Generic Forest Health Surveys Guidebook
Table of Contents
The minimum requirements for summaries of forest health survey information are:
survey method, including sampling intensity and locations;
forest health factor(s) detected and percentage incidence (ha, trees);
forest health factor location(s); and
expected impact on resource values and treatment recommendations.
The following methods are provided to assist in fulfilling the minimum requirements for creating FDPs.
Aerial surveys have provided an invaluable tool for detecting and monitoring forest health factors for many years. This section provides the standards for aerial surveys and how they should be conducted. It incorporates the aerial survey procedures for forest health provided in Forest Health Aerial Overview Survey Standards for British Columbia, Version 2.0 (CFS and BCMOF 2000; Resource Inventory Committee [R.I.C.] approved). Training for aerial overview surveyors is conducted by the B.C. Ministry of Forests (BCMOF) and training materials are provided in the Aerial Overview Surveys Training ProgramParticipants Guide (BCMOF 1997a).
For operational surveys, surveyors should contact the regional forest health specialist for codes and mapping standards for that region.
Aerial overview surveys are generally conducted from early July through September to coincide with the optimum damage symptom expression of major forest pests and damage in British Columbia. By that time, sufficient knowledge about current pest conditions has been gained from early-season surveys and anecdotal reports to allow management planning. Occasionally, special flights are conducted to address specific pests that express themselves either earlier or later than the mid-summer period.
Co-ordinated planning is essential to a successful aerial survey. Generally, for overview flights being performed by the BCMOF, initial aircraft selection and charter arrangements are done by the BCMOF through the Regional Fire Centre, which will ensure that aircraft and pilots meet specific training and safety requirements. Reserve an aircraft well in advance of the expected flight, because planes could be in short supply during fire season and alternative arrangements may have to be made. Mapping personnel should also be given as much advance notice as possible to accommodate flight scheduling. Aircraft charter companies should be informed that surveys are weather-dependent and that final decisions on suitable flying conditions cannot be made until the day of flight. The aircraft user must also be aware that short-notice cancellations may cause the air carrier to bill the Ministry a detention fee or other charge as allowed for in their tariff schedule.
Map scale will be determined not only by availability, but also by product requirements. For both national and provincial overviews, pest information is usually recorded on coloured provincial or national topographic series maps of 1:100 000 or 1:125 000 scale, or on 1:250 000 when those scales are not available. While larger-scale maps allow for greater accuracy and detail, the use of scales such as 1:40 000 or 1:50 000 is more appropriate for operational surveys that require greater accuracy. The 1:100 000 topographic maps are produced by Canadian Cartographics Ltd. in Coquitlam, B.C. (1-877-524-3337). More lead time may be required to gather the necessary maps if they are scarce. If maps are unavailable, the last option is to order customized maps produced through a Geographic Information System (GIS). Availability of these maps is highly variable or may take a long lead time if workloads are high. Consult the local forest district office to determine if maps are available at the scale required.
At least two copies of each map are needed, one as a working map and the other as a clean summary for digitizing. As an aid to detection, the working copy may include the previous years infestations plotted by GIS. This enables the observer and pilot to plan the flight efficiently and accurately locate areas where pest damage is to be mapped. It also allows any more expansions and changes over time to be checked and better identified. Flight lines with directional arrows are recorded on the map as the flight progresses. The date of the survey and the names of observers should also be noted on the map. Since space in an aircraft is at a premium, excess paper on the map edges is often trimmed away and the maps are folded.
Each observer should be equipped with a supply of pens and sharp pencils, binoculars, camera, extra film, amber-tinted sunglasses, a lunch, and motion sickness medication if needed. The aircraft must be equipped with radio headsets. Communication with the pilot and other observers is very important. If radio headsets will not be worn throughout the flight, foam ear-plugs or some other form of hearing protection should be used. Aircraft must be equipped with a radio programmed for BC Forest Service District and Air Operations/Fire Centre check-ins.
Aircraft selection may be largely determined by local availability, but should be of high-wing configuration for ease of lateral and downward viewing, have seating capacity for at least four, and be capable of sustained flight of 8090 knots. In remote coastal applications and some northern locations, a float or amphibious version is often more desirable, due to better fuel availability and landing opportunities. In the central and southern interior of the province, wheeled aircraft with fixed or retractable landing gear are preferred.
Some phases of the aerial survey require low-level detection or reconnaissance. This type of flying falls into a speciality flying category for Forest Service flights. Low-level reconnaissance exposes both pilot and observer(s) to a higher degree of risk versus high-level mapping. This risk is deemed to be acceptable, given that the pilot has been trained and certified by the company in the operational and safety aspects of this activity, and that other Operations Manual requirements are met. Ministry Fire Centre staff should be able to assist in determining the certification status of a pilot. In addition, observers must be briefed on the operational and safety aspects of this procedure. As in any Forest Service flight, passengers on board shall be limited to pilot(s) and essential personnel only.
Type of terrain and area of coverage will determine performance characteristics of the aircraft. Over flat and rolling landscape or small drainages, a Cessna 180 or equivalent may be sufficient, while in mountainous terrain, an aircraft with stronger performance such as a Cessna 210 or twin-engine Cessna 337 may be more appropriate.
Although aircraft availability and type may be limiting factors, safety should never be compromised. To avoid fatigue and loss of concentration, daily flights should be limited to 5 hours duration. Also, ensure that you are well rested before flight and avoid changes in diet prior to flying. For overview surveys, helicopters are not cost-effective and are usually limited to the occasional pest identification or assessment in otherwise inaccessible areas or as a follow-up after the initial fixed-wing flight.
