Growth Dynamics and Productivity of Riparian Dominant Tree Species in the Pacific Northwest Coastal Rainforest
Estelle Balian, Master's Project

Faced with the conflicting demands of residential and urban development, timber harvest, and adequate habitat for many species, the integrity and management of riparian zones has become an issue of increasing importance.

Despite the contemporary importance of riparian forests, few published studies exist with information regarding the productivity and growth processes of riparian forests. Most of the studies on riparian forests address riparian successional dynamics (Agee, J.K, 1988, Fetherston) or productivity of specific species, especially hardwoods and shrubs (Campbell and Franklin, 1979), or concentrate on a particular species like Douglas fir Means (1996). The few available observations of density and basal area of riparian tree species in the coastal rain forest suggest that younger stands are dominated by red alder (Alnus rubra) and older stands are dominated by sitka spruce (Picea sitchensis) or western hemlock (Tsuga heterophylla). Cottonwood and bigleaf maple which are more susceptible to selective browsing appear as intermediate species in older stands. In other regions, riparian aboveground biomass ranges from 100 to 300 t/ha (Brinson, 1990) and belowground biomass ranges from 12 to 190 t/ha. Aboveground riparian production varies between 6.5 to 21.4 t/ha/year.

One of the most critical roles of riparian forests is to supply the stream with large woody debris (LWD) that influences channel morphology, sedimentation and habitat complexity. The impact ofLWD on a stream depends on size and residence time which is strongly related to the source species (Hyatt, 1999). The time riparian tree species need to reach the mature stage that can provide the river with wood is a determinant factor of LWD frequency and density in the channel. We propose to address this question by studying the growth of riparian dominant trees in an environment with relatively low human impact.

The objective of this study is to understand spatial and temporal dynamics of growth (increment in diameter and height) and production processes of riparian forests (volume and biomass). We will identify growth patterns for dominant tree species and production processes for riparian forests related to geomorphic and physical characteristics. We will provide a spatial description of production process changes taking place over time using a semi-quantitative model. This model will allow us to estimate the time riparian species need to start supplying the system with woody debris. Production processes and rates of riparian forests will also be compared to those in upland areas.

Our hypotheses are:

Study site

The research area is located in the low valley of the Queets River, on the western coast of Washington's Olympic Peninsula (Figure 1). The study site and most of the upstream watershed is part of the Olympic National Park, and is one of the least impacted lowland floodplains in the western continental United States. The Queets watershed covers 1164 km2, its headwaters flow west from glaciers on Mount Olympus through high and low elevation forests before reaching the Pacific Ocean. The area is characterized by relatively dry summers, and high precipitation (~3m/yr) during the fall and winter. The study site is approximately 25 km from the Pacific Ocean. The floodplain is approximately 1km wide at this point and has experienced repeated disturbance events over the past few years.

Alluvial soils characterize west-flowing rivers such as the Queets. Coarse sediments can be mobilized during the fall and winter storm events resulting in a riverbed that is highly mobile.. The river's active channel shows a complicated network of side and back channels separating forest patches of different successional and environmental characteristics. The river's soil is mainly composed of coarse sediment with high hydrologic conductivity resulting in an extensive hyporheic zone in the floodplain areas. (Bechtold, 1999).

The study area is located in the Picea sitchensis (Sitka spruce) zone. The P. sitchensis zone has the mildest climate of any northwestern vegetation zone. The climate is uniformly wet and mild. Constituent tree species are Picea sitchensis, Tsuga heterophylla, Thuja plicata, Pseudotsuga menziesii, Abies grandis and Abies amabilis (Franklin and Dyrness, 1973). Cottonwood and big leaf maple can be intermediary species. On the younger stands more susceptible to flooding disturbance, Alnus rubra and Salix sp. are are also abundant.

The study site forms an island resulting from large woody debris. It is situated between the mainstem Queets River and Pebble Creek, a small tributary flowing through the left bank forest. During high flow periods, the backchannel formed by the creek also carries a portion of the Queets mainstem flow when discharge penetrates the upstream logjam.

Methodology

We identified four physical templates using aerial photographs:

During last summer's field season, four permanent plots were established on each of these templates for a total of 16 plots. These plots were selected in order to cover the different vegetation associations and environmental characteristics related to each template. These plots are superimposed on our already established long-term study sites for nutrients and hyporheic flow. (Figure 2- Table 1)

We had to adapt the size of the plots to the physical characteristics of the stand (high density, landform constraints) in order to keep homogenous plots representative of only one template.

