Long-term monitoring of large, remote areas with minimal funding: hope and encouragement for natural area managers Steven Fancy National Monitoring Program.

Long-term monitoring of large, remote areas with minimal funding: hope and

encouragement for natural area managers

Steven FancyNational Monitoring Program Leader

National Park Service

State of the Parks Report (circa 2000)

• 80 (1/3) of the “natural resource parks” had no professional natural resource manager

• Another 84 parks had only 1 or 2 natural resource professionals.

• Almost all projects/studies were short-term; staff mostly deals with the “crisis of the day”.

• Sampling designs mostly done by expert opinion “representative sites” or convenience.

No Time No Money No Clue

NPS Natural Resource Challenge

• Accelerate Inventories

• Design/Implement Vital Signs Monitoring

• Collaboration with scientists and others

• Improve Resource Planning

• Enhance Parks for Science

• Assure Fully Professional Staff

• Control Non-native Species

• Protect Native and Endangered Species

• Enhance Environmental Stewardship

• Expand Air Quality efforts

• Protect and restore Water Resources

• Establish Research Learning Centers

Revitalize and expand the natural resource program within the NPS & improve park management through greater reliance on scientific knowledge

Economics 101

• Without integration and cost-sharing, parks can only monitor a few things; too few to adequately track condition of resources;

• Park buy-in and cost-leveraging through partnerships are critical; must be relevant to park managers and flexible to allow integration and partnerships;

• Establish 32 “monitoring networks” that share funding and staffing among parks to gain efficiencies and consistency.

Funding level (avg. $100,000 per park) would allow each park to hire one professional position (GS-9 or 11)

plus about $30-40 K operating $$

Conclusions/Strategy:

The Wedding CakeAn alternative to “One Size Fits All”

National

Servicewide Core Variables

Network/Ecosystem Core Variables

Network/Ecosystem

• Primary use of data is at local level for park management

• Indicators and protocols most relevant to each system are very different; e.g., Northwest Forests vs. South Florida vs. Colorado Plateau

• Cost-leveraging through “Opportunity” very important for early success and efficiency

Spatial Sampling DesignsWhy do we Need Them?

We can only sample a very small proportion (often <1%) of most natural areas; but,

Our job is to protect, restore, understand, and inform others about the entire area, not just some convenient portion of it.

Need to make scientifically-defensible inferences to areas beyond the actual areas we sample.

Statistical, design based inferences can only be made to areas that have a chance of being included in the sample.

Especially important because we are designing LONG-TERM MONITORING

Reality Check: Sometimes you just can’t get there

Take Home MessagesA few Lessons Learned about Sampling Design for Long-term Monitoring

• Design should allow inferences to entire park, not just some accessible portion of it

• Some sort of probability sample should always be taken, even if you can only start with very small sample sizes

• Judgement sampling, using “representative” sites selected by experts, should be avoided

• Do not stratify on biological criteria (e.g., vegetation maps). Use stratification sparingly, if at all

• Permanent plots that are revisited over time are recommended to increase precision

• It is not necessary to visit plots every year - rotating panels increase sample size

• Co-locate sample sites for various components of the monitoring program

Three Examples:

• 2-stage “mini-grid” design for Central Alaska Network – 3 parks >21 Million acres (size of Indiana)

• GRTS design – large river parks in Missouri and Arkansas

• Unequal probability sampling of upland sites in Northern Colorado Plateau parks (site selection based on accessibility, climate, geology and soils)

First-stage systematic grid with 20 km First-stage systematic grid with 20 km spacingspacing

Second-stage “mini-grid” sample (25 points)

# # # # #

1 0 1 2 Kilometers

Buffalo River Streams: 100 sites

•Geomorphology•Aquatic Macroinvertebrates•Fish Communities•Ozark Hellbender (giant salamander, T&E)•Mercury•Lead

Sampling of stream segments with co-location and co-visitation among monitoring indicators:

Northern Colorado Plateau Network (NCPN) Sampling Design

• Collocated Vital Signs – vegetative structure, composition, condition of focal communities; soil/site stability; upland-hydrologic function; soil-crust structure; nutrient cycling

Integrated Upland Sampling Design - NCPN

• Systematic grid of points (50-m spacing) overlaid on park used to define anchor points of sampling plots

• Unequal probability of selection of sampling plots• provides ability to add sites in the future (if needed)• probability defined by accessibility (high, low)• probability defined by climate, soils, and geology

(rel. static surrogate for vegetative communities)

• A split-panel design will be employed that features• spatial interpenetration• the ability to evaluate inter-annual variability• a balance between trend and status assessments

And that meets Vital Sign sampling objectives

Integrated Upland Sampling Design – NCPN

Temporal Design

Split panel [(2-7)9, (1-8)9]

X X X 1

10…987654321Year/Panel

X = 10 plots; 30 plots per year spatially distributed ca. equally among 4 Ecosites

Weight of Evidence

http://science.nature.nps.gov/im/monitor