Annual Whooping Crane Survey


Wintering Whooping Crane Abundance Survey.  

Introduction and Justification
The Aransas-Wood Buffalo population of whooping cranes (Grus americana) is the last wild, migratory flock of this species. This wild population migrates between Canada’s Wood Buffalo National Park, where they nest, and the Aransas National Wildlife Refuge in Texas, where they spend the winter. In 1941, the population was reduced to 15 individual birds. It was at this time that people decided to intervene and whooping crane recovery efforts began. Today, biologists estimate there are more than 250 birds in the wild flock.

It is important that a self-sustaining population of whooping cranes be established, one that doesn’t require any human help. Biologists believe the Aransas-Wood Buffalo population must have at least 1,000 individual birds and 250 reproductive pairs before the species can be considered for down-listing under the Endangered Species Act.

Previous Methods (1950-2011)
For more than 60 years, the Aransas-Wood Buffalo population of whooping cranes has been counted in the same way. In 1950, when whooping crane aerial “census” efforts began, all of the whooping cranes’ known territories were located on the Blackjack Peninsula of the Aransas National Wildlife Refuge. The survey was coined a “census” as it was assumed that every bird in the population was seen and counted.

Surveys were done with one observer and the pilot and groups of cranes were marked on paper maps. The population size reported each year was based on the observer’s years of experience and knowledge of the individual whooping cranes and the territories they occupied each winter. If during a survey a bird was not seen, the observer could account for the missed birds in the next survey or by going in search of that bird (like all wildlife, whooping cranes move around when feeding). The survey was not restricted to an area but was instead a search for the birds. At the time, this may have been a reasonable approach because of the small size of the whooping crane population and the observer’s intimate knowledge of the birds. This way of surveying the birds had been the primary way the U.S. Fish and Wildlife Service (USFWS) conducted the census.

Counting each individual bird was easier when the wild population of whooping cranes was small, but the population has grown on average 4.6% every year. The birds are no longer concentrated in one easy-to-count location and have spread out well beyond the Black Jack peninsula. For the last several years, whooping cranes have been documented well outside their traditional wintering grounds, some as far away as Granger Lake, near Austin, Texas.

The most rigorous scientific methods must be applied in order to get the most accurate census of the crane population. It’s not only important to try and gauge the population, biologists must also get information on where the best habitat is, where the birds are feeding and more. Because of the slow but steady recovery of the whooping cranes, the USFWS has had to reconsider how they estimate the population of the wild flock. It is no longer feasible to count individual birds.

Obtaining a true census of any natural, free-ranging wildlife population is exceptionally difficult to achieve for two primary reasons. First, most study areas are so large that you can’t guarantee that an animal won’t enter or exit the area during the survey. Second, various circumstances might lead an observer to miss an animal, including the unpredictable behavior of wildlife, the density of vegetation and observer fatigue.

These and other factors were considered in 2011 when the USFWS decided to change how they collected this important data. Some of the areas to be improved upon included: 

  • Developing Survey Protocol – Data are only as strong as the methods used to collect them. Written protocols provide critical details regarding survey methods to help train future observers and guide the appropriate use of survey data. A protocol should include clear definitions of the survey’s objectives and which data will be collected to meet those objectives. 
  • Standardizing Data Collection – The methods used to estimate the whooping crane population should be able to be repeated by other observers on subsequent flights. The survey should be designed to increase the accuracy and reduce biases caused by environmental conditions, observer experience level, etc.  
  • Estimates of Precision, Accuracy, and Bias – The U.S. Fish and Wildlife Service recognizes that it is important to consider data and its associated uncertainty in the management and conservation of species and their habitats. Therefore, survey methods should be based on a sampling design that could be defended statistically and can provide adequate precision. 

