WEST NILE SURVEILLANCE
PLANT ECOLOGY & WEST NILE
The relationships between plant ecology, the vectors and hosts linked to west nile, and the chances for human infection by west nile are not as simple as epidemiologists often present them to be. We typically think of west nile as the result of the mosquito carrying the cause for the disease (a virus), and so we tend to relate this disease to the water-features of the mosquito habitat, and at times, the effects of the weather on mosquito breeding and hatching activities. We do not take into consideration the impact of plants on mosquito ecology, like the way that plants can impact the distribution of animals fed upon by mosquitoes or on the ability of plants to make an area more or less inhabitable by either the vector or the host.
Plants have much to do with mosquito behavior because of the impacts they have on the feeding and mating-breeding environments. When we first learn about mosquitoes, we are often given the example of a Cattail-dependent species that likes to punch holes or wedge its way into the leaf base of the cattail in order to lay its eggs. There are also mosquito species that in larva form prefer waterbound plants like the reeds and grasses because these plants shade their natural setting and help to keep it at a comfortable temperature in spite of days with a hot sun. In areas where Malaria is a major disease problem linked to mosquitoes, it is often taught that residing near particular tree species where the vector likes to swarm increased you chances of getting the fever. Oftentimes, plants play an important role in the ecology of mosquitoes; the major problem is that few people know anything about this aspect of natural and human ecology.
In general, plants are attractive to some insecta due to their stench (like some arums), the heat they produce (Symplocarpus foetidus-the skunk cabbage), their pleasant smell (i.e. the flowers), their color and appearance, their resemblance to another food or nutrient source (the venus fly trap and drosera) and even their ability to provide the insect with an essential nutrient used to produce their own sets of hormones. Other plants are capable of producing natural repellants for insects, the most famous of which is the mints, especially pennyroyal or pulegium, which irritates their respiratory apparati when the essential oil in evaporated form enters the sensitive respiratory structures.
However, the primary reason plants can be related to mosquito activities is the shade it can provide. Plants that have large crowns (like trees) can make an area stay cool for much of the day, and due to the combined effects of transpiration (putting out water in vapor form from the leaves) and reduced evaporation rates for the grounds beneath, these trees help to produce a more humid environment which is more supporting of the mosquitoes and other insecta in general.
Both tree type and tree form help to define an area as being mosquito-friendly or not. Trees with exceptionally high crowns may be less attractive than low-liars, should the crown bottom be so high there is too little temperature and/or humidity control. Shrubs and trees with lower crowns are more apt to produce cooler environments. But just how dark and cool does this environment need to be?
Animals frequent areas with well-defined shade trees; beneath low-lying crownbearing shade trees, they may also be less noticeable to predators. In the case of tree or shrub type, species is very important in demonstrating whether or not the species of tree/shrub may even be related to certain mosquito species behaviors. To go one step further in this reasoning, some local trees such as Sassafras officinalis (Sassafras), Lindera benzoin (Spicebush), Ailanthus glandulosa (Tree-of-heaven) and the various conifers (most often Pinus spp. (pines), Tsuga canadensis (Eastern Hemlock) and Juniperus spp. (the junipers)) put out fragrances that are irritating to insects respiratory passages.
Often in New York habitats, the largest shade-producers and thickest forest-makers consist of just several species. These large crown-growers, residing in woodlands areas, are also major shade producers, and since they are often in uninhabited regions, they provide enough space for mosquitoes to fly at various elevations in the forest, both above and below the canopies at specific heights or in relation to specific types of trees such as conifers or not. For this reason, the dominant tree genus of a trapsite may be related to specific vectors species captured at that site, and this information used to engage in point-related spatial analyses of mosquito behavior within specific plant settings and around specific shade trees.
Another relationship exists between plants (trees) and trap sites relating to the animal host species. It is possible that the tree is important due to the animal host or carrier of the west nile, and not just the needs of the vector. One might example of this research might pose as a question for this study: “Can heavy crowns trees like those of the Oak trees be linked to areas where a specific species of mosquito can also be found and trapped?”
The following map shows one way in which this analysis can be rapidly carried out, assuming detailed ecological data was gathered for each trap site. Other methods to use for following up on this question might include a use of landsat or other remotely sensed images in order to define well-shaded regions due to tree canopies and then relate this visual object to the production of traps beneath it.