Weather is one of the most critical factors governing the success of an aerial survey and an essential part of pre-flight planning. Regardless of the prevailing weather, a daily weather forecast describing flying conditions should be obtained to ensure that there is good visibility and a minimum ceiling of about 1000 m (3000 feet). Local weather information can be obtained by calling the forest district office in the area scheduled for mapping. Clear and sunny days are preferred, to maximize detection of defoliation and bark beetlekilled trees for mapping and photography, but solid, high overcast, giving the forest a monochromatic look, is also acceptable. Broken cloud conditions, where one is constantly shifting between sun and shade, are extremely difficult to map under, as the eyes are forced to adjust every time the light changes. Such conditions are very fatiguing, and important infestations can be missed in the blind spots. Since shadow from low sun angle can obscure features early or late in the day, especially in areas of significant topographic relief, the optimum flight period is between mid-morning and mid-afternoon, when the sun angle is highest. Typically, some flight adjustments may be required when dealing with unstable air in the afternoon.
All BCMOF use of aircraft and the operational procedures associated with that use shall be planned comprehensively and in detail by the aircraft user, pilot/air carrier, flight watch authority, and/or local Fire Centre. Aircraft users must also receive adequate pre-job instruction, including safety procedures, before actual operations commence.
Prior to each flight, the pilot and observer(s) shall review the proposed mission. The material to be covered must include but is not limited to the following:
All BCMOF personnel who use aircraft should be familiar with the BCMOF Aviation Safety Manual (1997b) produced by the Aviation Management Section, Protection Program. This document describes BCMOF policy on minimum requirements for air carriers, pilots, speciality flying, flight safety and guidelines, and other safety matters.
Prior to each flight, onboard personnel and the ground communications centre should know the intended flight plan and duration. Known as "positive flight following" or "flight watch," location updates are radioed every 30 minutes to either the BCMOF Fire Centre or the appropriate district office, depending on the local protocols. Radio headsets must be requested for both hearing protection and flight communication. At the very least, in the event the aircraft is not equipped with radio headsets, noise abatement equipment such as foam ear-plugs should be used.
Forest Service passengers travelling over remote areas of the province must wear the appropriate clothing and footwear needed to survive in the event of a forced landing. Passengers should avoid wearing synthetic fabrics such as nylon or polyester. Natural fibres such as wool and cottons offer some measure of protection. Employees who fly regularly in the course of their duties or are involved in low-level operations should consider wearing flame-resistant flight suits or coveralls.
Amber-coloured sunglasses are often used for both eye protection and to enhance colour differentiation on the ground. Be in frequent communication with the pilot regarding direction, altitude changes, air speed adjustments, fuel considerations, meteorological conditions, and ferry time estimates. Do not hesitate to ask questions or discuss with the pilot anything that causes you concern. While the observer who chartered the aircraft has jurisdiction over the basic flight procedure, the pilot is ultimately responsible for the aircraft and the safety of the passengers, and may overrule any aspect of the survey plan with respect to aircraft operation and safety. Conversely, if you feel that the aircraft is not being flown in a safe manner, you should terminate the flight and report the incident to the BCMOF Fire Centre and the charter company. Please forward an incident report to the Fire Centre, as described in the Aviation Safety Manual (BCMOF 1997b).
While the normal flying height is usually between 500 m (1500 feet) and 1000 feet (3000 m) above the terrain, a minimum flying height of 160 m (500 feet) above ground level must be observed as a safety precaution, such as when crossing ridges between drainages. Depending on the type of aircraft used, minimum airspeed should range between 70 and 90 knots.
The primary limitation of the overview aerial survey is one of perception, particularly as it pertains to bark beetles. Some forest managers may expect to be able to make stand-level decisions on the basis of the overview, when this is clearly beyond its scope. Generally, only estimates of current damage are given, while older tree mortality is usually not included in the total area figure. However, mortality estimates, if applicable, are made following the collapse of defoliator infestations. If the intent of the overview survey program has been consistently met, estimates can also be made of cumulative mortality caused by bark beetles in specific stands by overlaying successive years of damage. Additional ground survey assessments are needed to calculate the total extent of pest incidence and damage. In the absence of more detailed information, aerial-sketch mapping results should not be extrapolated beyond reasonable bounds and expectations.
Throughout the province, standardized coding of identified infestations is required to allow the creation of a provincial summary. The standardized procedure and codes are provided in Forest Health Aerial Overview Survey Standards for British Columbia (CFS and BCMOF 2000). Some regions and districts, such as the Cariboo Forest Region (Howse 1995), have specific aerial survey procedures that may exceed the provincial standard.
Ideally, two observers are used, one on either side of the plane, to expedite coverage and improve accuracy. The forward observer is usually the more experienced individual for the particular area, and has the overall responsibility for flight direction, altitude, and speed. With attention to elevation, map contours, and natural features, the location, relative size, severity and damage, and probable cause are delineated on topographic maps. As infested areas are detected, they are plotted on the map, either as a polygon or as a dot representing infestations of less than 1 hectare. When two people are mapping, ensure that their field of view does not result in double mapping.
Plan a flight line that covers the survey area. Topography will usually have an influence on the route. Over level terrain, flight lines are usually flown on a parallel grid with some overlap, so that no area is missed. In mountainous terrain, contour flying is most efficient with one or more passes through a watershed, depending on its size and lighting. In some instances, a zigzag flight through a valley may be sufficient when only one pass is made. This action gives the opportunity to map pest damage behind and below the aircraft as well as laterally. Flight lines should always be marked on the map with arrows showing direction. Some oblique photography or video is recommended for a visual record, for a training guide, and occasionally to refine sketch maps and the assessment of damage. After each flight, both mappers are to compare their respective maps and produce a composite that later will facilitate GIS entry at the office.