Updates to Survey Methods
Developing Survey Protocol
A formal whooping crane survey protocol is currently under development, the first draft of which has been completed (and can be obtained by contacting the Aransas National Wildlife Refuge). This protocol is undergoing professional peer review to ensure the methods are appropriate, scientifically defensible, and professionally valuable. The survey protocol will describe measurable survey objectives, sampling design, field methods, data management, data analysis, and reporting procedures. Reviewer recommendations will be addressed prior to finalizing the protocol. The current protocol prioritizes four objectives for the survey:

1. Estimate whooping crane abundance in the surveyed area
2. Estimate the number of whooping crane pairs that bring juveniles to the wintering grounds
3. Estimate the number of juveniles relative to the rest of the winter population
4. Identify habitat characteristics used by whooping cranes

Standardizing Data Collection
In more recent years, surveys were typically extended to areas where whooping crane territories were known to occur. The boundaries of these areas are the basis for the sampling area. Additional sampling areas were included where biologists expect whooping cranes may soon colonize. To ensure each area was surveyed in a uniform and consistent pattern, transects that are 1 km apart were established for each sampling area.

To standardize the search effort and increase the efficiency of how we collect the data, each survey is conducted by a pilot and two observers looking out different sides of the plane. Upon detecting a crane, the observers record the bird’s location on a high-resolution satellite image using a touch screen laptop equipped with a wireless GPS. Surveys are conducted at approximately 60 meters altitude and 160 km/hr.

Historically, most of the birds have arrived on their wintering grounds of Aransas National Wildlife Refuge by mid-November. Because this is when the most birds are likely to be within the sampling area, future surveys will be conducted the last two weeks of November and the first two weeks of December. Surveying during this time frame should provide the most scientifically defensible estimate of the whooping crane population within the surveyed area

Understanding Imperfect Detection
By itself, the number of cranes seen on an individual survey does not provide a scientifically sound estimate of the whooping crane population. To be scientifically valid, a survey method has to consider the number of birds that might have been missed. Different things can affect the number of birds seen like the weather, habitat type, observer fatigue and more. A well-respected scientific model used for estimating wildlife populations is Distance Sampling.

For more than 30 years, scientists and biologists have been using Distance Sampling to survey other endangered species, including fin whales, Karner blue butterflies, lesser prairie-chickens, and raptors. This scientific model does not factor in an observer’s assumptions about the birds’ territoriality and it does incorporate the possible effects of weather, search effort, observer fatigue.

To understand factors that influence seeing a bird on the ground from an airplane, refuge biologists did experimental surveys using a known number of whooping crane decoys. During September 2011, 104 decoys were placed at randomly selected locations on the Blackjack Peninsula at Aransas National Wildlife Refuge. The two observers conducting the aerial surveys did not know the location of the decoy groups. Biologists conducted four surveys and each time the whooping crane decoys were moved.

The four surveys were analyzed in three different ways and revealed a ‘census’ of the decoys at 111 decoys, 104 decoys and 96 decoys. Statistical analysis also revealed that the distance from the transect line, position of the sun, and the number of whooping cranes within a group influence the outcome of the survey. An observer is less likely to see a whooping crane on the ground when the observer is facing the sun, medium when the sun is overhead, and highest when the observer has the sun at their back. However, there was little difference between detection probability when the sun is overhead and when the sun is at the observer’s back. Therefore, survey efficiency will be improved by using two observers (one on each side of the aircraft) when the sun is overhead (i.e., midday surveys).

Future Directions
Identifying Quality Whooping Crane Habitat
Understanding the habitat whooping cranes prefer and where that habitat exists on the landscape can help guide management decisions and conservation efforts. If we know quality whooping crane habitat exists outside of the surveyed area, it helps to prioritize and effectively implement habitat conservation efforts. 

Note:  This summary was submitted November 2012.  The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service. All data and conclusions contained in this report are preliminary and subject to revision. The assessment is provided on the condition that neither the U.S. Fish and Wildlife Service nor the United States Government may be held liable for any damages resulting from the authorized or unauthorized use of the assessment.