high precipitation, cloudiness, long frost-free periods and absence of extreme cold. Soil type varies depending on the location but best growth is usually found on alluvial soils with a high nutrient status. Growth is poorest where there is a high humus content. Pure stands occur only where disturbance has taken place. Under more stable conditions it is more commonly associated with western hemlock (T. heterophylla) (2).

Sitka spruce was first discovered in 1792 at Admiralty Inlet, Puget Sound, Washington. It was first introduced to Europe around 1830 by David Douglas (6). In the intervening one hundred and fifty years it has been planted extensively outside its natural range. In Britain and Ireland it is a species of major importance, while on continental Europe it is of moderate importance in Denmark, Norway, Germany, the Netherlands, Belgium and France. In New Zealand and Australia it is only of minor importance. In its native habitat it reaches its maximum commercial importance in Alaska and northern British Columbia.


Prior to the establishment of the Sitka spruce Working Group, study of the species variation was limited and confined in the main to Europe. The function of the Working Group, established in 1971 at the 15th IUFRO Congress at Gainesville, Florida, is to coordinate provenance research activities with the IUFRO provenance collection of Sitka spruce seed. This collection was carried out in two phases, the Washington - Oregon distribution in 1968 by Mr. J. Turnbull and Mr. B. Hansen and the British Columbia - Alaska distribution in 1970 by Dr. A. Fletcher and Mr. N. Danby.

Figure 1. Picea sitchensis. Species range and provenance location

Figure 1

The objective of the collection was to sample the species natural range, both extensively and intensively. This objective was more than successfully achieved with 81 provenances being collected, their distributions being 19 in Alaska, 41 in British Columbia, 11 in Washington, 8 in Oregon and 2 in California. In addition the introgression sone with P. glauca was sampled along the upper reaches of the Skeena and Naas rivers, British Columbia. The depth of this collection allowed interested researchers with either limited or extensive knowledge of the species to choose the number of provenances which suited best their requirements.


Experience from earlier and older experiments had shown that this species exhibited a fairly well defined pattern of continuous variation in certain characteristics. It was therefore decided to sample the species range at intervals of from 50 – 80 km. Extra samples were taken where the species was found growing in very dry areas or where only remnant populations remained. Collection areas were categorised under three headings.

  1. Commercial areas: Suitable for commercial collections yielding up to 50 kg of seed from good quality trees at low cost.

  2. Collectable areas: Marginally suitable for commercial collection, yielding 5 to 30 kg of seed from good quality trees at a cost twice that of (a).

  3. Scientific areas: Where stand size was too small to warrant commercial collections - collections mainly for special research projects.

Within each collection area, every effort was made to ensure that the sample was representative of the species growing in that area, by collecting from up to 20 trees representing as typical a cross section of the population as possible.

Chosen trees within each sub-population were widely scattered to minimize the likelihood of half-sib matings and at the same time to maintain a broad genetic base. Collections were made mainly by climbing, though in some instances squirrel caches were used. Approximately 400 cones were collected per tree and kept separate at this stage. It was only after seed extraction that bulking of seed per provenance took place. Single tree samples of 15 cones were kept separate for further study by the University of British Columbia.

To ensure that no mistakes occurred at any stage of the collection, lots were labelled from collection to extraction. The extraction was carried out under the supervision of the collection team. Data included detailed listing of location and site details, covering such aspects as soil, geology, slope, aspect and forest type. Individual tree descriptions included height, diameter, crown dimensions and form.


When the working group on procurement of seed for provenance research informed member institutes of IUFRO of the availability of the collection, 18 institutes in 14 countries expressed an interest in obtaining provenance lots (Appendix 1). The number of provenances requested ranged from a low of 3 to a high of 81. Within provenances sample size varied from 500 to 3000 seed lots. It was at this stage that the Sitka spruce Working Group was activated. A meeting was held in Ireland in September 1972 to plan the activities of the Working Group. Representatives of eleven of the eighteen member institutes of the working group attended. Due to the varying numbers of provenances taken by each institute and the varying requirements of each institute's plan, it was agreed that it would not be possible to organize an international experiment containing all the provenances.

It was therefore decided to confine the working group's activities to ten provenances (Table 1).

Table 1: Provenances included in international provenance experiment of Sitka spruce

IUFRO NumberLocationLatitudeLongitudeAltitude
3003Forks, Washington48.07124.30 137
3008Hoquiam, Washington47.08124.77     6
3012Necanicum, Oregon45.82123.76   46
3024Duck Creek, Alaska58.37134.58   30
3030Ward Lake, Alaska55.42131.70   15
  3040/IUsk Ferry, Skeena R., B.C.54.63128.40 137
3044Inverness, B.C.54.20130.25   15
3049Link Rd., Graham Is., Q.C.I.53.50132.17   91
3056Holberg, Vancouver Is., B.C.50.62128.12   30
3062Big Qualicum R., Vancouver Is., B.C.49.38124.62     0

/I Replaced by 3038 Pacific, Skeena R., B.C., in Germany and Netherlands.