The detection and location of damage should be accurate to the scale of the map used. However, when using smaller-scale maps such as 1:250 000, the size of infestations is frequently exaggerated, especially when small pockets comprised of 550 trees each are mapped. This was found to be true in comparisons of selected outbreaks shown on aerial photographs versus sketch mapping. Harris and Dawson (1979) found the total area sketch mapped to be 34% larger than measured on photographs, and similar results were obtained by Gimbarzevsky et al. (1992) in comparison data from ground plots, aerial sketch mapping, and various types of remote sensing. As a rule, the largest topographical map scale available should be used, normally up to 1:100 000. Occasionally, larger-scale maps up to 1:50 000 are used, but the large number of maps required for overview coverage makes organization and sorting in the cramped environment of a small plane difficult and time consuming.
While classification of damage is a subjective judgement by the observer, past surveys have shown that experienced personnel can estimate damage intensities fairly accurately. Help in maintaining accuracy and consistency can be obtained by referring to photo examples (CFS and BCMOF 2000), by taking periodic flights with others, and through quality check flights.
Observable damage symptoms can vary among the different bark beetles and between bark beetles and defoliators. Some defoliators can be differentiated by their damage patterns. It is important that the observer recognizes these differences. Some of the types of damage visible from the air include:
Pine foliage killed by mountain pine beetle initially appears chlorotic, then gradually turns yellow and fades to red within 1 year of attack. Trees killed by Douglas-fir and spruce beetles have variable foliage colour, and red or brown trees can still contain live beetles. Consult the Bark Beetle Management Guidebook or a regional entomologist for more details on the colour changes of trees killed by bark beetles. Because new attacks are not detected by aerial surveys, ground assessments are made to determine current infestation status. For aerial survey purposes, a red tree is one that was attacked and killed the previous year. These are the trees that are mapped. Grey trees are those that have been dead for 2 or more years and that should have been mapped during a prior survey. Small infestations of up to 50 trees may be located on the map as a dot, with the number of trees and the abbreviation for the appropriate tree species beside it. All dots (point sources) are classified as severe. For GIS input, the following scale is applied to area estimates:
230 trees = 0.25 ha;
3150 trees = 0.50 ha.
For larger areas, a polygon is drawn around the infested trees and marked
with the appropriate damage classification, as follows (only red trees are recordedor note if otherwise):
Light = 110% of trees recently killed
Moderate = 1129% of trees recently killed
Severe = 30%+ of trees recently killed
Grey (Old) = tree mortality 2 or more years old (generally not mapped)
Defoliated trees, stands, or hillsides assume a reddish tinge as a result of active feeding on the foliage. Only the current years feeding damage is mapped. In areas where severe defoliation has occurred for several years, trees with little or no remaining foliage may appear grey. In light infestations, close observation is necessary because defoliated trees do not readily stand out. Defoliation intensities also tend to fade into each other, and subjective delineations must often be hastily made between areas of differing intensity. When possible, ground checks should be done to verify identification of the defoliator, particularly in new infestations. Following are the severity classes normally used to help classify an infestation:
Light discoloured foliage barely visible from the air, some branch tip and upper crown defoliation.
Moderate pronounced discoloration, noticeably thin foliage, top third of many trees severely defoliated, some completely stripped.
Severe bare branch tips and completely defoliated tops, most trees sustaining more than 50% total defoliation.
Classification of tree mortality caused by defoliators is the same as that for bark beetles. Accurate classification of mortality for deciduous trees is difficult to achieve from the air unless the trees are obvious snags, the surveyor has prior knowledge about the previous years defoliation, or a ground truthing is available.
While bark beetle and defoliator infestation assessments are the main targets of the aerial survey flights, other types of damage are noted if the observer considers this damage significant. Other forest disturbances mapped during regular aerial surveys include blowdown, winter damage, animal damage, flooding, foliage diseases, root rots, and pollution damage. Observable damage symptoms can vary considerably between each cause, or be very similar.
Whereas blowdown and flooding are usually easy to recognize due to their physical characteristics or association, others, such as winter damage and foliage diseases, are more difficult to identify because they can mimic other types of damage, such as defoliation (and vice versa). Root rot disturbances are also difficult to map, due to the scattered nature and various stages of decline of infected trees.
In summary, the following damage agents and conditions sometimes observed during aerial surveys can be confused with damage caused by insects:
· porcupine feeding;
|· large cone crop;
· needle diseases;
· root rots;
· fire damage;
· flooding; and
· pollution (ground-level ozone).
The overview aerial survey is designed to incorporate mapping of visible damage from as many forest pests as possible in one flight. However, the period when damage (primarily from insects) is most visible varies with the pest species and its geographic distribution. In most cases, there is sufficient overlap of defoliator damage and bark beetlekill to properly schedule both types of damage in the same survey. The normal aerial survey period (the "biological window") in British Columbia is between early July and late August, which provides maximum detection of common pests with a minimum of duplicate flying (Table 2). Winter moth, tent caterpillars, spruce aphid and lodgepole pine needle disease are examples of some common pests that do not fit the general biological window and that may require separate surveys prior to July 1.
Ideally, detection of damage should be performed during the optimum colour change or defoliation related to the pest. Delays in surveying may result in significant underestimation of damage levels. For example, wind, rain, or snow storms may remove red or damaged foliage, making accurate defoliator damage identification very difficult.
