Step 4: Select and document supplemental monitoring methods; estimate sample sizes; set sampling frequency; develop implementation rules

Step 4a: Select and document supplemental monitoring methods (if required) 

  • Decide whether‚ÄĮsupplemental indicators‚ÄĮare necessary to meet management and monitoring goals. Keep in mind that adding supplemental indicators will require additional work in the field and beyond (see below). ‚ÄĮ¬†
  • If supplemental indicators are necessary to meet management goals and monitoring objectives, first evaluate the core and contingent methods to determine if these supplemental indicators can be calculated using a core or contingent method.
    • If a necessary indicator cannot be calculated from the core or contingent methods, select a‚ÄĮsupplemental method:¬†
      • Select supplemental methods that are used by other monitoring programs, state regulatory agencies,¬†and are documented clearly in a peer-reviewed method manual.¬†
      • Other desirable characteristics of supplemental indicators and methods¬†include:¬†relevance¬†to Land Health Standards;¬†ability to be¬†measured¬†objectively and consistently in many ecosystems by different observers; scalability; and¬†applicability¬†to multiple objectives.¬†
      • Be sure to document the rationale for including the supplemental indicator as well as a citation for the method. We strongly advise against creating new methods or modifying existing methods.¬†
  • Additional tasks to complete in order to implement supplemental indicators include:¬†
    • Practicing¬†the supplemental method‚ÄĮin the field to establish how it will work with AIM plot¬†or reach¬†layout and requirements (e.g., not walking on left side of a terrestrial transect, for lotic¬†sampling,¬†collecting water quality before instream sampling begins)¬†
    • Identifying‚ÄĮdata management protocol and tools for the supplemental method,¬†including:¬†data recording, electronic data capture, data storage, quality assurance and control, and analysis and reporting.¬†
    • Establishing‚ÄĮcalibration standards‚ÄĮfor the supplemental method.¬†
    • Identifying¬†capacity to provide‚ÄĮtechnical support‚ÄĮfor the supplemental method (e.g., who will answer questions about it during the field season).¬†
    • Planning¬†sufficient training‚ÄĮfor successful implementation of the supplemental method. This cannot be during a core methods training, although we recommend that it follow soon after.¬†
    • Considering¬†the‚ÄĮadditional time required‚ÄĮfor a crew to complete the supplemental method at each sampling¬†location¬†to ensure that the cumulative impact of supplemental methods does not impair the crew‚Äôs ability to visit the desired number of¬†plots or¬†reaches..¬†
    • Using the process outlined in tables 1, 2 and 3, determine how data generated from supplemental methods will be used to‚ÄĮinform management decisions¬†

Step 4b: Estimate sample sizes (Completed by National AIM Team) 

  • The number of monitoring¬†locations needed for a monitoring design is a function of several factors: 1) the amount and quality of existing or legacy¬†monitoring information, 2) the amount of resource that needs to be monitored, 3) statistical considerations, and 4) funding and personnel limitations.¬†
  • If significant amounts of comparable, high quality monitoring data already exist, the required sample size may be smaller than when such data is not available. Make sure to inventory pre-existing monitoring data when you begin to plan your AIM monitoring efforts.¬†
  • If stratification is used, the number of points required in each stratum should balance the proportion of the resource that will be represented by the stratum with the weight of the points (see below for more information).¬†
    • For more information about statistical considerations, see Step 5.¬†
    • A terrestrial field crew with 3 people can monitor approximately 50 plots per season.¬†
    • A lotic field crew with 2-3 people can monitor 25-35 reaches per season.¬†
  • The default method for allocating sample sizes is to‚ÄĮproportionally allocate points‚ÄĮbased on the area/length that the¬†sampling frame or¬†stratum covers.‚ÄĮFor example, if you plan to sample 50 points in a season,¬†chose to stratify,¬†and¬†have¬†a particular stratum¬†that¬†covers 10% of your study area,¬†then¬†you would sample 5 points¬†in that stratum.¬†
  • The recommendation is to start with the proportional allocation approach and then adjust sample sizes up or down as needed. Frequently, the number of sample points will need to be increased in areas that cover a small percentage of the study area in order to achieve a‚ÄĮsample size‚ÄĮsufficient enough to provide information for management decisions.‚ÄĮFor example, black sagebrush areas often occupy a small portion of the landscape but provide important sage grouse habitat, and thus will need to be well represented in a design that is focused on sage grouse.¬†
  • If you increase the desired number of points in one stratum, others may have to be reduced, to keep the total number of points sampled the same.‚ÄĮChanging sample sizes will affect point weights (see below) in each¬†stratum, and¬†should be done with care.¬†
  • Allocating zero points to any strata is not recommended because it will limit your ability to draw inference to the entire¬†landscape, and¬†should not be done unless: 1) the stratum is not part of the target population defined by your monitoring goals and objectives (e.g., open water in a terrestrial monitoring effort) or, 2) the stratum is being monitored as a part of a separate monitoring effort.¬†
  • Point weights‚ÄĮare the area (in acres or hectares) or length (in stream kilometers) represented by an individual sample¬†location.¬†Weights are used¬†to generate statistical estimates of resource status or condition across the landscape (i.e.¬†proportional estimates). Specifically, weights are used to adjust the relative influence that each point has on the final estimates; points with larger weights have more influence, and points with smaller weights have less.‚ÄĮThe weight of each point depends on the¬†specifics¬†of the¬†design,¬†how it was implemented (see final designations),¬†and¬†the reporting area of interest.¬†

