All six species of wild deer in Australia have been living in free ranging herds for over a century. For most of that time numbers were low and distribution was limited to a handful of areas. Most Australians, including hunters, had little idea that there were wild deer in this country. That situation has changed dramatically over the past couple of decades.
The distribution of deer throughout the country has increased significantly, particularly in the eastern states. Deer have become a topical subject for many people, hunters and non-hunters alike. Wild deer are now on the natural resource management agenda at national, state and local government level. With an increasing awareness comes the desire to know more about these animals, particularly ‘how many deer are there?’ This question can be answered by a wildlife survey, a tool widely used by natural resource managers.
However to fully appreciate the results of a wildlife survey it is important to have some basic understanding of what they are and how they work. This article is the first of four which aim to provide deer hunters with this knowledge.
Regardless of the species of interest, the size of area or the reason to know, there are some basic principles that apply to all wildlife surveys. These principles apply equally for whitetailed deer in North America, red deer in Europe or any species of deer in Australia. This article will explain these basic principles starting with the difference between an index and an estimate and then an explanation of data collection and sampling.
Not all Surveys are Created Equal
There are many kinds of surveys used by wildlife professionals, and even more factors influencing the choice of survey used. One of the most important concepts to understand is that surveys may be used to find either an Index of Relative Abundance or an Estimate of Absolute Abundance. Because both of these surveys are widely used confusion can occur for someone trying to understand or use survey results if they don’t have a good understanding of the difference. An index quantifies abundance without actually stating how many deer there are whilst an estimate of abundance will state how many deer there.
Index of Relative Abundance
An index, in survey terms, is a method for quantifying a population without estimating the actual number of animals. The important element of data collected for an index is that it must increase or decrease in relation to the actual density of deer. Deer particularly can be very cryptic and the habitat they live in can be very dense, therefore only spoor can be quantified such as tracks or scats. The principle of an index relies on the assumption that more deer equals more spoor and less deer equals less spoor.
For many management applications an estimate of actual numbers is not required and an index of relative abundance is sufficient. An index can be used to compare different areas at the same point in time or to track changes in the same area at different points in time. This ability makes indexes useful for tracking the outcome of management actions. For instance the question may be asked ‘are sambar increasing in number in the XYZ National Park’? This fictitious park could be in Thailand where efforts have been made by local authorities to reduce the amount of bush meat hunting by local villagers. Alternatively the park could be in Victoria where the access to recreational hunters has been closed due to perceived safety concerns. In both instances there has been a management action taken that is predicted to affect the number of sambar present. The habitat is dense and the deer hard to observe so the choice of survey methods is limited. However a survey design using an appropriate Index of Relative Abundance will answer the original question.
Estimate of Absolute Abundance
When it is possible to actually see deer or quantify them as individuals in some way then it is usually possible to estimate the actual density or number of animals present in a defined area. Surveys designed to achieve this require that animals are not counted accidentally more than once and that individuals don’t arrive or leave during the survey period through immigration, emigration, births or deaths. This means the survey will be conducted within a clearly defined timeframe.
Imagine a Scottish estate where sheep are grazed in the same habitat as free ranging red deer. There are no predators present so the numbers of deer will increase without harvesting.
The owners of the estate like hunting the deer but want to make sure the harvest doesn’t reduce the population size. To know how many deer can be safely harvested in any given year an accurate estimate of total abundance is required. An index of relative abundance would not answer this question. However there are many survey types that will provide data sufficient to estimate the actual abundance of red deer on the estate. When the population size is known a sustainable harvest can be calculated.
Some of the more common survey types used by wildlife managers.