Specific tree types may be of interest to the field investigator. In the Dutchess County area, maple and oak trees play an important role in most of the local vegetation regions. Whereas maples are more common to forested areas, oaks are important both in woods and open field settings. In the open field setting, oaks can serve as important shade trees making the difference between whether or not a site is inhabitable by mosquitoes versus a site that is there just for a temporary animal host or pass-through.
The reason a tree like an oak can be important ecologically in thie field setting has to do with the nature of its tree crown as a shade provider and air cooler for this particular habitat. Contast this use for the oak with another setting, such as oaks surrounded by other trees, this primary effect of the oak tree crown may also be related to other neighbors like maple and hickory.
There are distinct chemical differences between major tree types. We normally do not think of this, but when we camp, placing a tent under a conifer does have a different effect on the ambience and air quality (smell) once nighttime arrives, than oak (which can stink and be a little ‘sourish’ (vomit-like)), or maple (fairly non-aromatic, even when it flowers). Do these features impact mosquito species habits? The following was a review of the sites with oak trees.
In this review of some of the most productive sites with tree ecology, notice the presence of Acer saccharum (sugar maple), for most of these sites, and its close relative Acer rubrum (red maple) at the remaining non-Acer saccharum containing site.
TOPOGRAPHY & WEST NILE
The impact of changing elevation on vector species distribution along three major streamedge settings with exceptional elevation changes in shoreline and floodplain terrain.
Ecological Studies performed lengthwise along two Creek-edges.
The Fishkill Creek and the Wappingers Creek are the two main streams that cross Dutchess County between eastern county regions, in and close to Connecticut, and the Hudson River forming the county’s western edge. In spite of the relative proximity of each of these creeks to each another, each has its own distinct series of habitats and environmental settings situated along the creek. The more southern Fishkill Creek has a slightly more meandering-like behavior, with widely-defined floodplains formed along much of its central and southwestern portions before empying into the Hudson River. The more northern (actually central Dutchess County) Wappingers Creek traverses large farming and livestock areas in the eastnortheastern to northnortheastern portions of the county and in the central county passes through approximately 6 to 10 miles of heavily population suburban settings (depending on where you draw the urban/suburban boundary), half of which includes several sections of moderate to deep ravine settings (25′ to >50′ depth from water surface), before making its way through the last mile or so of creek which crosses deciduous forest areas, numerous gravelly creek-edge settings, one significant waterfall, a ravine and combined riparian-estuarine setting before emptying into the Hudson River. Most unusual for the Fishkill creek is that it passes through a channel formed at the base of a mountain ridge for approximately 3 miles, including one narrow channel in a well-forested area before reaching the river, with much of the region well shaded by trees. In terms of human ecological features, after passing through the agricultural setting, the Wappingers Creek passes through ten miles of numerous, consecutive, heavily populated settings, followed by an old manmade lake setting, and a narrow gorge with a channel leading into an old industrial setting (where the water and company boilers used this water to service entire factory setting during the 19th century). This creek then continued to the river along much slower flowing channel, much of it manmade, before emptying into the Hudson.
The ecological settings along the Fishkill creek demonstrate fairly natural settings crossing large flood plains (some portions 3+ miles wide), in spite of significant development of this region; wooded areas are less impacted by housing contruction and flood plain topography is transected but not changed throughout much of the river edge. In one area, a well-known dump site can be located (primary west nile vector setting). The Wappingers Creek agricultural settings also bear a floodplain, although perhaps only half the width and area of the largest Fishkill Creek floodplains. The lower area lacks floodplains (except immediately adjacent to the river), and bears significant housing along much of the river edge. Most of the large suburban settings in this former river floodplain were constructed on landfill sites, formerly serving as wetlands, therefore much of the wetlands for these natural sections of the river have been minimized. Mosquito populations along each of the two creeks demonstrate the difference in mosquito ecology between these two creeks. The manmade, seminatural settings make the Wappingers Creek an area where studies focused on human population and special landuse and topographic features can be reviewed in relation to mosquitoes and west nile ecology. The Fishkill Creek is an example of a series of fairly low-impact natural settings, defined by a combination of broad floodplain wetlands sections interspersed with several community suburban settings and a few very unique perimontane and ravine or channel ecology settings.