Instructions for filling out the remainder of the Sample Design Table: 

When no stratification is used, fill in the first row of the Sample Design Table with information regarding your entire sample frame. 

  • Proportional area or length:‚ÄĮDivide the number of acres or stream km represented by each stratum by the total number of acres or stream kilometers in the entire study area to get proportional areas/lengths.¬†
  • Proportional points per stratum:‚ÄĮCalculate the proportional number of points per stratum by multiplying the proportional number of acres or stream km by the total number of points to be sampled. ‚ÄĮ¬†
  • Final Points per stratum (optional):‚ÄĮIf a proportional allocation of points will not satisfy your monitoring objectives, adjust the number of points that will be monitored for each stratum.‚ÄĮCalculate the number of sites you would like to sample in each stratum, taking the four factors mentioned above into account. ‚ÄĮIn the event that points are allocated in a way that is highly disproportionate to the proportion of the landscape that is represented by a given stratum, the proposed point allocations should be reviewed by someone at the NOC.¬†Final point numbers normally refer to the total¬†number of sampling sites visited within one sampling cycle (e.g.¬†over 5 years). If specifying point number for a different time¬†period¬†this should be specified in the sample design table¬†
  • Point weights:‚ÄĮOnce all of the other columns in the‚ÄĮSample Design Table‚ÄĮhave been finalized, point weights can be calculated as the total number of acres or stream km within the stratum, divided by the number of points to be monitored for that stratum.¬†For assistance¬†in completing this section contact the NOC,¬†particularly¬†for more complex¬†revisit designs.¬†

Step 4c: Define revisit parameters (Use the Revisit Frequency Table to document decisions made in this section) 

