Herbivore Census
Techniques
How to Skills Session - I
Wildlife Census
● Need information on the numbers of animals for research, management and
conservation
● Information can be: total numbers, size and structure and distribution and
movements
● How do we get such information then?
● 1st point to clearly know what the objectives of the census are.
Other factors to consider:
● resources
● size of area
● vegetation
● relief
● target species
● Decision making on method should include the following aspects:
Ground based census techniques
● Census team travel along a predetermined transect
● Stratification – transects to cover various habitats in study area
● Driving speed is important (30 km/hr)
● Animals cited are recorded by spp. and group size, structure
● Sighting distance & sighting angle in relation to travel direction
Road counts: what the method entails
● Transect are traversed on foot, horseback, bike or vehicle
● Transect is searched, each animal seen provides one
measurement of perpendicular distance to the transect
● Animals are not seen often along the transect
Distance sampling techniques
● Estimating populations based
on line transects-ground based
technique
● Observer moves along a
transect, dots are individual
animals
● Arrows are distances from the
transect
● Most animals are not detected
Representation of field recordings
● Record the following
1. Sighting distance (ri)
2. Sighting angle (θi)
3. Perpendicular distance (xi)
Animal sighting
Transect
Detectability
● In practice animals are not
always detected
● Weakness of a fixed strip
width
● Under-estimation occurs
● Detectability will fall off with
distance from the transect
● Shaded area shows the
detection function for wildlife
populations
Basic assumptions
1. Animal directly on the transect will never be missed (detection probability = 1)
2. Animals are fixed at the initial sighting position- they do not move before being detected,
none are count twice
3. Distances and angles are measured exactly – no measurement error, rounding errors
4. Sighting of individual animals are independent events
Density calculation
If the assumptions are met, then simple density estimation is :
density =
Where n = Number of animals seen on transect
L = Total length of transect
a = Half the effective strip width (a constant)
The problem is estimating the transect width
✔ Constant- total area under detection function
Herbivore Census
Techniques
How to Skills Session- II
Practical – Distance software
● Distance software - GUI
● Create a project
● Import data
● Run project
● Extract results
Distance Software
● Download distance software: https://distancesampling.org/
● Here we using a graphic user interface (GUI), for more details, take
time to explore the links below:
○ https://synergy.st-andrews.ac.uk/ds-manda/
○ https://www.fws.gov/midwest/endangered/insects/kbb/pdf/KBBDist
anceSamplingGuide11June2008.pdf
Arranging data in excel
● There is one column for each of:
o stratum name,
o stratum area,
o transect name,
o transect length,
o distance (perpendicular distance () and
o cluster size
Extracting & Reporting
● For an example of how such results are extracted & used, see
the link below:
○ https://drive.google.com/file/d/1t7ZabGbWGIMKQfZrrb0yKmKu8zFkW
pA6/view?usp=sharing
○ https://drive.google.com/file/d/1I2IJDrtdOBBghdr5o5loZc7I9hu3XATc/vi
ew?usp=sharing
○
Common pitfalls
● Warning: Size bias adjustment has increased expected cluster size
○ large clusters are more likely to be seen at large distances than small
clusters;
○ the average of detected clusters is likely larger than the average size of
clusters in the population “Size bias problem”
○ Meaning: adjusted estimate is larger that the observed average cluster size
(estimate of the “expected cluster size”)
Reporting Format
Probability of detection
The relative frequency that
observed groups were within the
stated distances from the transect
line (histograms) and the fitted
detection functions (bold lines) for
the nine commonest species of
large mammal
Data Representation
Alternative Reporting/Representation