Table 2. General biological window for aerial survey mapping of bark beetle and defoliator damage in British Columbia
(Pest codes with a "?" indicates that a species code is not available).
mountain pine beetle
|western white pine||IBM||mountain pine beetle||Julyearly September|
|whitebark pine||IBM||mountain pine beetle||Julyearly September|
|ponderosa pine||IBM||mountain pine beetle||Julyearly September|
|IBW||western pine beetle||Junemid-August|
|lodgepole pine||IBI||engraver beetles (Ips spp.)||JulySeptember|
|white spruce||IBS||spruce beetle||mid-Juneearly September|
|Douglas-fir||IBD||Douglas-fir beetle||mid-Juneearly September|
western balsam bark beetle
|grand fir||IBI||fir engraver||early Julylate August|
|Defoliators||Douglas-fir||IDW||western spruce budworm||early Junemid-August|
|IDT||Douglas-fir tussock moth||mid-Julylate August|
|IDZ||false hemlock looper||mid-Julylate August|
western blackheaded budworm
|IDL||western hemlock looper||mid-Julyearly September|
|IDG||green-striped forest looper||mid-Julyearly September|
|?||saddleback looper||mid-Julyearly September|
|?||phantom hemlock looper||early Julylate August|
|?||gray spruce looper||late Augearly October|
2-year-cycle spruce budworm
|IAB||balsam woolly adelgid||Aug. through September|
|IDE||eastern spruce budworm||late Juneearly July|
|amabilis fir||IAB||balsam woolly adelgid||Aug. through September|
|grand fir||IAB||balsam woolly adelgid||Aug. through September|
pine needle sheath miner
|IDS||conifer sawfly||mid-Julylate August|
|pine butterfly||July through August|
March through June
|white spruce||IDB||2-year-cycle spruce budworm||June|
|IDE||eastern spruce budworm||late Juneearly July|
|IDP||larch sawfly||late Julyearly September|
|larch budmoth||early Julymid-August|
|Deciduous||IDF||tent caterpillar||early Juneearly July|
|IDU||satin moth||early Junemid-July|
|IDX||large aspen tortrix||early Junemid-July|
|IDN||birch leaf miner||early Junemid-July|
pine needle cast
May through June
|?||winter drying||April through July|
Daily mapping results should be compared among observers, and a composite (master map) drawn after each flight while visual image retention is still good. The product should be a quality sketch map suitable for digitizing or photocopying. Each map should have a standard colour-coded legend representing each pest mapped. Additional data include date of flight, names of observers, type of aircraft used, and comments on weather and visibility. Upon completion of the composite map, current infestations and areas of damage are entered into a GIS, from which the completed data are ultimately distributed. Because GIS-generated maps appear clean and professional, it is easy to make assumptions about their veracity, but the results are only as good as the data entered.
Pest data from maps are recorded by digitizing the polygons and assigning attributes of pest severity, year, forest region, and map reference. From these data, searches or compilations of any combination of desired attributes can be made. During digitizing, the current and previous years infestations can be viewed on the screen, providing an opportunity to make changes before entry into the database. A final edit of the digital map against the sketch map is required. A legend should be produced to accompany the map, according to the standards outlined in Appendix 2 of the aerial overview standards publication (CFS and BCMOF 2000). Observers should input their own data, so that errors and omissions can be minimized during digitizing. However, increasingly, the input of map data will be by people other than those participating in the actual mapping and this will leave little basis for decision making if discrepancies occur. GIS reproductions at various map scales are distributed to co-operating agencies such as the forest industry and BC Parks. Using report generators, area and polygon tallies can be derived for selected areas, map sheets, administrative regions, or the entire province/territory.
The BCMOF has now assumed the data custodianship responsibility for recording, reporting, and storing aerial overview survey information. A set of digital data standards has been produced that will be followed to facilitate the seamless roll-up of all new overview survey data collected throughout the province. As well, standards for metadata to accompany mapping data are available. Finally, data transfer standards should be adhered to. This information is provided at the website address: http://srmwww.gov.bc.ca/risc/standards.htm
Aerial surveying is not an exact science, but an observer should do everything to ensure that the best calls are made. Credibility comes from following established criteria:
Aerial sketch mapping can be enhanced with the use of aerial photographs, especially in areas of extensive pest damage on even terrain with few geographical features. Up-to-date aerial photos can be useful in showing logging, burns, and other details that observers can delineate from infested timber. If available, custom-drawn GIS maps (scale 1:100 000) that highlight cutblocks, roads, water bodies, and other landmarks greatly improve the observers ability to orient themselves quickly and thus enhance the accuracy of pest polygon placement.
Studies have shown (Harris and Dawson 1979; Harris et al. 1982; Gimbarzevsky et al. 1992) that defoliation estimates are frequently exaggerated during sketch mapping, while counts of bark beetlekilled trees are low when compared to aerial photographs, ground plots, and some remote sensing techniques. For a given area, assessment of aerial survey accuracy and presence of bias are best determined using a multi-stage sampling procedure, which is comparing sketch mapping, aerial photography, and ground plot data.
Periodic check flights of overview surveys should be done by experienced observers to maintain the accuracy and precision of pest assessments within acceptable parameters such as the qualitative and quantitative criteria listed above. The recommended process is as follows:
1. Make a flight audit no more than 2 weeks after the initial survey and remain within the biological window of the pests mapped.
2. Identify the pests to be assessed from the map legend.
3. Use the same map scale and any previous data.
4. Randomly select sample polygons or dots representing 510% of the total area mapped. This can be done by plotting transects through infestations, and mapping only intersected polygons. Normally, this should not amount to more than the equivalent of 1 weeks flying for all of areas of British Columbia.
5. Analyze and compare both maps against established criteria such as pest and host identification and damage intensity levels.
6. Ensure that the level of accuracy is proportional to the degree of mapping difficulty (e.g., scattered occasional defoliation versus extensive defoliation mapping). However, acceptable limits of accuracy are expected to be within 30% plus or minus the check flight assessment. When those limits are exceeded, the observer should be re-assessed to determine the source of discrepancy.
