Forest Landscape Baseline No. 18
Brief Progress and Summary Reports 1999
|
Forest Landscape Baseline No. 18 Brief Progress and Summary Reports 1999 | |
USING MULTI-LAYER VEGETATION DATA AND SLOPE POSITION TO REFINE BIODIVERSITY ASSESSMENT IN CENTRAL ONTARIO: PRELIMINARY RESULTS
T. Lee and P. A. Quinby
Introduction
There is general consensus among natural heritage and conservation experts in Ontario that the amount of protected area in most parts of the province must increase if biodiversity and ecosystem integrity are to be maintained in the long term. This requires accurate and efficient assessment of species and ecosystems and planning for their conservation. To carry out biodiversity conservation planning in central Ontario, the Ontario Ministry of Natural Resources (OMNR) stratifies the landscape first by landform type and then by forest overstory community type as determined by Forest Resource Inventory (FRI) maps. These maps are made from interpretation of 1:20,000 scale aerial photos and show the composition, age, and density of the tree species in the upper forest canopy. This information is then used to identify and map the boundaries of forest stands. We know, however, that forest vegetation also includes plants below the uppermost canopy such as smaller trees, saplings, shrubs, herbs, mosses, liverworts, lichens, snags and logs. In addition, FRI stands may often reach 100 ha (250 acres) or more in size. This is much larger than some ecosystems that occur within them such as riparian communities along streams and pockets of forested wetlands. For example, tree species in riparian zones are often absent from the FRI stand composition data. Thus, assessing and characterizing biodiversity in central Ontario using FRI maps in combination with landform types in order to identify reserve boundaries will, in some cases, result in the misguided placement of reserve boundaries and the eventual loss and degradation of genetic, species and ecosystem diversity.
To address this problem, we collected field data during 1997 and 1998 in order to determine how the vegetation classification based on FRI data would be affected by (1) additional field-based vegetation data including all trees (not just the uppermost canopy), the sapling/shrub layer, understory, snags, and logs, and by (2) stratification of FRI stands into slope position habitats. These results will used to propose a more refined approach to assessing and characterizing biodiversity in central Ontario and will be applied to identify and characterize a conservation corridor stretching from the eastern shore of Lake Superior to the Ottawa River. A major feature of the Superior-Temagami Corridor will be the protection and linkage of ancient forest landscapes in and between the Algoma Highlands, the Lower Spanish Forest, and the Temagami region.
The central axis of the Corridor has been identified, however, the northern and southern boundaries have not yet been determined. The most logical way to identify these boundaries in the context of conservation is to apply the principle of ecological representation. In other words, those ecosystems and species that will be included within the Corridor reserve should reflect the full complement of species and ecosystem types that are found throughout the broader Lake Temagami Site Region. Our current studies will provide the methodology required to place these boundaries so that ecological representation is achieved more effectively than currently practiced. This preliminary report addresses only the results of data analysis for the white pine-red pine community type (Maycock 1979) sampled in Temagami, Ontario.
