METHODS OF ANALYZING FOREST CHANGES

A. Short-term historical methods

1. Written documentary sources (explorer's descriptions, diaries, etc)

Useful for gross changes, but reliability for subtle change is doubtful.

2. Historical maps of vegetation

Can be excellent, but early maps often used mapping units that were too broad (e.g. "tall forest," "closed forest," without sufficient detail on composition).

3. Terrestrial photographs

If available, excellent for documenting gross changes and generating hypotheses to explain the changes (especially when combined with other data).

Problems:

a. Locations may not be representative (e.g. along roads, railroads)

b. Changes are not easily quantified

4. Aerial photographs/remote sensing images

Excellent for quantifying changes in cover type. Problems:

a. Aerial photographs older than c. 1935 are rare.

b. Rarely allow identification of individual species.

5. Re-measurement of previously surveyed areas

a. Old vegetation surveys for earlier ecological studies or timber inventories.

b. Use of General Land Office surveys

B. Monitoring of permanent plots (including experimental plots such as exclosures)

Generally, the best source of data but records are scarce and usually short.

Considerations in installing permanent plots:

1. High cost

2. Location:

representative sites;

objective vs. subjective placement of plots to test particular hypotheses;

accessibility and ease of re-location by future works;

long-term protection of the site.

3. Measurements:

tree i.d. and permanent markers on trees;

diameter at breast height (dbh; accuracy);

tree locations (x and y coordinates);

tree ages and ring width samples for growth measurements;

tree heights;

understory composition, tree seedling populations (often with smaller plots);

importance of clear measurement instructions (e.g. size class definitions of seedling vs. tree, measurement of bifurcated trunks, how to measure dbh of trees with buttresses, etc.)

C. Inference from stand structure

Required information:

1. The species pool and where species "segregate out along environmental gradients" (i.e. which species can tolerate a particular site).

2. Basic autecological information such as dispersal mechanisms, vegetative reproduction, longevity, seed production and periodicity, etc.

3. Quantitative data on stand structure

Types of data collected:

i. tree population age structures (sometimes supplemented with size data)=the basic data

ii. tree spatial patterns (clumping, associations of size/age classes by species and micro-site)

iii. disturbance history (documentary, photographic, dendroecological, etc.)

iv. understory data (interference from non-tree species, seedling abundances and micro-sites)

v. tree canopy structure (density, cover; estimated or measured)

vi. vertical structure (relative tree heights; emergent, dominant, suppressed, etc.)

SEE “AGE STRUCTURE ANALYSIS” HAND OUT FOR TYPES OF AGE STRUCTURE

Limitations to Age Structure Analysis

A. Confusion of variation in past mortality and past recruitment rates.

Static-age structure curves (frequency distributions of numbers of trees in age classes) are shaped both by variations in past recruitment (e.g., past “pulses” of recruitment) and past mortality (e.g., effects of a past drought). This is the fundamental reason why static-age structure curves are not the same as a survivorship curve determined from a fixed cohort life table; tree mortality rates cannot be derived from static-age structure.

B. Errors in determining tree ages:

1. Ring-counting errors.

2. Missed pith.

3. Time required to grow to coring height.

C. Missing age data due to rotten centers.

D. Great sampling effort required to age trees over a large area.

4. Size (dbh) structure analysis (i.e. substitute tree size for tree age).

Restrictions/guidelines for size structure analysis.

A. Minimize dependence on size data and maximize collection of age data.

B. Test age/dbh relationship by sampling ages of large range of tree sizes in a uniform habitat.

C. Adjust the detail of the interpretation (and analytical procedures) according to the strength of the age/dbh relationship. Even if there is a strong relationship between age and dbh (e.g., correlation coefficents of 0.9) you can only make general inferences (e.g., persistence versus non-persistence) from size structure analysis and you cannot reconstruct the details of stand history or reach conclusions about the quantitative stability of the population.

D. Interpret data on a stand-by-stand basis before attempting to lump data from dispersed sample sites (likewise this is necessary in age-structure analysis). Problems with lumping result from:

 age/dbh relationship changes with site;

 disturbance history changes with site;

 regeneration modes become obscured;

 always try to combine age/size analysis with precise dating of disturbance events.