surveys and monitoring (lecture 7) Flashcards
Point-in-time surveying:
- why survey?
- look at single site
- new species, management plan, impact assessment - compare multiple sites
- biodiversity, habitat type/condition, status of key species
surveys essential for conservation planning
- need fine scale distribution data
- expensive but cost effective
romeo error
- surveys can rediscover species thought to be extinct
What is monitoring?
- regular repeat surveys to measure change
- used to quantify loss & impacts, identify causes and identify counter strategies
What are monitoring targets?
- same as surveys: species richness, habitat condition etc
- plus additional: demography, phenology etc
Biases in UK surveys:
taxonomic
- breeding birds
- plants
geographic
breeding birds:
- breeding success & population size monitored annually
- distribution only 3 times since 60s
plants:
- breeding success not monitored at all
- population size monitored annually for some rare
species, otherwise haphazard
- distribution monitored twice since 1960s, but
haphazard
there are also geographic biases
Why do surveys and monitoring share core principles?
- monitoring is a repeat survey
- good survey design allows for future repeats
What is accuracy?
- how close estimate is to truth
- inaccurate results are biased
- difficult to assess accuracy but possible to assess if survey methods likely to create bias
- bias usually arises from inappropriate site selection/counting methods
What is precision?
- how close different estimates are to each other
- unrelated to true value
- easy to assess: 95% confidence intervals/standard errors
- precision increases with increased sample size and reduced variation within the sample
What are the core principles of designing surveys/monitoring schemes?
- Define objectives
- realism vs ambition - Define study area
- can be difficult - Choose survey locations
- avoid bias:
- systematic, random, or random stratified sampling - Minimise all other sources of variation
- observer quality: training & clear guidelines
- weather: only suitable conditions
- time of day/season: keep constant
- effort: keep constant
What is:
- systematic sampling?
- random sampling?
- random stratified sampling?
systematic sampling:
- sampling in a prespecified pattern from starting point
- prespecified pattern may match ecological variation therefore creating bias
- not ideal
random sampling:
- equal probability of selection
- can miss rare habitat types
random stratified sampling:
- increases surveying of rare habitat types
- random sampling of sites within categories e.g. habitat type
- decreases sampling where variation is low e.g. where focal species unlikely to occur
Methods: assumptions and common problems
- territory mapping
- multiple visits to site
- record location/activity of all sightings of an individual
- define clusters of sightings
- clusters = territories
- excellent technique for an appropriate species
BUT - time consuming
- doesn’t work for colonial species (e.g. meerkat), polyterritorial species (e.g. wood warbler), species whose habitats can’t be detected
Methods: assumptions and common problems
- mark and recapturing/resighting
- population size estimated from number of marked individuals recaptured/sighted in a subsequent sample
- depends on whether populations are closed or open
- e.g. influence from mortality, immigration, emigration, recruitment
- labour intensive but great for elusive species
- camera traps for large animals
- pit fall traps for small mammals/reptiles
What types of marks are used for mark/recapture surveys?
natural:
- e.g. natural markings on coat, scars etc
artificial:
- temporary e.g. fur clipping, toe clipping
- permanent e.g. pit tags
What method to use when designing mark-recapture in an open population?
- jolly-seber method
- key assumptions:
i) marks last during the sampling period
ii) capture probabilities are constant across
individuals
iii) emigration losses are permanent
Methods: assumptions and common problems
- transects and point counts?
- previous methods are complicated & time
consuming - counting all individuals detected in an area simpler
- transect: route along which focal species occurrence is recorded
- point count: transect with length of 0
What are the core assumptions of transects/point-counts?
- all individuals exactly on the route are
detected - individuals do not move before detection
- individuals are not ‘double-counted’
- individuals are detected independently
- distances measured accurately
