BIS 219 Conservation principles

and realities
Survey and monitoring
 BIS 219 Conservation principles
and realities
Survey and monitoring

Learning objectives:
describe reasons for conducting surveys and monitoring – and their differences

Apply key principles of site selection

Select appropriate methods for a given situation

 SURVEY - one time point
Why survey?
1. Single site: new species, management plan &
impact assessment
10
9
# new bird species

8
7
6
5
4
3
2
1
0
2000 2014
 SURVEY - one time point
Why survey?

2. Compare multiple sites:

• biodiversity
• habitat type/condition
• status of key species
Essential for conservation planning - need
fine scale distribution data
 SURVEYS ARE EXPENSIVE….
15,000km2 Ugandan
forest
woody plants, large
moths, butterflies, birds
& small mammals
100 person years
1million $US
 …BUT COST-EFFECTIVE
species in reserve network
Cumulative # important

No data on local species

distributions, i.e. random
selection of sites

Cumulative reserve area

 …BUT COST-EFFECTIVE
With complementarity
species in reserve network
Cumulative # important

Complementarity
selection of additional
sites with reference to
which species already
Random site selection occur in selected sites

Cumulative reserve area

 …BUT COST-EFFECTIVE
species in reserve network

Target #
Cumulative # important

species

a = extra
land needed
a

Cumulative reserve area

Balmford & Gaston 1999 Nature 398, 204-205
 …BUT COST-EFFECTIVE
species in reserve network

Target #
Cumulative # important

species
If it costs less to do
surveys than to
purchase and
manage land of
a = extra area A then surveys
land needed are cost effective
a

Cumulative reserve area

Balmford & Gaston 1999 Nature 398, 204-205
 The Romeo error – rediscovering
extinct species

Cebu flower-pecker
1906 last seen
1959 declared
extinct
1992 re-
discovered
 MONITORING
Regular repeat surveys to measure change

Quantify loss & impacts Monitoring

Identify causal mechanisms

Devise strategies to reduce impacts

 MONITORING
Monitoring targets:
1. Same as surveys (species
richness, habitat condition
etc.)

2. Additional reasons, e.g.

phenology and demography
 TAXONOMIC BIASES IN
MONITORING
UK BREEDING BIRDS
success - annual
pop size - annual
distribution - 3 times since 1960s

UK PLANTS
breeding success - none
pop size – annual for some rare
species, otherwise haphazard
distribution - twice since 1960s, but
haphazard
GEOGRAPHIC BIASES IN
MONITORING




 




Avian population monitoring schemes

SURVEY & MONITORING DESIGN

Core principles identical because:

i) monitoring is a repeat survey

ii) if conducting a survey best practice to
use a design that can be repeated in the
future
 KEY TERMS: accuracy
How close the estimate is to the true value;
inaccurate results are biased.

Very difficult to assess directly, but can

assess if survey methods are likely to create
bias
 KEY TERMS: accuracy
How close the estimate is to the true value;
inaccurate results are biased.

Very difficult to assess directly, but can

assess if survey methods are likely to create
bias

Bias largely from inappropriate site selection

or inappropriate counting methods
 KEY TERMS: precision
How close different estimates are to each
other, unrelated to true value.

Easy to assess, e.g. 95% confidence

intervals or standard errors.

Precision increases with:

i) sample size
ii) reduced variation in the sample
 Low
High variance across samples
Precision

High

Low variance across samples

Low

Sample size
 CORE PRINCIPLES
1.Define objectives - realism versus ambition
 CORE PRINCIPLES
1. Define objectives
2. Define study area – not always that easy

Sheffield city
council boundary
 CORE PRINCIPLES
1. Define objectives
2. Define study area – not always that easy

Use land cover to define

urban

Spatial grain size crucial

 CORE PRINCIPLES
3. Choose survey locations

Must avoid bias

Three main methods:

i) systematic
ii) random
iii) random stratification
 SYSTEMATIC SAMPLING
Sampling in a pre-specified pattern from a
random starting point
X X X

But, pre-specific pattern may match pattern

of ecological variation, creating bias
Not ideal
 RANDOM SAMPLING
Equal probability
of selection