  1. Set the revisit frequency and the number of years sampled per cycle-Most monitoring efforts need to be spread out across several years to accommodate field crew capacity and to¬†ensure that interannual variability is captured by the monitoring data. Once the total number of sample points and the point weights have been calculated, determine how many years of sampling¬†might be necessary to achieve the desired sample size.‚ÄĮFactors to consider when setting¬†revisit frequency¬†include:
    • Reducing¬†bias from year-to-year climate variability (e.g., drought)¬†by using¬†a rotating‚ÄĮpanel‚ÄĮdesign (where a certain number of points, all contributing to the same design, are sampled over several years) is typically recommended. Rotating panels help ensure that sample points are randomly distributed across the entire project area every year.¬†
      • For example, a¬†20year design¬†with a¬†5-year revisit frequency¬†would¬†consist of 5¬†revisit¬†panels, where each point is assigned a specific year in which it should be sampled.¬†All points in the same panel will be revisited every 5 years¬†for a total of 4¬†data collection efforts¬†(cycles)¬†at each point over the 20year design.¬†See the¬†Monitoring Design Worksheet example¬†for more details.¬†
      • In contrast, when specific geographic areas are sampled in only 1 or 2 years rather than during every year of the design, bias from climate variability can affect condition estimates.¬†However,¬†it¬†may be appropriate¬†in¬†Lotic¬†sample designs¬†to sample only a proportion of the years in each sampling cycle based on logistical and funding limitations¬†e.g.¬†2 years sampled out of 5.¬†¬†
    • Detecting change in condition through time (i.e., trend) is a common monitoring objective that requires setting an interval for revisiting points over time.¬†Questions¬†to consider when setting¬†revisit frequency¬†include:¬†
      • What¬†revisit frequency¬†makes sense relative to the disturbance or management event? For example, ES&R monitoring dictates annual re-visits for three years, whereas monitoring stream geomorphic changes following livestock removal might occur on a¬†3¬†to¬†5-year¬†basis, and changes in upland condition might occur over 5-10 years.¬†
      • How resistant and/or sensitive to disturbance are the areas that you are monitoring? How resilient are those areas following disturbance events? You may want to consider establishing more frequent revisit intervals in areas that are more sensitive or less resilient to disturbance than in areas that are highly resistant and resilient.¬†
      • How¬†variable¬†and/or sensitive¬†are the indicators¬†you will use to evaluate your management objectives?¬†You may want to¬†consider more frequent revisit intervals for indicators that are particularly sensitive¬†to¬†inter-annual variability¬†in¬†abiotic conditions.¬†¬†
      • What resources will be available (e.g., funding and personnel)?¬†
    • The default revisit interval for Resource Management Plan effectiveness monitoring is every 5 years for terrestrial systems and every 5 years for lotic systems, unless natural conditions or management actions occur that would elicit landscape-scale responses on shorter time-scales.¬†
  2. Set number of cycles and the total duration of your design‚ÄstA cycle is a¬†defined time period over which a group of panels are visited¬†g.¬†5¬†years. The number of cycles in your design depends on both your revisit frequency and¬†total design duration such that¬†numbers of cycles = design duration¬†√∑¬†revisit frequency.¬†¬†
    • Typically for terrestrial revisit designs the standard number of cycles is 4 with a total design duration of 20 years (using a 5-year revisit frequency).¬†
  3. Set the proportion of your design points which will be revisited¬†–¬†Depending on objectives, only a subset of points may need to be revisited. In general, trend assessments¬†are most effective when¬†by revisiting¬†approximately¬†80% of the¬†points sampled¬†within a year or cycle.¬†Factors to consider when¬†determining the proportion of design points which will be revisited¬†include:¬†
    • Revisitation involves resampling existing points¬†and can help to explain changes over time.¬†The higher the proportion of revisit points, the more statistical power you have to detect trend.¬†¬†
    • Non-revisit points add new sampling locations across the landscape and help to explain spatial variability in resources.¬†The higher the proportion of non-revisit points the higher the precision¬†of¬†condition¬†estimates.¬†
    • What are your management goals?¬†If trend assessment is a priority and existing trend data is unavailable a higher proportion of revisits¬†will be beneficial. Conversely, if management goals are more focused on¬†precise condition assessment at a single point in time, a higher number of non-revisits points¬†will be¬†preferred.¬†
    • In general, a good balance between trend and status estimates is reached using 80% revisits and 20% non-revisits each year.¬†
    • Some monitoring efforts will not need to determine sampling frequencies on account of various project constraints or intentional design.¬†

Step 4d: Develop implementation rules 

  • Review standard AIM implementation rules, including rejection criteria, on aim.landscapetoolbox.org under‚ÄĮData Collection.¬†
    • Proper design implementation involves documenting the fate of each point in a given design. Documentation of point fate should be tracked using the¬†Terrestrial Plot Status populated in¬†Collector Map, and the Lotic¬†Office Evaluation¬†WebMap.¬†For more information and to download these tools and their instructions, visit the‚ÄĮPoint Evaluation and Rejection page.¬†
  • If the implementation rules need to be customized to meet you monitoring objectives, consult with the NOC when developing the additional criteria to ensure the design will remain statistically valid.¬†For terrestrial design implementation consult the¬†TerrADat¬†Ingestion Tree¬†to¬†review the minimum requirements for¬†Terradat¬†data ingestion.¬†¬†

  

Step 4 Example: Supplemental Monitoring, Sample Size, Monitoring Frequency, and Implementation Rules

Read more...

Step 4a: Select and document supplemental monitoring methods (if required)