Examples of relative abundance indexes
Examples of estimates of absolute abundance
Passive soil plots
Strip transect
Faecal pellet index
Line transect (Distance sampling)
Catch per unit effort
Mark recapture survey
Spotlight counts
Index-manipulation index
Remote camera data
Change of ratio method
Walked line counts
Known to be alive
Data Collection
There a many ways data can be collected when surveying wild deer. Observations can be made whilst driving, riding a horse, flying or walking at night or in the daylight. Some surveys are conducted from a fixed location whilst others collect data without people present (for example, by camera trap surveys). Data collected from any of these methods can be utilised in different ways as either an Index of Relative Abundance or an Estimate of Absolute Abundance. You may hear of someone doing ‘aerial surveys’. This description of the survey only describes the platform the data was collected from, not the type of survey method used.
The collection of some field data is quite specific and will detail not only the number of deer but the sex and age category of those seen as well. When using camera traps to estimate the actual abundance of deer present the data is used to identify individual stags. This level of detail is unusual with most surveys counting deer or spoor only. It is important when designing surveys to clearly define the units of measure and how they will be used. For example there is no point collecting detailed data on colour forms of fallow deer if the final survey result only reports on all deer present regardless of colour.
Sampling
In an ideal world a wildlife manager would be able to answer the question ‘how many deer are there in this location’, with the absolute answer ‘there are x number of deer’. Unfortunately there are very few instances where this is possible. Deer present on a reasonably sized island or number of deer within a fully fenced area would be a couple of examples where this might be attempted with some success. When it is possible to count all animals present the survey is known as a census. However in the majority of cases surveys rely on sampling to find the number of deer within a subset of a population. This subset is then used to extrapolate to the whole population. The accuracy of the answer attained describes how close the estimate is to the actual number of deer present. The precision of the answer is a description of how repeatable the result would be if the survey was replicated. The best surveys are both accurate and precise however most are a compromise due to the many limitations a wildlife professional must work within. High precision is usually the attainable goal most survey designs aim for. A survey design with good precision can effectively track changes in animal density over time.
Transects and Quadrats
Once a species and area has been defined a sampling regime will be decided upon to collect field data. This may be collected from transects through the area or quadrats within the area. Transects are lines that run through a survey area that data is collected from. A transect for instance could be a fixed wing aircraft flying from one side of an area to another with observers recording deer seen along the way. A quadrat could be a 10 metre by 10 metre section of pasture where all deer droppings are removed at time one and all new droppings counted at time two. The selection of where transects or quadrats are positioned and how many there are is critical to the effectiveness of the survey.
If the defined area to be surveyed has different habitat types then a good survey design would stratify the area. Each strata represents one habitat type. Within each strata transects or quadrats would be located to ensure effective coverage for all of the habitat types. The survey design may incorporate more transects (that is, greater sampling effort) within some strata to achieve greater accuracy if required.
Transects ideally would traverse the grain of the habitat and not run along the grain. For instance a transect should travel across a creek line and continue through a forest or travel up a hill not contour around the hill. Following the creek line or contouring around the hill would not represent the full habitat and would bias the sampling. There are many surveys that utilise vehicle tracks through an area due to the convenience of access. These transects are not ideal and will be biased due to propensity for tracks to follow contours and creek lines. However the data collected can still be very useful so long as the limitations are recognised.
ADA Stalwart Neil DeCoite measuring a transect
Bias in Surveys
All surveys have inherent bias within the design and application. Bias can take many forms based on the habitat, weather, sampling effort, observer ability and survey method. A spotlight survey through a mixed habitat of open pastures and forest will be biased towards the pastures because animals are harder to see with a spotlight in a forest. Two observers may be seated either side of an aircraft counting deer below. One counter may have 20/20 vision and the other blind in one eye, there is clearly a bias present here.
As previously mentioned a survey along a vehicle track will be biased in most cases. A good survey design identifies all possible bias that may be present in the survey and attempts to reduce this by stratification and standardising effort. Accepting there is bias means management decisions can be made accordingly. For instance many survey results will underestimate the number of animals due to inherent limitations of observing wild animals. These surveys are biased towards a conservative estimate or index.
These notes have described some basic principles that all wildlife surveys are based on. To reinforce the main point of this article: an index quantifies abundance without actually stating how many deer there are whilst an estimate of abundance will state how many deer there.