1 2 3 4 5 

6 7 8 9 10  

11 12 13 14 15 

16 17 18 19 20 
 RANDOM SAMPLING

 

Random sampling can miss rare habitat types

RANDOM STRATIFIED SAMPLING
1 1 2 2 3
Use to increase sampling
of rare habitats
3 4 5 6 4 Randomly select sites, but
1 within, i.e. stratified by,
habitat type
 1

Also decrease sampling where variation will be

low, for example where focal species is unlikely to
occur
 CORE PRINCIPLES
1. Define objectives
2. Define study area – not always that easy
3. Choose survey method
4. Minimise all other sources of variation
observer quality – training & clear guidelines
weather – only suitable conditions
time of day/season – keep constant
effort – keep constant
 METHODS :
assumptions & common problems
 TERRITORY MAPPING
• Multiple visits
• Record location & * *
activity of all sightings * * *
* *
• Define clusters of
sightings
• No. clusters = no.
territories * ** *
*
* *

simultaneous sighting
same individual
 TERRITORY MAPPING
• Multiple visits
• Record location & * *
* * *
activity of all sightings * *
• Define clusters of
sightings
• No. clusters = no.
territories * ** *
**
* *

simultaneous sighting
same individual
 TERRITORY MAPPING
Assumptions not always valid
Colonial Poly-territorial
species species

Can’t detect territories But if focal species chosen well

excellent technique

But time consuming

 MARK & RECAPTURE/
RE-SIGHTING
Estimate population size from % of marked
individuals recaptured/resighted in a
subsequent sample

Are populations closed or open, i.e. influenced

by mortality, recruitment, immigration &
emmigration
 MARK & RECAPTURE/
RE-SIGHTING
If completely open population use Jolly-
Seber method, key assumptions are:
i) marks last during the sampling period
ii) capture probabilities are constant across
individuals
iii) emigration losses are permanent
MARK & RECAPTURE/RE-SIGHTING
Types of marks- natural
MARK & RECAPTURE/RE-SIGHTING
Types of marks – artificial
Temporary, e.g.
fur/toe clipping

Permanent
mark e.g. pit tag
MARK & RECAPTURE/RE-SIGHTING
Labour intensive, but great for elusive species

Camera traps: Pit fall traps: reptiles, small

nocturnal mammals mammals etc
 TRANSECTS & POINT COUNTS
Previous methods are complicated & time
consuming

Idea of counting all individuals detected in

an area has the appeal of simplicity

x
A point count is
a transect of
Transect
zero length
 TRANSECTS & POINT COUNTS:
CORE ASSUMPTIONS
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
TRANSECTS OR POINT COUNTS?
More individuals counted per unit time with
transects as recorded whilst moving, so
are preferable unless:

i)habitats are fragmented

TRANSECTS OR POINT COUNTS?
More individuals counted per unit time with
transects as recorded whilst moving, so
are preferable unless:

ii) access difficulties

create route diversions or
makes it difficult to move
and count simultaneously
 TRANSECTS & POINT COUNTS:
Can we just count?

Yes to calculate an index

of population trends

But no if we want to:

i) compare densities
between species/habitats
ii) calculate population size
 TRANSECTS & POINT COUNTS:
DISTANCE SAMPLING
Because detectability:
i) declines with distance from transect
ii) varies between species & habitats

100%
Conspicuous spp
probability

or open habitats
Detection

Elusive spp or
dense habitats

Distance from transect

 TRANSECTS & POINT COUNTS:
Distance sampling corrects for variation in
detectability by estimating number of
undetected individuals
Allows estimation
of absolute density
probability
Detection

estimates

Distance from transect

 BOTANICAL SURVEYS
All general design principles apply to plants
Territory mapping & mark/recapture
irrelevant
Can use DISTANCE sampling, e.g. cacti,
although need to take plant size into account
 BOTANICAL SURVEYS:
small plants
Quadrats: most common
method
Optimum size is a trade-off
between time taken per
quadrat & sample size
Usually record % cover not
number of individuals
 SUMMARY
Surveys and monitoring are essential &
often cost effective

The most important core design principles

are to minimise bias & have clear objectives

Numerous valuable methods are available,

but selection needs careful consideration