The principal hosts for west nile may be interpreted in two ways. There are the active hosts that serve as carriers of the west nile virus and in turn if bitten by a potential vector in their natural setting may result in spreading the disease into this new environmental settings. There are also hosts that are infected directly by an infected mosquito vector, such as a dog or horse, and in some cases a human.
The monitoring of activities and mortality rates for crows and various raptor species of birds in the region is an example of host surveillance. This is done to provide information on the location of the west nile relative to the local ecological setting. A crow or other host may be infected with the disease after it collected following its recent death, but not pass this on to the local mosquito population and ultimately to other animal populations. This crow is different from a crow that represents one of several crows that recently died in a small area and where most likely a mosquito population infected by west nile is currently residing.
Utilizing traditional kriging tools, census block data, and moving windows for the identification of host clusters in high risk population settings.
There is a value of studying crow behaviors when trying to understand west nile diffusion patterns. In Dutchess County, crows as well as blue jays and hawks are the primary carriers of west nile. When bitten by an infected mosquito they have a significant likelihood of developing the west nile infection and transporting this virus to other locations situated a considerable distance for the nidus or site of origination. A number of studies suggest that crows in the New York area typically travel about 10 miles +/- a day, and usually return to the same nesting place for the night. Some studies show that crows may travel considerably more distance if migration to a new area is the intention of the crow and/or its mating and living partners. For the Dutchess County area, the general impression has been crows capable of carrying west nile to the county do so by travelling from west to east across the Hudson River. For this reason, it was often assumed during the first two to three years of west nile surveillance that these were the chief carriers and hosts of west nile for the region.
Each year, a large number of crows in fact die, due to natural causes, age, and numerous other environmental and human factors. Yet, the percentage of crows that died each year for the county was really not well understood, until they became the chief carriers of west nile and their carcasses, found in yards, fields, woodlands and along roadsides, became prime research materials for west nile ecologists. In 2001 and 2002, attempts to collect crow carcasses increased significantly, enabling as many as fifteen to be documented and shipped to testing labs each week in search of the immunogenic markers for west nile found in crow blood. In fact, numerous more were found and documented during the peaks of west nile infection periods, but typically not saved for shipping due to their condition and whether or not they still had viable, testable blood.
By collecting crows, the heavy reporting that occured enabled special ecological settings for crows to become better understood. It ends up that, at least for the Dutchess County area, the places where crows like to aggregate, converse, feed, mate, and recreate, are not necessarily the places where they will become infected by west nile or pass it on to mosquitoes and then humans.
Crow Call density by Zip Code Tracts in the most populated section of the County
The crow density map for all of the carcasses reported for two years shows that crows like to reside around people and urban areas, in close proximity to feeding sites, but in places where other important social settings are also within a short distance of them. The reason they reside in large numbers close to the primary urban region of the county also has to deal with the local forest setting and the large amounts of available human garbage and debris found near the city’s edge and interior, within which they can find food. One other topographic feature important to this area is the large elevation it has above the river edge, providing them with a place to rest while on longer trips, with a view that encompasses much of the immediate area for about 10 to 15 miles.
Understanding this latter requirement for crow roosting habits is helpful. Once I engaged in some research for an entire summer that documented crow roosts and migration paths, by walking these paths and documenting where their feathers are dropped. There are definitely signs of this that suggest crows tend to favor certain types of paths. Although they are willing to simply travel as the crow flies from one spot to the text, they also show the tendency to prefer routes that follow large narrow open spaces and roadways. They like to aggregate in well-forested, but not too densely forested areas such as parks, and they definitely like to roost and spend the night in areas different than those where they spend much of their middle and late day hours. Like people, crows like to frequent the most forested-open space park areas, where they can gather in significantly large numbers at dusk (by the hundreds). It is also possible to find single roosting trees where they like to sit; these places are easy to recognize because it is not unusual for them to have hundred of feathers of all sizes just beneath the dominant (most densely crowned) tree of that small area. Such an area may be only the size of a small yard.
In this map of the crows in Dutchess County, the positive testing areas are not indicated (some of them are on another map though with the 183 trap sites documented). A subsequent study was done using the GIS to identify census block areas with large numbers of crow calls, and ecological studies of each of the areas where just the positive tests species were found, especially when these positive carriers could be linked spatially to known equine, animal hospital, and human cases in the immediate region. Other areas researched ecologically, in single trap, transect and grid form, include a small privately owned international zoo, several privately-owned equine racehorse training sites, two livestock/farm animal veterinarian settings, and a privage facility specialized in raptor research.
Crow Density based on dead crow calls and field investigation
The use of nearest neighbor analysis and moving window monitoring techniques to identify animal host-defined west nile clusters.
One way to determine where cluster settings might exist is to monitor crow death frequencies and locations and rtelate this to census tract block information. Places situated in urban and periurban settings, with significantly higher percentages of people at risk based on the cnesus data, are those where special attempts must be made to monitor crow deaths, research sites where the numbers of deaths undergo a sudden increase, and determine if there is a cluster of dead bird reports that may have relevance to whether or not problems can impact a small heavily populated area.
Nearest neighbor analysis is used because it is not the numbers of crows that die on a specific block that matter. Instead, this is done in order to determine whether or not blocks could possible have a shared location or origin of the deaths. reports related to the people residing in that area and their nearby ecological settings are the first and most important features to pay close attention to, even before test results come back telling you whether or not the dead birds collected test positive.
Utilization of a Moving Window ArcView Avenues Extension to define risk areas based on Crow Calls
Another way to survey for the same type of features is to first make use of an ArcView Extension designed to do a moving windows search, by which it marks on the map using nodal points (the hard-to-detect red crosses) to form the corners of gridcells containing a certain number of the species you are searching for or more. (This would be based on the point data of these deaths which have been mapped.) With these positive testing cells identified, the blocks underlying can then be reviewed, as well as land use and water-vegetation features, providing you with a better idea on where and how to search for the locations with the highest risk for infection. [WARNING: One major disadvantage of the early version of this Moving Windows program is that it does not produce the nodal points as a shapefile; they are permanently placed on the view of your end product.]
The other pages:
.
.
Brian Altonen, MPH, MS 