The scheduling of harvest and management of the forest resource is largely based on the information available in the BCMOF forest inventory system (FIS).1 This inventory is used to generate forest cover maps, and displays information by polygon of tree species composition, age, site index, area, and other information. Long-term planning is based on the information obtained from the FIS; however, this inventory lacks some information on forest health factors (pests). The incidence and relative severity of pests must be quantified at the forest landscape level in order to incorporate their management in the planning process.
The survey for pest incidence (SPI) identifies pest-specific incidence and severity at the landscape level, based on a continuous series of 100-m-long plots that vary in width depending on the age and density of the stand.
The SPI identifies pest-specific incidence and severity at the landscape level. It is designed to build a database of accessible information, linked to forest inventory type groups, that can be used in the planning process to enhance existing annual aerial survey information and hazard/risk rating systems.
The survey attempts to record and quantify incidence and severity of all damaging agents within an area and gather stand information, including stem density, species composition, and stand structure and age. Pest incidence can be assessed on a simple presence or absence basis or by severity. For some forest health factors, there is relatively little information correlating incidence and impact; therefore, only occurrence (detection) is noted. For other forest health factors, this information is available, and specific severity categories are available that can give an estimate of impact. Although a SPI may be done in any stand type or age class, it is not intended to be used in very young regeneration, but rather in established plantations and forested areas (greater than 10 years old).
The land use priorities of the area to be surveyed should be known to the surveyors, because this might affect the forest health factors that are given priority for identification. For example, if the survey area includes identified valuable wildlife habitat, forest health factors that may be affecting vegetation particularly important to the habitat should be identified in the survey.
Results from SPI provide an overview of the state of forest health within the area of interest by providing a snapshot in time of the forest health factors that are present and that have occurred in the past. As the database becomes more comprehensive, the confidence level for predicting the occurrence of forest health factors in specific forest types will improve, thereby highlighting forest health opportunities in the planning process.
The SPI can also yield information on stand dynamics if survey lines are re-visited at specific time intervals (e.g., 5- to 10-year intervals). More detailed studies (e.g., permanent sample plots), usually done over a period of time, are required to quantify the impact of a specific pest or pest complex.
[1 Tree Farm Licences will have separate forest inventories retained by the holder of the licence.]
The following procedures are recommended for a landscape-level assessment of pest incidence. First, identify the geographic area of interest. The area of interest can range from a cutblock, a mapsheet unit, and a single drainage to even larger geographic units. Second, determine the number, distribution, and relative proportion of inventory type groups (ITGs) in the interest area (Appendix 2). This information can be retrieved in digital format from the forest inventory database. Plot intensity per ITG should be proportional to the area (ha) occupied by each ITG. The total number of plots is dependent upon the confidence level desired. The SPI can also be applied at the stand level to determine pest incidence and severity for writing prescriptions. The intensity of sampling should be higher in this case, to obtain the detail of information required.
The area of interest will be defined as one mapsheet unit for the purpose of describing the following SPI procedures.
a) For the mapsheet, retrieve the following inventory information, by polygon;
A forest cover "1" map file (FC1) may be obtained from BCMOF Resource Inventory Branch - Forest Resources Geographic Information Systems (FRGIS) with an accompanying Forest Inventory Planning (FIP) summary report file. The FIP file is sorted by species, ITG, and age class, and gives the total area for each grouping (Table 3). Please note that these files are gradually being replaced by the Vegetation Resource Inventory system (FIP by Vegetation Inventory Files [VIFs], and FC1 by Vegetation Inventory 1 File [VG1]).2 [2 Consult Appendix 1 for a list of acronyms used in this guidebook]
Table 3. Example of a FIP map summary report, showing forest cover data from the FRGIS system
|Species||ITG||1||2||3||4||5||6||7||8||9||Total area (ha)|
The data summary obtained from the FIP file also gives discrete polygon information useful for selecting plot locations.
Table 4. Example of data from a FIP file listing attributes by polygon
|MAP NUMBER - 082L052
POLYGONS MEETING SELECTION CRITERIA INCLUDE:
|SPECIES Fd||PLAT||ATPLFD||S ATB||ATE||FDS B||S FDE||PLS||S FD||FDS||S FD||S FD|
|TOT VOL. 1||768||5457||17662||321||974||15959||1167||0||2655||9339||2729|
|% CONTR. 1||0||100||100||100||100||100||100||100||100||100||100|
b) Sort the above inventory information in a hierarchical fashion by:
Exclude all land attributes that are not within the forest to be surveyed, such as alpine, inoperable, lakes, private land, and open range.
c) The inventory data are grouped into strata of similar attributes for the purpose of determining a sampling matrix. Division into sampling strata is a subjective exercise based upon management objectives, composition or diversity of the mapsheet, desired sampling intensity (Appendix 2), and relative size (ha) of strata breakdowns.
The following is a decision matrix for creating sample strata:
d) Determination of the optimum number of strata is dependent on several criteria;3
[3 The SPI survey was developed on a mapsheet basis and typically 2025 strata per mapsheet were selected. A 1:20 000 forest cover mapsheet has approximately 16 000 ha gross area.]
Regardless of the number of strata selected, each plot within a stratum is ultimately tied to polygon number, ITG, and mapsheet. Therefore, this is simply an exercise to achieve proportional sampling distribution over the interest area.
Each mapsheet is divided by ITGs so comparisons can be made among mapsheets. This gives the SPI a common language. The total area in each stratum is compared to the netted-down mapsheet area and is then assigned a proportional number of plots for that stratum.
Use the following steps in Figure 3 to determine optimal number of plots (see also Appendix 2):
Figure 3. Calculation of optimal number of SPI plots.