Most of the ecological aspects of these two creeks were evaluated for vector species spatial distribution and ecology, and reviewed extensively for other natural ecologic features focused on the impacts of topography, channel/ravine form and ecology, peririparian montane ecology, and plant-related geographic conditions on vector-host species spatial distributions. These in turn could be related to the human population features linked to west nile transmission behavior within this middle county in lower New York State. Due to human ecological features, the species along the Wappingers Creek are more likely to be West Nile carriers. Due to natural ecological and topographic features, the mosquito species along the Fishkill Creek, although capable of carrying west nile, is found to be in considerable competition with numerous other species, with a diversity of mosquito species more variable than that of the Wappingers Creek. Moreover, these species are less likely to carry the disease and transmit it to humans due to the distribution of human settlement areas in relation to the actual sources for Culex pipiens-restuans, the natural vector of west nile in this county. Cx. pipiens-restuans is more apt to reside and infect people with west nile in the more urban-suburban densely populated settings, where its primary hosts (crows and blue jays) tend to be in closer contact with human ecology related features such garbage-ridden former woodland sites, illegal dumpsites (former industrial settings), and stagnant water bodies that have little contact with a nearby creek feature.
One the lower end of the Fishkill Creek it comes close to the small mountain range, which includes the best known peak of the area Mount Beacon. This region has a number of neighborhood setting situated along both sides of the creek edge. One is accessible from the north and east for the most part, the other requires access from the southside from the downstream and the roadways commencing from the southsoutheast, travelling upstream with the creek to your left.
A number of studies were performed more upstream, on the other side of a village setting in a fairly well-developed and often trashed region of the creekline. This study was done for comparison; this area chosen differed considerably in land use, land form and natural setting related features.
The traps were set on the east shore of the creek, about 30 to 40 feet from water edge, and along a fairly straight line heading away from the stream, across a boulder-ridden talus area covered with shade due to heavy tree cover, and up a gravelly slope about 200 to 250 feet from the talus. the overall elevation above the creek surface is about 250 to 300 feet, with a slope on the mountain face measruing about 30 to 45 degrees (slightly more in some spots).
The reason for setting these traps is identical to the study in which a DEM used to display the results (done at the more northern part of the Wappingers Creek). It was to determine how elevation impacts species type and density. The results of this study showed that species type and density changed considerably all along the mountain and creek edge on up to the first terrace at 75-100m. It was suspected that canopy may have been a factor in influencing the species change, due to either high elevation-related selection features instinctive to certain species (again, see coverage of the same in Medical Geography) or simple microclimate, subcanopy-related differences.
Not unexpectedly, the riches diversity of mosquitoes was found at the creek edge, where Anopheles quadrimaculatus was captured. With the exception of one trapsite, Aedes trivitattus was slowly replaced by Anopheles punctipennis as one ascended the slope.
A transect study of mosquito species distribution across a broad flood plain.
The broad flood plain of the Fishkill Creek has its advantages in the more heavily populated regions. These regions, although densely packed at times with housing developments, are less impacted ecologically by these major changes in land use patterns. For the most part, these housing communities continue to maintain their dense foliage produced by the local trees, and often lack areas with sizeable developments due to the large wetlands that exist. These section fo the creek edge differe greatly from regions where wetlands are completely filled in or modified from their natural ecologic conditions, conditions mroe typical along the Wappingers Creek. Given this peririprian ecological setting, one would expect, given the ecological setting, that this portion of the creek flood plain would be rich in positive testing vector species and animal host cadavers. Such is not the case.
To develop a map of vector species ecology along the creek both lengthwise along the edge of the creek and laterally across the floodplains, at least two series of trapping and analysis have to be carried out. Traps must be set along the creek edge as much as possible along the creekbanks, and traps have to be set going away from the creek to determine how distance from the creek hels to determine species/vector population form and density patterns. In theory, various sections of the creek are going to be different. In areas well-developed, it is expected that the west nile vector species will more likely be trapped as the primary resident of the area, or making up a large amount of the mosquito population found in these developed regions. In undeveloped settings, even though the presence of Culex may be documented, the natural conditions make it less likely to be trapped as part of the mosquito population due simply to probability-related features. In fairly well-established mosquito ecology settings, numerous species will reside in such settings, competing spatially with the Culex vector species.