Methods
To determine how well biodiversity is characterized by one particular landform-FRI community type, we sampled four FRI stands (79 plots) in the field representing the white pine-red pine community type on moderately broken shallow sandy till uplands. These stands are located in the Cassels-Rabbit Lakes area that is a few kilometers due east of the Town of Temagami, Ontario. Each stand was stratified into slope positions in order to evaluate the influence of habitat diversity on species composition. Slope positions included hilltops, upper slopes, saddles (upland valleys), lower slopes, and lowland valleys. In addition to sampling the upper-most canopy trees, which are assessed by FRI mapping, biodiversity was
further characterized by field-sampling sub-canopy trees, the sapling/shrub layers, understory, snags, and logs. Species and dbh for upper canopy and sub-canopy trees (10+ cm dbh) were recorded within a 20 x 20 m randomly located plot representing a particular slope position. Percent cover of all sapling/shrub layer vegetation (>.5 m height and <10 cm dbh) was determined by species within five 2.5 x 2.5 m sub-plots that were located within the larger 20 x 20 m plot such that four sub-plots were positioned in the corners and one was positioned in the centre. All understory (<.5 m height) vascular plants, mosses/liverworts as a group, and lichens as a group were assessed for % cover by species in 15 1 x 1 m quadrats that were located in three parallel rows of five along each side and down the middle of the 20 x 20 m plot. Snags within the 20 x 20 m plots that were greater than 10 cm dbh and taller than 2 m were identified to species when possible, measured for dbh, assessed for decay class, and assessed for woodpecker activity. Logs within the 20 x 20 m plots that had a minimum diameter of 15 cm at the large end and a minimum length of 1 m were identified, measured for length, measured for diameter at each end, and assessed for decay rank. Habitat conditions in the locale of each plot were also described and recorded. The samples from all four stands were ordinated using Detrended Correspondence Analysis. To evaluate the influence of adding biodiversity information, however, only two stands (stands 0549 and 4518; 40 samples) were compared. The FRI composition of stand 0549 (25 ha) located near Blueberry Lake was 50% white pine, 20% red pine, 10% white birch, 10% white spruce, and 10% poplar. The FRI composition for stand 4518 (83 ha) located in the White Bear Forest was 50% white pine, 20% red pine 10% white birch, 10% white cedar, and 10% balsam fir.
Results and Discussion
A comparison of the tree data ordination (representing the FRI data; Fig. 1) with the ordination of all the vegetation data (trees, understory, shrubs, snags, and logs; Fig. 2), shows that differences between FRI stands increase substantially as more biodiversity information is added to the analysis. This is shown by the increase of unique polygon ordination space for each FRI stand. For the tree ordination, 56% of the plots (22) are located outside of the shared ordination space of the two stands. When all of the data is considered, the percentage of plots located outside of the shared ordination space rises to 77% (30 plots).
When compared to the ordination space of the white pine-red pine community as a whole using all data sets (Fig. 3), it is obvious that the stratification of the FRI community by slope position (Fig. 4) can be a useful method of refining biodiversity assessment. This analysis reveals that no ordination space is common to all slope positions (Fig. 4), and that all slope positions show at least some unique ordination space. Valleys are completely unique among all slope positions. Hilltops and saddles show a slight overlap of ordination space with one another, while upper and lower slopes show the largest amount of overlap with other slope positions.
The results of these preliminary analyses show that the OMNR's forest vegetation classification, refined only to FRI community types, does not recognize significant ecological variation within the landscape at scales of resolution finer than the typical FRI forest stand. A vegetation classification system that recognizes the influence of slope position on vegetation composition will increase classification resolution, vastly improving assessment of forest biodiversity in the central Ontario forested landscape. In addition, the data required to achieve this improved classification discrimination is readily available at no expense to the government in the form of digital elevation maps. Only when we more accurately identify and characterize the species and ecosystems present on the landscape, can we improve our success at conserving species and ecosystems.
A common argument against this approach is that field work is very expensive and therefore, its application over large regions becomes operationally prohibitive. The development of geographic information systems, the availability of digital forest tree species (FRI community types) maps, and the availability of digital elevation maps, however, has made this refined approach much more feasible now than previously. For example, once we clearly establish relationships between vegetation composition and slope positions using field studies, these known relationships can be used to predict finer resolution vegetation units. For example, a digital slope positions map, created from a digital elevation map, could be combined with a digital FRI map to create a forest stand habitat map. The spatial nature of known forest stand habitat conditions can then be applied to more accurately determine the boundaries of protected areas that will include species and ecosystems that are representative of the central Ontario landscape.
Reference
Maycock, P. F. 1979. A preliminary survey of the vegetation of Ontario as a basis for the establishment of a comprehensive Nature Reserve System. Provincial Parks Branch, Ontario Ministry of Natural Resources, Queen's Park, Toronto, Ontario.
Produced by Ancient Forest Exploration & Research, main office: 154 Wright Ave., Toronto, Ontario M6R 1L2; field centre: R.R.#4, Powassan, Ontario P0H 1Z0, phone/fax (705) 724-5858
|