1) Total area of interest = 4221 ha
2) @ 0.05% coverage = 0.0005 x 4221 ha = 2.11 ha
3) For plot area of 0.02 ha (2 x 100 m):
2.11 ha = 105.5 plots (round to 106)
4) Divide plots proportionally among the strata (Appendix 3):
749.7 ha = x ; therefore x = 18.8 (round to 19)
4221 ha 106
|Strata total||Estimated number of plots|
|71.0 ha||1.8 = 5a|
a The minimum number of plots in a stratum is five.
Planning line and plot locations. Location of SPI lines and plots are determined as follows:
If a line continues from one stratum into another, have at least a 100-m "no tally" portion between the strata break to allow for compass or map error. This will ensure that the next plot is actually within the new stratum.
Each SPI line needs a unique POC that is clearly marked and labelled with SPI line number, compass bearing and distance to Plot 1, date of survey, and area identification (mapsheet and polygon number). Plots must be numbered consecutively along lines and require only a semi-permanent marker between plots. At the beginning and end of each plot, or "no tally" portion between plots, record line number, plot or "no tally" section completed, plot or "no tally" portion to commence, bearing, and distance from the POC. "No tally" sections include the portion of the SPI line from the POC to the start of Plot 1 and any other portions of the line between plots that are not surveyed. The surveyor should insert a "no tally" section when crossing from one stratum type into another (minimum 100 m) or when features such as swamps, roads, or other significant non-forested types are encountered.
The SPI is a continuous series of 100-m-long plots. Plots can vary in width from 1 to 5 m and should contain between 20 and 50 trees, depending on the age and density of the stand (Table 5). At the beginning of each 100-m plot, the width can change, but the width cannot change within a plot. A more detailed description of plot widths is provided in the subsection entitled "Plot parameters." Every tree greater than 1 m in height within the plot boundaries will be examined and tallied, by species, for pest occurrence and severity. The POC for all SPI plots and the point of termination (POT) should be a minimum of 25 m from the edge of forest cover polygons, roads, and unnatural openings.
The plot centre-line should be marked with ribbon accurately and clearly in a highly visible colour. Endeavour to place ribbons exactly on the centre-line, and at least the last two ribbons placed must be visible. This ensures that the line will be straight for accuracy of plot assessment. A different ribbon colour should be placed at the commencement of each new plot or "no tally" section.
The SPI line is established using a hip-chain measuring device. The line should be measured and marked prior to recording pest incidence and tree data. While establishing the line, the surveyor should note stand density and composition per plot in order to estimate the plot widths that will yield the minimum number of required trees, and to identify potential sample trees. On field notepaper, sketch the location of the line, noting bearing, plot locations, "no tally" sections, and any pertinent physical features.
The objective of the particular SPI survey will affect the plot parameters.
You should decide what are the forest health factors and host trees of interest. Decisions should also be made on the need to collect data on tree layers (including seedlings and vets), and any other information on the stands.
Tree layers. Depending on stand age, SPI plots may be divided into two layers. In mature stands, layer 1 (L1) includes overstorey trees (dominant, co-dominant, intermediate, and vets). Layer 2 (L2) is defined as understorey or suppressed trees (Figure 4). In a mature, single-layer stand, there is often no true L2; therefore, simply assess understorey trees as they are encountered within the width of the mature strip. In this case, as long as L1 tree minimums are met, the 10-tree L2 minimum is not critical. In a true multi-storeyed stand, 10 trees in the L2 layer should be tallied. Note that the SPI procedure is not designed to examine regeneration; therefore, there is no L3 layer.
Immature stands usually contain only one layer or age class, and this layer is designated as L1. If two distinct layers (age classes) are present, then both an L1 and L2 layer are recorded, as in mature stands.
A minimum number of trees must be assessed in each 100-m plot. Minimum tree numbers have been set for L1 and L2 (Table 5). Plot width is chosen by considering the density of the stand and the minimum number of trees needed. The strip width of L1 and L2 can differ in any given plot to achieve the minimum number of trees. For L1, the minimum strip width is 1.0 m and the maximum is 5 m. However, if the minimum number of trees has not been met, do not exceed the plot width of 5 m. Layer 2 often (and Layer 1 rarely) has high stem densities; therefore, the minimum strip width for L2 (or L1) can be as low as 0.5 m. For example, L1 could be 4 m and L2 could be 0.5 m in a mixed-age stand with a relatively open overstorey, and a dense understorey. Tally trees in the L2 layer only for the first 10 m of the plot if the L2 density is exceedingly high.
Table 5. Target number of trees per 100 m SPI plot, by stand type
|Type||Target number of trees per plot|
|Immature stands (single age class)||SPI layer 1 = 50|
|Mature single-age stands / mixed age stands||SPI layer 1 = 20, excluding vets|
|Understorey or suppressed trees||SPI layer 2 = 10|
Figure 4. SPI plot showing minimum plot length and widths for L1 and L2 layers.
The following describes the procedures and type of information needed for completing a SPI form (Appendix 4).
a) Line Location Information:
b) Pest Incidence Information:
Figure 5. Enlargement of pest code and severity section of SPI form.
(see Appendix 4 for a copy of the SPI form and Appendix 5 for a list of damage agents, their codes, and pest severity ratings.)
The following describes the example information contained in the Figure 5 example form: I
c) Sample Tree Information:
Tree tally. The surveyor tallies all trees and stumps, by species, within the plot. Trees are in the plot if more than 50% of the tree base diameter falls within the plot boundary. Assess each tree or stump for pest occurrence. Individual pest codes incorporate the tallying of dead trees and stumps uniquely according to causal agent. When no pests or damage are detected, tally live and dead trees by species and layer. When pests or damage are identified, record (Appendix 4):
The objectives of the particular SPI survey should be determined in detail prior to the field work. For example, young tree regeneration is not normally included in the tree assessment and tally. If it were included, it would probably be an L3 layer (separate plot). Normally, the survey will consider trees either as part of a one-layer stand (L1), or a two-layer (L1 and L2) stand. The tally and assessment for the layers should be recorded as separate overlapping plots. Plot widths for the layers may differ, and the minimum tally number for plot trees may have to be achieved separately.