These were selected to represent the major physiographical regions of the species' natural distribution. Each institute received a 500 seed sample of each lot and agreed to follow the prescriptions of the detailed working plan drawn up by the working group chairman (4).

The working plan laid down detailed prescriptions on all aspects of the experiment from the pre-sowing stage through to the forest stage. The plan aimed at ensuring that all cultural and management treatments were carried out uniformly at all locations. The objective of the international experiment was to study provenance by site interaction between countries, plus genotype stability under different ecological conditions. Evaluation was and would be based on three stages:

  1. seed bed
  2. lining out, and
  3. forest stages.

The experiment would have a life of twelve years from seed. Details of assessments proposed are listed in Appendix II. In addition to the International Ten Provenance Experiment, each Institute drew up its own plan for the other provenances they had obtained.


The experiment was successfully sown in all participating institutes in 1973 and the nursery stage was completed by either autumn 1974 or 1975. This allowed the forest stage to be outplanted the following year. In addition to the traditional style nursery experiment, a phytotron study was also carried out at Petawawa, Canada. The results at the end of the nursery stage were presented to the 16th IUFRO Congress, Oslo, June 1976, as a report of the Working Group's activities since it was initiated in 1971 (5).

The report contains the results of each individual institute, presented as a separate paper. In addition it was also possible to have a joint analysis of the data carried out. The results of the phytotron study were included by kind permission of the National Research Council of Canada, to complete fully the report of the Working Group's activities.


Though the working plan stipulated a specific regime of assessments, it was not possible for all collaborators to carry them out exactly for varying reasons. These ranged from killing winter frosts, to lack of suitable facilities and staff, to an unfortunate accident with a herbicide.

Seed Parameters - weight - length - width

The parameter most commonly assessed in this section was that of seed weight. The general consensus of the results indicated that there was a significant difference between provenances in seed weight, with generally the Alaskan provenances being heavier. There was however no significant correlation with latitude of origin, subsequent height growth or any of the other seed source variables suggested in the working plan.

Where length was assessed so also was width, both being measured by a high percentage of the participants. Seed length variation between provenances was generally greater than that for width. They were generally correlated significantly with each other but not with latitude. When variation within seed parameters was examined it was found to be greatest among the Washington, Oregon and lower Skeena valley provenances. The former may be due to the different ecological conditions under which the southern end of the species evolved, and the latter to the fact that the Skeena river valley is considered to be an introgression zone between Sitka spruce and white spruce.

Germination Parameters

Assessments under this heading were germination rate, capacity and cotyledon numbers. Most collaborators assessed at least one of these parameters. Results tended to follow a similar pattern at each location.

Germination rate between provenances was significantly different but no clinal pattern of variation was apparent. The Big Qualicum River, Vancouver Island, B.C. and Usk Ferry or Pacific, Upper Skeena River valley provenances, were the fastest with those from Holberg, Vancouver Island and Inverness, Lower Skeena River valley the slowest. Rate was not correlated with any of the seed parameters, latitude of origin or one year height growth.

Though there was a significant difference between provenances in germination capacity, there was no correlation between this parameter and latitude, or any of the seed parameters. Capacity was found to vary inversely with mean annual precipitation and mean April through August precipitation. General trend was for northern and outer coast provenances to have poorer germination.

Cotyledon number ranged from 3 to 8 with a general mean in the region of 5. It was not correlated with latitude. Degree of relationship with seed length and seed weight varied between institutes. It was however found to be positively correlated with July mean maximum temperature. The heritability value of this parameter was also extremely high.

At germination some chlorosis and seed cap retention was observed. It was confined to only one or two seedlings per provenance and was in no way correlated with any of the variables used in the experiment.

Growth Parameters - height, diameter, growing points

Height growth at any of three stages was the most frequently assessed parameter of the three listed under this heading. Most institutes carried it out at end of 1 and 2 years.

Where climatic or accidental damage occurred, assessments were confined to end of 1 year growth. At all assessment dates there was a highly significant difference between the provenances. The time of assessment of this parameter had a considerable effect on the degree of the relationship which height growth had with many of the other variables. At two months height growth was significantly correlated with germination rate but not with latitude. A negative correlation with latitude had developed by the end of 1 year's growth. By the end of two years latitude was having a major effect as indicated by the high negative correlation coefficients obtained by most collaborators. Over this period the effect of seed parameters had decreased rapidly. Generally it was found that height growth was a function of length of growing season which was in turn a function of latitude of origin. Total height at two years was also found to be negatively correlated with annual rainfall and rainfall in frost free period. Ranking of the provenances indicated that Big Qualicum River, Vancouver Island and the Washington and Oregon provenances were best, particularly where no frost damage occurred. At all locations poorest growth was found in the Alaskan provenances (Figure 2).