Terrestrial –¬†Supplemental methods – Terrestrial monitoring indicators and methods can be found in BLM Tech Note 440 and the Monitoring Manual for Grassland, Shrubland and Savanna Ecosystems (MacKinnon et al. 2011; Herrick et al. 2015). ¬†The core terrestrial indicators are sufficient for evaluating most of the terrestrial management and monitoring objectives, as discussed in Steps 1 and 2. ¬†However, several supplemental indicators were identified that were not addressed by the core indicators: ¬†sagebrush shape, distance to nearest sagebrush patch and distance to nearest juniper/tall structure. ¬†These supplemental indicators inform sage grouse habitat questions as part of the Sage Grouse Habitat Assessment Framework (Stiver et al. 2015). ¬†Distances to nearest sagebrush patch and distance to nearest juniper/tall structure can be measured using GIS with minimal additional field time, especially with the help of good notes taken by the field crew (Stiver et al. 2015). ¬†The Field Office GIS staff will capture this information in an Excel spreadsheet following the field season. ¬†A standard method for describing sagebrush shape, consistent with the HAF is available in the National Resource Inventory (NRI) protocol (National Resources Inventory 2016). ¬†This information will be recorded every time a sagebrush plant is hit while doing Line Point Intercept, and electronically captured using DIMA. ¬†Supplemental training for field crews and field office staff will be made available to ensure that these methods are implemented successfully. ¬†¬†

Lotic – The lotic core indicators will be sufficient for evaluating most of the lotic management goals and monitoring objectives. However, the Utah Standards of Rangeland Health require that the BLM assess whether streams are meeting State water quality standards for e. coli. Since e. coli exceedances are only likely to occur in streams that are impacted by grazing, e. coli samples will only be collected at streams that are heavily used by cattle. E. coli samples will be submitted to the Utah Department of Environmental Quality, Division of Water Resources for analysis. The Aquatic Ecologist in the West Desert District will train the field crew on e. coli sample collection methods.

Step 4b: Estimate sample sizes

Terrestrial – Sample sizes were determined for each stratum based on field crew capacity and the final proportion of acres represented by each stratum (Sample Design Table).

Sample points per BpS unit will be proportional to BpS area on the landscape except where monitoring objectives suggest alteration (Figure 3; Sample Design Table; see Steps 4 and 5).  Within the broader District, a decision was made to intensify monitoring within sagebrush-dominated strata (mountain big sagebrush and Wyoming big sagebrush) to address sage grouse habitat management objectives. Sampling intensity was reduced in the Shadscale-Winterfat strata.

Lotic Р To balance personnel capacity, statistical power, budget etc., we worked with NAMC/NOC to select a sample size of 50 stream reaches for the district. We anticipate that a sample size of 50 will allow us to estimate the proportion of stream km in a given condition category with 80% confidence. Supplemental points can be added to increase the precision and accuracy of estimates as needed.

There were no river km on BLM managed lands in the West Desert District, therefore we will only monitor small and large streams. Proportional allocation of points to small and large streams would have resulted in fewer than 10 points on large streams which seemed like an insufficient sample size, so we increased the point allocations in the large stream category to obtain a larger sample size and decreased the number of points allocated to small streams to keep the total number of sample points to the desired number of 50 points for the district.

Step 4c: Set sampling frequency

Terrestrial РOne field crew can collect approximately 50 points in a monitoring season, so we intend to sample 100 plots per field season, for 5 years. Thus  we are planning to hire and run two three-person field crews to collect terrestrial monitoring data for each of the 5 years.

After the initial 5 year effort is complete. We plan to revisit 75% of terrestrial points on a five-year rotating basis in order to estimate trend.  

Lotic – In our area, one lotic field crew can collect data at approximately 25-30 reaches in a year. Thus to accomplish the desired sample size of 50 points, we plan to hire one, 2-person crew for the next two years.

The objectives for trend monitoring will be determined after baseline conditions are established. Specifically, follow up monitoring will be focused on any indicators that raise red flags. The temporal scale that we will use to implement subsequent monitoring will depend on which indicators need to be monitored and the temporal scale that we expect them to change in response to natural environmental variability and/or management actions. For example, recommendations might be to assess water quality indicators on a monthly basis if exceedances are observed. In contrast, bank stability would be assessed on an annual or semi-annual time-scale.   

Step 4d: Develop implementation rules

Terrestrial and Lotic – We will use the standard AIM implementation rules, the Terrestrial Plot Tracking Tool, and the Lotic Design Management Spreadsheet to implement the designs and to document the fate of all design points (see aim.landscapetoolbox.org).

Additional implementation rules are that the supplemental indicators will be collected at each terrestrial plot (see step 7).

Helpful Documents and Links

Monitoring Design Worksheet

Monitoring Design Worksheet Instructions

Monitoring Manual for Grassland, Shrubland, and Savanna Ecosystems

Shiny Tool for calculating sample size

Java applets for power and sample size

BLM Tech Note 440

Sage Grouse Habitat Assessment Framework

National Resources Inventory 2016 РSagebrush shape protocol 

2020 Terrestrial Data Management Protocol

Comments are closed.