Tally the total number of live trees per plot. Tally dead trees, where the cause of mortality is known, with the particular damage agent codes. Tally dead trees in the "dead" column (where no damage agent code is identified) only where the cause of mortality is unknown.
All standing trees in a 100 m transect plot area are assessed as living or dead. Dead trees greater than 3 m in height are considered trees, and dead trees or stumps less than 3 m in height are considered stumps. Recently created stumps must be assessed for pathogens. Recent windthrow, whose point of germination lies within the plot boundary, must be assessed for pest occurrence (Appendix 5).
Tally every standing, live tree encountered, by layer, in the total live trees column. This will avoid "double-tallying" of trees that have more than one pest. Trees with multiple pest occurrences get tallied once under each appropriate pest and severity column. Therefore, a tree with three different pest problems would be tallied three times but only once in the total tree column. You cannot sum the number of trees unaffected by pests (clear) and the trees tallied for each damage agent and expect to correctly determine the number of live trees per plot.
When a tree is free of any pests, it will be tallied as "clear" in the pest incidence information section. If a dead tree, stump or windthrow is encountered and the causal agent cannot be determined due to the tree being "dead" for some time, tally in the dead column in the pest incidence information section.
Within each plot, one relatively pest-free tree representing the majority species and size of the dominant or co-dominant layer will be chosen. For this tree, the following data will be taken and recorded in the sample tree information portion:
Vets are not to be chosen as sample trees.
This information can be used for inventory purposes, and for SPI analysis of plot data.
Severity is a measure of the forest health factor's impact or abundance on the host tree. For each pest encountered, an assessment of severity is done. This assessment is reflected in the column (1, 2, or 3) under which the tree is tallied on the SPI form (Figure 5; Appendix 4). Columns 1, 2, and 3 reflect most severe to least severe impact or abundance of the pest, respectively. Not all forest health factors are necessarily divided into three severity categories. Some may have only one or two severity categories. For example, balsam woolly adelgid (IAB) is tallied only under column 1, indicating presence of the pest. No further assessment of its impact or abundance is noted because there is presently little information on this pest's impact. Lodgepole pine dwarf mistletoe (DMP) on the other hand, is recorded in columns 1, 2, or 3 as follows:
Column 1number of trees rated at 5 or 6 on the Hawksworth scale.
Column 2number of trees rated at 3 or 4 on the Hawksworth scale.
Column 3number of trees rated less than 3 on the Hawksworth scale.
Follow the example below to rate the tree:
Figure 6. Hawksworth dwarf mistletoe severity scale.
In the case of lodgepole pine dwarf mistletoe, all three columns are utilized because the Hawksworth scale is a well-quantified methodology for assessing dwarf mistletoe impact.
For the survey area, determine the overall Hawksworth scale rating by summing the trees per column and averaging the ratings.
To interpret pest incidence and impact at the landscape level, a common measure of comparison must be used. To facilitate this comparison, a numerical pest severity rating (PSR) was developed (Table 6; Appendix 4).
The PSR has four categories:
1 = Mortality;
2 = Major volume loss;
3 = Minor volume loss; and
4 = Insignificant volume loss; defect.
A PSR was assigned to each assessment level (columns 1, 2 and 3 on the SPI form) for each pest. For example, when the number of trees with lodgepole pine dwarf mistletoe is recorded in column 1 or 2 on the SPI form, this identifies a PSR of 2, reflecting that this pest causes a major volume loss. When the number of trees with DMP is recorded in column 3 of the SPI form, a PSR of 3 is identified, indicating a minor volume loss (Table 6).
Table 6. Example of a selection of pests according to types of
(see Appendix 5 for complete list)
|SPI form column|
Microsoft ACCESS® could be used for the storage and analysis of SPI data. Quoting the ACCESS® Help introduction: "Using Microsoft ACCESS, you can manage all your information from a single database file. Within the file, divide your data into separate storage containers called tables; view, add, and update table data using online forms; find and retrieve just the data wanted using queries; and analyze or print in a specific layout using reports."
Someone suitably trained on ACCESS could develop a series of linked "tables" of the SPI data. Alternately, a generic ACCESS database has been developed in the Kamloops Forest Region, and a copy can be obtained from the region. Local modifications of that database to suit your needs can be made without great difficulty.
The database will have the following six elements:
Technical staff can obtain data-loggers to enter data in the field. On the data-loggers they would have four worksheets:
From the data-logger, the data can be easily downloaded to EXCEL®, reviewed and edited, and then imported into the ACCESS program. ACCESS files can also be read by ARCVIEW® and other programs to create geographically referenced reports.
The data collected from SPI give a wide variety of information about the interest area. This ranges from information on species composition and stem density, to pest incidence and severity, to tree volumes. The number of trees affected by a given forest health factor gives incidence in percent of trees affected (by species) and number of trees affected per hectare.
The PSR can be used in the analysis phase to show general trends and highlight problem areas or ITGs. For instance, a frequency distribution of PSRs by ITGs could be done. A high frequency of PSR1 in a particular ITG would indicate a high-priority management concern.
The various types of summaries (using both the query and reports functions of ACCESS) and analysis that can be obtained from SPI data are listed below:
Relative frequency of forest health factors (must interpret):
Relative severity of forest health factors(s) recorded:
GIS display of data:
In conclusion, data gathered from the SPI must be interpreted in the context of the management objectives for the area. Sampling at the landscape level will usually not give enough resolution to make a stand-level prescription. As more areas are surveyed and the database grows, the confidence level of interpretation will also increase.