SAMPLING

We will measure the annual change in diameter and density of all trees in each plot. As tree diameter is correlated to tree height, we will be able to estimate the average height of trees in a plot using a height-diameter curve for each species. We will take 20 height measurements in all diameter classes for each dominant species in order to test height estimation validity.

For the plots with a low density of trees and where stems are distinctive (old growth, sitka spruce and alders plots), we will record the number of living trees and the diameter at breast height. In the plots dominated by young willows and mixed willows and alders, the high density of stems makes such an exhaustive count impractical. Each plot will be subdivided into sub-units (25 sub-units for a plot of 625m2). Density and basal area are calculated for a random sample of 10 sub-units. Young willows are difficult to separate in individuals as they sprout from the same root system. Stems connected to the same main root observed at the soil surface will be identified as one tree. For this same tree the basal area will be the sum of the stem cross sections, where stems have a diameter larger than 1cm.

At least Twice annually (April and October) we will measure the diameter of approximately 100 individually-marked specimens of each species (across the types of physical templates) for growth using either dendrometers or sequential changes in diameter at breast height. The growth variables are the basal area increment based on diameter change and the height increase. Tree height will be estimated using the height-diameter curved established for each species.

Tree cores will complete data on past growth. 15 trees per plot will be sampled for past growth by increment boring at breast height or by cutting same diameter stems outside the plot.

Production refers to volume or biomass increment at the stand level. Using allometric equations relating DBH to above-ground biomass (Means et al. 1994) and volume (Hush, 1993), we will estimate species-specific and site-specific, above-ground production rates.

DATA ANALYSIS

We will compare growth rate and productivity for the past 5 years between species and for one species between plots to determine which species grow faster and which template is the most productive.

Growth rates and productivity will be compared to the sub-surface water flow gradient that characterizes each plot (Clinton, in press).

We will also compare growth rate based on basal area increment and productivity in riparian area and in upland forests for species present in both environments.

The relationship between growth rate and approximate age for dominant species will allow us to assess production changes over time.

PRELIMINARY RESULTS

The first field sampling was conducted between April 1999 and November 1999. We established the 16 plots but only 9 plots were surveyed for growth for the 1999 growth season because the 7 other plots were established in late August.

The results for density (number of trees per hectare) and plot basal area (square meters per hectare) are shown in figure 3. In the graph, the plots are organized by template and sorted by successional stage from the stand initiation stage with young willows to the old-growth mature stage.

The plots consisting of mixed young alders and willows exhibit the highest density but low basal area, whereas old growth and sitka spruce plots have the lowest density but the highest basal area. Young willow plots and older alder plots seem to have intermediate density but alders plots show a higher basal area than the mixed willow and alders plots. These results are consistent with successional patterns (Oliver, 1996): during the stand initiation stage, invaders plants colonize open area until one of the physical factors becomes limiting, which characterizes the stem exclusion stage. Competitive interactions in the stem exclusion stage result in mortality increase and density decrease and the stand evolves to the mature forest stage or old growth if no disturbance occurs. Old-growth stands are characterized by an important variability in diameters and heights, a lot of wood debris and standing snags and a clumping distribution of dead trees. The highest density is then reached at the stem exclusion stage, before competition increases mortality.

Basal area is larger in old growth plots where dominant species more adapted to the environment have developed over a longer time and have reached large diameters.

Next year we will again sample density and basal area for each plot but we don't expect wide variation in the plots characteristics.

GROWTH RESULTS

A total of 564 trees have been sampled for growth during 1999 growth season between April and November. (Table 2)

The dominant riparian species we surveyed were:
Big leaf maple: Acer macrophyllum (ACMA)
Red alder: Alnus Rubra (ALRU)
Sitka spruce: Picea sitchensis (PISI)
Cottonwood: Populus trichocarpa (POTR)
Western hemlock: Tsuga heterophylla (TSHE)

Dendrometers were set up on 50 trees located in the old growth and in the young sitka spruce stands. We read increment in diameter for these 50 trees in July, September and November. Four of the species tend to show higher diameter increase between April and July. Cottonwood diameter increased more from September to November following the first major precipitation event. For all species except cottonwood the total increment in diameter is approximately 0.3 cm. Cottonwood tends to have a lower growth in diameter. (Figure 4)

Sequential measurements are not available yet.