Root collar diameter results indicated that this parameter was positively correlated with height growth. Its relationship with latitude was non-significant. When correlated with climatic data, it was found to have a negative correlation with mean annual rainfall and rainfall in the frost free period. Heritability levels for this parameter were the lowest of all assessed parameters.

The number of growing points assessment was carried out by 3 institutes mainly at the end of 2 years growth. There was a highly significant difference between provenances. This parameter was positively correlated with height and diameter growth and negatively correlated with latitude, total rainfall and rainfall in frost free period.


This characteristic was assessed under two headings:

  1. flushing, and
  2. bud set.

Flushing time was found to be significantly different between provenances. Ranking of provenances varied between experiment locations but overall no clinal pattern of variation was evident. It was significantly negatively correlated with height at 2 years. Generally this parameter was found to be under strong genetic control.

Differences between provenances in date of bud set were highly significant. Variation pattern was of a clinal type with a strong negative correlation with latitude. The length of the growing season was positively correlated with height growth but only moderately correlated with diameter growth. The more southerly provenances tended to have a longer growing season and consequently would be expected to suffer from early autumn frost damage in northern climes. The conclusion was drawn that date of bud set was under photoperiodic control.

Other Assessments

The effect of frost and winter cold was recorded by two institutes. At one location, Riga, Latvia the combined effects of these climatic parameters killed the greater percentage of the transplants in all provenances. Effect was somewhat less in the most northerly provenances. At the second location, Stend, West Norway, autumn frost and winter cold damage was significantly related with latitude of origin of provenances. The most southerly provenance suffered extensive damage and was not available for outplanting.

Figure 2. Relationship between height at second year and latitude of provenance for 10 ha provenances

Figure 2

Combined Analysis

The objective of the combined analysis was to investigate the country x genotype interaction using the technique of joint regression analysis without replication. The investigation of the interaction between the assessed parameters and location of experimental nursery was limited to those parameters assessed by each collaborator at the stipulated time.

Various unforeseen mishaps accounted for most of the missing data sets. Variation in height growth was attributable to location and provenance, while their interaction was found to be small. This decreased further in the second year in proportion to the other effects. At year 1 the main contributor to the interaction was Big Qualicum River, Vancouver Island, B.C.

At year 2, the main contributors were Big Qualicum River, Necanicum (Oregon) and both the Alaskan provenances. Height growth was significantly, linearly related to latitude, longitude, average maximum temperature and average minimum temperature. The significant linear correlation between height growth, mean monthly rainfall and temperature was an illustration of the importance of climatic effects at the seed origin. The effect of nursery location on height was described by the following relationships.

  1. with longitude - a positive linear correlation;

  2. with latitude - a quadratic relationship with peak values at 52 – 53°N;

  3. with rainfall - a negative linear correlation.

The trend in root collar diameter growth, as expressed by a quadratic relationship with latitude of nursery origin, was similar to that for height, with peak values around 53°N. Provenance x location interaction was less evident. Alaskan and Skeena River provenances were smallest with the southern provenances being the largest. Number of growing points showed little indication of provenance x location interaction. Alaskan provenances generally had fewest growing points with those from Oregon the greatest number. The growing season was shortest in those provenances from Alaska, while those from Washington and Oregon had the longest on all sites.


In conclusion it can be stated that the results of the nursery stage follow a rather similar pattern at all locations. The full effect of local climate had not manifested itself at this stage. However it would play an ever increasing role in the forest stage. Even at this stage the limitations of southern provenances were very evident for northern Europe. Provenances from Vancouver Island and north Washington would appear to be best for mid Europe, while southern Europe would suit more southerly provenances. The within provenance variation might be studied further to extend the range of more southerly provenances to more northerly locations.


(1) Daubenmire, R. 1968  Some geographic variations in Picea sitchensis and their ecological interpretation. Canadian Journal of Botany 46.

(2) Fowells, H.A. 1965  Silvics of Forest Trees of the United States. USDA Forest Service, Agriculture Handbook 271.

(3) Heusser, Calvin, J. 1952  Nunatak flora of the Juneau Ice Field. Alaska. Torroy Bot. Club Bull. 81 (3).

(4) O'Driscoll, J. 1972  Working Plan for International Ten Provenance Experiment. Forest and Wildlife Service, Dublin, Ireland.

(5) O'Driscoll, J. 1976  IUFRO Sitka spruce International Ten Provenance Experiment - Nursery Stage Results. Report, Forest and Wildlife Service, Dublin, Ireland.

(6) Veitch, J. 1881  A Manual of the Coniferae.

1 Address: Sidmonton Place, Bray, Co. Wicklow, Ireland

Previous Page Top of Page Next Page