As a part of the FDP, the purpose of the walkthrough is to verify forest health factors identified during operational aerial surveys. A walkthrough would likely be undertaken if the district manager requested a forest health assessment of the area under the FDP. In addition, other damaging agents not visible from the air may be identified during the walkthrough. The scope of this survey differs from other similar walkthrough surveys in that it is conducted at the landscape rather than the stand level.
Landscape-level, generic forest health walkthroughs that may be requested by a district manager are done to the standard set by the district manager. Currently, there is no specific generic walkthrough procedure that is widely used. Survey procedures for specific forest health factors are available in forest health guidebooks. A list of available surveys and the guidebooks in which they are published is presented in Table 1. Standardized procedures have been created in the past, and are available from the Forest Practices Branch.
Walkthroughs or probes related to silviculture prescription approval, particularly for bark beetles, may also be undertaken at the request of a district manager. Procedures associated with stand bark beetle walkthroughs or probes are provided in the Bark Beetle Management Guidebook.
Windthrow is an important abiotic forest health factor (abiotic pest) in numerous timber supply areas, and an assessment of the consequences of windthrow may be needed for both the FDP and the silviculture prescription. Survey information on windthrow is provided in this guidebook because the survey techniques lend themselves to general forest health factor assessments, and there are no other more specific guidebooks appropriate to this topic. The survey information provided in the guidebook can be applied equally well to a landscape or to a cutblock.
There are three basic approaches to windthrow risk evaluation: observational, empirical, and mechanistic.
The FS 712 windthrow field cards (Appendix 6) use the observational approach and therefore indicate relative windthrow risk. The assessment can be made more quantitative by first calibrating along the boundaries of existing cutblocks. This feedback process improves prediction of future outcomes based on past experiences in similar situations. The FS 712 field cards are based on the premise that the management and biophysical (environmental) factors contribute to the wind loading and wind resistance of trees. The biophysical factors can be further subdivided into topographic (wind exposure), soil (strength of anchorage), and stand (acclimation to windloads) components. Each of these components is individually assessed and then integrated into an overall assessment of biophysical hazard that is the inherent susceptibility of the stand in a given location to windthrow. The degree to which management activities will increase wind loading is then considered. For example, windward boundaries on large cutblocks will experience greater wind loads than will parallel or leeward boundaries following harvesting, and therefore have a greater windthrow risk.
Healthy trees are capable of acclimating to routine wind loads. Healthy, open-grown trees are seldom damaged by routine winds, and are typically resistant even to extreme winds. As stand density increases, trees are sheltered by their neighbours, and shed wind loads through inter-tree contact (damping) as they sway. Competition also reduces the resources available for structural increment. It is commonly observed that stand vulnerability to wind increases as stands grow taller. However, veterans, emergents, and large dominant trees within stands are generally more windfirm than other crown classes. Very high density stands of sapling and pole-sized trees are often relatively windfirm if left unthinned because damaged trees hang up in the canopy during windstorms. Prior to thinning high-density stands, consider the importance of inter-tree shelter and damping.
Where soils restrict the depth of anchorage, trees form plate-like supporting root systems. However, plate root systems do not provide a high degree of stability. Deep, well-drained soils provide the best anchorage, and soils with highly fluctuating water tables the worst. The strength of fine-textured soils is lower when the soils are wet. This leads to site-, season-, and weather-related variations in rooting strength. Stem and root decays reduce wind firmness. Rootwads of overturned trees should be examined for evidence of rooting restriction or root decay. Road cuts also provide an opportunity to examine the soil profile and restricting conditions.
Most high wind events in British Columbia are generated by large-scale weather systems such as Pacific low-pressure systems or the movement of arctic air masses. Endemic windthrow is caused by peak winds that recur regularly (e.g., every 13 years), while catastrophic windthrow is caused by infrequent peak winds. The FS 712 cards evaluate risk from endemic winds. Regional winds are accelerated or slowed by local topography. The relationship between wind speed and direction and local topography is complex. Analysis of the orientation of existing windthrow within the forest, or along edges created by harvesting, assists greatly with identifying local wind patterns. The severity of damage along existing windward cutblock edges is a very good indicator of biophysical hazard. Uprooted trees typically create pit-mound pairs. The orientation of pit-mounds can indicate damaging wind directions long after stems have decayed.
The components of biophysical hazard are not necessarily additive. Trees in open-grown stands of healthy trees generally compensate for high wind loads in areas of high topographic exposure by forming short stocky stems or flagged crowns. Similarly, they often compensate for soils that restrict rooting through buttressing or root interlocking. The most susceptible stands are tall, high-density stands growing on high-productivity sites where rooting is partially restricted. Evidence of recent windthrow in such stands prior to harvesting indicates a high degree of instability.
The design of openings or partial cuts can greatly increase the wind loading on stand edges and residual trees. For example, on level ground, windward edges are partially sheltered if openings are less than five tree lengths wide. In uniform thinned stands, wind loads increase in direct proportion to inter-tree spacing. Crown modification techniques can be used to reduce wind loads on residual trees. Removal of 30% of the upper crown mass can reduce wind loads by 50%. Windthrow is a natural disturbance agent, and potential impacts and the level of acceptable damage should be incorporated into prescriptions for damage mitigation.
It is a good practice to maintain a landscape-level map showing windthrow and windthrow salvage locations. Over time this will assist in identifying local wind patterns and those portions of the landscape that are more susceptible to damage. Windthrow can be observed on 1:15 000 scale photographs, and these can be used to examine the edges of cutblocks adjacent to areas proposed for harvest.
Because of the complex interactions of climatic, biophysical, and management factors that contribute to windthrow risk, there is always uncertainty in prediction. The wisest approach is one of prediction, experimentation, observation, and feedback.