Monitoring Avian Populations in Utah s Riparian Areas

Transcription

1 Monitoring Avian Populations in Utah s Riparian Areas Why monitor riparian birds? Look at results from 10 yrs of monitoring Population trends: linear & non-linear Compare techniques: relative abundance vs. density Frank P. Howe

2 Introduction: In the Southwest, 51% of all the avian species are completely dependent upon riparian vegetation. In Utah, 66-75% of bird species use riparian at some time during the year. In riparian, bird species diversity and abundance is highest of all habitats in the West. During spring migration, riparian areas attract >10x the number of birds found in other habitats

3 Which birds use riparian? A wide variety of spp. from most bird groups E.g., herons, ducks, shorebirds, hawks, perching birds nearly 75% of all species in Utah

4

5 What is it about riparian? All of life s prerequisites in a concentrated area. Cover for nesting, foraging, escape Water many birds can get water from food but many others require free water Food fruits, seeds, leafy veg., flowers, insects

6 Diversity leads to Diversity A wide variety of plants are supported by conditions in riparian These plants in turn provide a wide variety of foods to birds Plants also provide substrates for insects most birds are insectivorous during breeding season

7 Vertical Layers Canopy Subcanopy Ground

8 Canopy (WIWA) Different birds use different layers Subcanopy (SOSP) Ground (FOSP) Combination (BTLH)

9 More on Importance to birds Riparian < 1% of land but up to 75% of bird species use it = disproportionate importance to birds Highest relative abundance of any habitat type Riparian: spp Emergent wetland: ~30 spp Upland forests: spp Pinyon/Juniper: spp Shrublands: 15 Grasslands: 10 spp

10 Riparian Obligates, etc. About 1/2 of the spp you find in riparian areas are obligates i.e., if you remove the riparian those spp are gone Many spp that nest in upland forests elsewhere nest only in riparian habitats at lower elevations At least 60% of riparian birds are neotropical migrants

11 Going down the drain Nationally about 70% of riparian lost or altered in last 150 yrs In arid regions, loss 80-95% Combine this with wetland loss (e.g., 40% UT)

12 Threats Many human activities affect riparian habitats Urban encroachment Agricultural encroachment Water development Overgrazing Road development Timber harvest Pollution Invasion of exotic plants Recreation

13 Impacts lead to Fragmentation of habitat Separation of riparian from adjacent uplands Narrowing of the riparian corridor Lack of regeneration

14 Riparian Landbird Monitoring # # # # # # # # # # METHODS 32 Baseline (49 total) Sites, located randomly 10 Points per Survey Site All birds by sight & sound (ID, flock size, and distance) 8 Minute Counts (3/5/8 min periods) (Variable Radius) Point Transects Each Site visited twice per year Analysis Parameters: Richness, Relative Abundance, Frequency, Density Also collect vegetation data 1 every 3 years Monitor population trends Also monitor trends in richness & vegetation

15 Trend Analyses Linear (parametric) Non-linear (non-parametric) First step: Estimate density using DISTANCE D ˆ = n P a = area under g(x) area under rectangle 1 w ap a w 0 g( x) dx w

16 Linear Trends Linear Trends Mean annual change in density - Used mean density for all years as the basis for comparison For each species Broad-tailed Hummingbird Density (birds/ha) 2 0 Mean

17 Non-linear Trends Annual population changes modeled with General Additive Model (GAM) Incorporated a smoothing function Focus on intermediate time scale Avoid assumption of linear change over 10 years Eliminate ephemeral effects

18 Non-linear Trends Then calculate annual abundance index Index measures change in density relative to reference year () as a function of time Identifies significant inflection points in population trajectory Can test whether inflections are different from zero change

19 American Robin Linear Trend Song Sparrow Results Broad-tailed Hummingbird Over 100,000 detections of 200 spp Examples of 6 common riparian birds American Goldfinch Black-headed Grosbeak Yellow Warbler

20 Linear Population Trends Linear Population Trends Yellow W arbler Density (birds/ha) Mean Broad-tailed Hummingbird Density (birds/ha) Mean American Goldfinch Density (birds/ha) Mean American Robin Density (birds/ha) Mean Black-headed Grosbeak Density (birds/ha) Mean Song Sparrow Density (birds/ha) Mean -2.5% P= % P= % P= % P= % P= % P= % P= % P= % P= % P= % P= % P=0.00

21 Population Trends Population Trends Black-headed Grosbeak Density (birds/ha) Mean

22

23 YWAR YWAR Yellow W arbler Density (birds/ha) Mean

24 BTLH BTLH Broad-tailed Hummingbird Density (birds/ha) Mean

25 Riparian-Dependent Riparian-Independent Big Movers

26 Relative abundance & Density Relative Abundance Point Count Technique Density Fixed radius (25 m, 50 m, unlimited) Point Transect Technique Variable radius Distance Sampling unlimited

27 Counts Relative abundance Output Mean number of birds/point/visit/site/year Transects Density ( D-hat D ) Birds/ha Probability of detection (P a ) The probability that a bird within the survey area is detected Effective Distance Radius (EDR) D ˆ = ap a Radius within which the number of birds missed equals the number of birds detected farther away n

28 Counts Assumptions Detectability remains constant over time (habitat, observer, etc.) Transects Birds on point are detected with certainty Birds are detected at their initial location Distances are accurately recorded

29 Counts Bottom line Assume detectability remains constant Transects Actually estimates detectability Allows us to determine if detectability is constant P a and EDR should remain constant among years for this (Count technique) assumption to be true

30 Methods Similar to linear trend analyses Data from - Trend = mean annual change in statewide Relative Abundance or Density Common scale (compare RA to D) Use percent change in the mean

31 Brown-headed Cowbird Warbling Vireo Count vs. Transect Results Examine whether techniques met their assumptions Examples from 4 riparian species Yellow Warbler Bullock s Oriole

32 Results: Trend Species D-hat RA25 RA50 RA.. BHCO BUOR WAVI YWAR

33 Species BHCO BUOR Results: Detectability & Effective Distance Year Pa EDR Species Year Pa EDR WAVI YWAR Constant detectability? Would not know without distance measurements!

34 Conclusions Both linear & non-linear trend analyses proved useful Linear Familiar, easily interpreted Trend for entire 10-year period Restricted by parametric nature Non-linear Indicate when populations change trajectory Useful in covariate analyses (get at causes) Data Data-driven Complex analyses (no packages) Black-headed Grosbeak Density (birds/ha) Mean

35 Conclusions Detectability is NOT constant Point count (relative abundance) assumption is not met Detectability varies and can be measured with point transects Distance sampling (point transect) largely met assumptions Provided unbiased estimates of density Leading to better trend estimates unlimited

36 Distance Sampling It s a 23 meters More work Collect distances in field Training observers (assumptions) Use of laser rangefinders Model detection probabilities Training in Program DISTANCE But reliable results Robust estimations of density and trend

37 Population vs Project monitoring Population vs Project monitoring Broad-tailed Hummingbird Density (birds/ha) Population monitoring Monitor populations on large scale (, (,, or ) Project Monitoring Assess impacts of a management practice Did management or birds or no effect Same principles of detectability apply Mean

38 Russell E. Norvell Dr. Jimmie R. Parrish Utah Partners in Flight Dedicated crews Numerous reviewers Original results in Auk Asilomar Proceedings Acknowledge DOD 11% BOR 10% BLM 9% USFWS 9% CUP 9% USFS 8% OTHER 5% UDWR 39%

39