 |
 |
 |
 |
 |
 |
 |
 |
||
 |
 |
 |
|||||||
|
Knowledge of population size is essential for the management of animal (and plant) populations. Direct count methods, such as strip counts, aerial counts or vantage point counts, can provide accurate estimates of animal abundance, but such methods are only applicable in more open ground areas, where the objects (animals) can be easily seen. An alternative approach which is likely to reduce any bias and improve the precision of abundance estimates is the use of distance sampling methods. |
|
|
|
|
||||
|
Above: Line transect sampling approach with a single, randomly placed line of length L. Here six objects have been detected and their distances from the transect recorded. |
|
In line transect sampling an observer counts the number n of objects seen while traversing a predetermined line of length L. The perpendicular distance of each object from the transect line is also recorded. When all objects located on the line are detected with certainty, the density of objects in the area surveyed (D) is estimated as (Buckland et al., 2001):
The parameter f(0), estimated by |
||
,
corresponds to the probability density function of the perpendicular
distances, evaluated at zero. f(0) is more readily interpreted as 1/µ,
where µ corresponds to the perpendicular distance from the transect
line within which the number of undetected objects is equal to the number
of objects that were detected beyond it. µ is termed the effective
strip half-width and, when multiplied by 2L gives the effective area
surveyed. Thus estimation of the density of objects can be easily obtained
from estimates of encounter rate (n/L) and f(0). For more information
about line transects, see encyclopedia article |
|
||||
|
Point transects are mostly used in avian surveys. They involve an observer standing at a given position and counting the number of objects seen, (n). This means there is no need to traverse difficult terrain as is required for line transects. Generally, all detections are recorded, regardless of distance from the point. Point counts to a fixed distance w are another form of quadrat sampling, these can be used if it can be assumed that all objects within a distance w of each point are detected. A point transect design consisting of k points, each making up a circular plot of radius w, gives a sampled area of kpw2. If n objects are detected within these circular plots the density, D is estimated using (Buckland et al., 2001):
For more information about point transects, see encyclopedia article Thomas, (2002). |
|
Above: Point transect survey |
||
|
|
||
|
RUWPA has developed distance sampling methods as above since its early days. Staff at RUWPA have expertise in all areas of distance sampling - line and point transects - and are continually developing and improving new research methods and software available. We also provide training for distance sampling users in the form of books and training workshops (both introductory and advanced). An email help-list is also available and maintained by RUWPA staff. Recently, RUWPA has worked on developed new methods for dealing with uncertain detection on the track-line (g(0) estimation problems), as well as new methods used for analysing acoustic and double-platform surveys. Automated survey design is also implemented. |
|
|
|||||
|
|
The 1994 Small Cetacean Abundance in the North Sea (SCANS) project aimed to estimate abundance and identify the main summer concentrations of small cetaceans in the North Sea. The main target species was the harbour porpoise Phocoena phocoena but all cetacean sightings were recorded. The harbour porpoise is a particularly difficult species to survey because of its small size, unobtrusive behaviour and the fact that it may move in response to the survey vessels. In summer 2005 a second survey (SCANS II) was designed and undertaken. Analysis is currently under way. Left above: SCANS
I survey. |
|
|||
|
|
||
|
Cue-counting is related to point transect surveys, although sometimes it is carried out simultaneously with line transect sampling. The method is suitable when multiple discrete (instantaneous) cues are observed during a survey. Estimates of cue-production rate per animal are needed to convert estimates of cue density to animal density. For example, with sperm whales, the cue used is clicks made by the animal - estimates are made of click density using point transect methods, and these can be converted to animal density by using click per animal per unit time. |
|
|
||
|
The advent of Geographic Information Systems (GIS) has facilitated the implementation of automated design algorithms that generate survey designs with known properties rapidly and simply. A number of frequently used survey designs were developed and implemented by Samantha Strindberg within the geographic survey design component of the Distance 4 program. The component can generate surveys based on a range of point- and line-transect designs. It can also perform simulations to compare the efficiency of different designs and to investigate design properties such as probability of coverage. |
|
|
|||
|
Songbird surveys often utilize point transects rather than line transects. It is often difficult to see the birds, due to difficult terrain, etc., and so counts are made of songs rather than sightings. Counts are also made over several minutes, to allow the observer to detect and identify objects. Point transect sampling is used widely for monitoring trends in abundance of song birds. RUWPA have experts in comparing methods and finding the most efficient, unbiased methods for different species. Right: Greenish warbler |
|
||
|
|
|||
|
|
In woodland areas, indirect count methods based on faecal pellet counts are preferred. This requires knowledge of defecation and dung decay rates in order to estimate deer density from the pellet group counts. Line transect methods are used to estimate faecal pellet density, which in turn leads to deer density. We have used line transect surveys of dung to estimate sika (Cervus nippon) and roe (Capreolus capreolus) deer abundance in parts of Scotland, working with the Deer Commission for Scotland and Forestry Authority Scotland. Left: Sika Deer |
||
|
|
||||
|
Above: Bottlenose Dolphin |
Ships and boats are often used for distance sampling surveys of marine mammals. Ships can be used to survey large areas of the ocean that are often inaccessible to land-based craft. Aircraft can also be used to survey large areas for marine mammals. Marine mammals are frequently below the surface, and so the assumption of seeing everything at distance zero is often not satisfied. RUWPA have developed methods of dealing with situations such as this, when assumptions are not satisfied. |
Above: Killer Whale with calf |
||
|
|
|||||
 |
 |
Distance sampling methods such as line transect surveys have been used for surveys of small objects, such as dung-piles or plants. These may only be visible within a metre or two of the transect line, and a common problem is that up to 40% of the counts may be made at zero distance. This makes reliable modelling very difficult, and RUWPA have developed methods for dealing with this, as well as providing training to observers who commonly encounter this problem.
|
|||
|
|
|||||
|
Above: Polar Bear, photo courtesy of Tiago Marques |
Tiago Marques recently took part in a distance sampling survey in Arctic Europe. Experts at RUWPA are needed to help conduct regular counts in order to monitor trends in the polar bear population. This, in turn, tells us about the state of the polar icecaps and potential changes in climate. A BBC report on this sudy is entitled, "Bear study sparks climate fears. A pioneering study by Scottish experts has highlighted the dangers posed to polar bears from global warming."
|
Right: Tiago Marques "The final goal should be that in 100 years or so people can still admire these animals in the wild" |
|||
| | about RUWPA | contacts & staff | research | conferences & workshops | publications | software | links | home | |