 |
 |
 |
 |
 |
 |
 |
 |
||
 |
 |
 |
|||||||
Below are some examples of current and recent RUWPA research relating to monitoring and trend |
|
|
|
|
|
||||
|
|
Estimates of density or abundance
from platforms of opportunity (PoPs) are often not reliable for a variety
of reasons. These include the facts that PoP data seldom come from well-defined,
randomized designs and they often lack the level of data quality control
found on dedicated surveys.
Nevertheless, PoP data are potentially useful
for estimation of trend and distribution, partly because they sometimes
come in such large volume and can have wider temporal coverage than
dedicated surveys. Spatial
modelling methods and spatio-temporal modelling methods can be
used to take advantage of the greater spatial and temporal coverage
offered by PoPs (provided data quality is adequate). Left: The figuret shows transects (black lines)
and harbour porpoise detections (green circles) from the Joint
Nature Conservation Committee, "Seabirds at Sea" surveys
off the coast of Britain. The red to orange coloring indexes relative
density. |
|||
|
|
||||
|
Above: Counts from a control (black) and study (blue) locations over days spanning an impact event (at the horizontal dashed line). |
Designs for assessing impact of event(s) on wildlife populations often suffer from the severe limitation that adequate replication in space and time is impossible. The development of Before-After-Control-Impact and other design analysis methods which are more intensively model-based than conventional analysis methods is an area of active research interest. Both Spatial Modelling methods and Integrated Population Dynamics methods are useful here.
Right: Simulated control sites (blue) and an impact site (red) for assessing the power to detect impact on a population (dots). Color represents temperature and dashed lines are contours - both of which affect expected counts. |
|
||
|
|
||||
|
|
Population trajectories can be estimated purely empirically, by linking closed-population estimates using some appropriate smooth. They can also be estimated using a model for the underlying population dynamics. Left: Empirical smooths to a count dataset.
|
|
||
| | about RUWPA | contacts & staff | research | conferences & workshops | publications | software | links | home | |