For a variety of reasons, the increase population density of mosquito species in these natural settings will impact human behaviors in such as was as to reduce the likelihood for human contact and infection. The second reason species rich regions are less likely to result in human cases relates to the likelihood that these more favorable natural settings provide the true vector species with alternative sources for blood, ranging from birds to both wild and domestic animals. Only once moves away from the ecologically rich creekedge does on see the likelihood for competition decrease and the possibility of infection non-human species decrease due to ecological changes. This important features of west nile ecology across large flood plains is illustrated by the following results of the study of the area outlined by the red box.
The slope of a landsurface is essentially a mathematical measure of the steepness of its hills. A 90 degree slops is a perfect cliffface, a 45 degree slope is a steep hillside angled at 45 degree from the flat surface forming the bottom of the hill-edge. Fairly shallow slopes are those which are less than 3 degrees. In the case of the Dutchess County maps providing slope data, low slope ranges may be labelled as 0 degrees (no slope), or ❤ degrees. In the case of maps with slopes provided as percentage values, we typically find 90 degrees represented as 100%, 45 degrees as 50%, and ❤ degree slopes and <4.5% or <5% depending on the version of maps being reviewed.
In the above slope maps, the darker the region the steeper the slope. One can see several areas where the Fishkill Creek passes through failry narrow, steep slope creekedge sites, but these are few and far between for the most part. For much of the creekedge, slopes remains fairly low and so much of the floodplain identified on the base map for this image, consists fo fairly low slopes, most well under 10% or 9 degrees. The base maps used to determine these slope areas were the Side-Looking Airborne Radar images (SLARs, the base map in the above image) and National Elevation Data maps (NEDs), both either obtained through Geomedia or downloaded from an adequate web sourse). This data is also available from the local soil maps (provided by the NY GIS datasource site, note: soil maps only provide average slopes for given soil-defined settings only as an average areal-based calculation), but is not reliable or accurate in the three-dimensional sense. Depending on the types of analyses that had to be performed, the following results were produced using either Idrisi32 or ArcView 3.2 with the Spatial Analyst extension.
Aspect is in part a result of the slope features of a given area. It is the measure of the landsurface as this surface is impacted by solar activity or energy. Surfaces with higher aspects face in a southern direction, with tapers heading in both the westerly and easterly directions. When using GIS to identify areas based on aspect, the traditional direction of the sun is used and selection is made for areas with high aspect or high solar energy; the areas not selected are therefore those areas where solar radiation is poor, regions where cooler microclimate settings exist and where most mosquito species are more apt to remain dormant when they are not feeding. These areas also possibly represent high probability breeding settings given the right hydrological and topographic features.
By combining the results of the slope map and risk map we produce a map that depicts the areas of highest risk for mosquito-exposure and potential west nile infection, based primarily upon areas with shallow slopes or fairly level regions (suggesting prime mosquito habitats settings), in combination with the right aspect features (suggesting more favorable solar exposure settings).
The remaining ecological feature important to this study of west nile ecology pertains to, once again, the phytoecological features of one fo the primary breeding areas along the creek. We expect a region with a large flood plain and broad assortment of mosquito breeding pools and the like would be more productive and would hae been one of the settings in which either one of the first west nile positive testing hosts (crows) was found, ot where a positive testing Culex pipiens-restuans could be located. The human ecology of this area is substantial enough to favor one or more of these findings in disease surveillance. Like another ecologically rich portion of this county (the Pine Plains trap settings), west nile is not easily found or located in spite of ongoing trapping and dead animal host testing.
One possible reason for this lack of positive testing vector and host cases (and human for now) in ecologically-rich settings is the competition related features of these wetland and peririparian ecosystems. It is possible that mosquitoes known to be carriers of west nile may not be as much a threat to the local human population in placed where natural ecological conditions set the stage for their biting habits. In rich ecological settings, the local environment provides them with ample animal hosts to bite instead of humans. Equally important, in areas where rich, dense populations foprm swarms, the presence of other competitive species in the same ecosystem makes it more that mosquitoes will find adequate food courses, changing the probability that a vector species will make the right contacts required to produce an infected animal and/or human case. The reasons for this lack of success as a vector relate to both the phenomenon of swarm density changes over space intertwined with species-related behavior changes induced by other highly competitive non-infectious swarms. The simple presence of large enough swarms may be enough to reduce the probability that humans will remain outside long enough and in the right state so as to experience large numbers of bites. Not that these bites faily to occur in highly diverse regions, but rather there is a likelihood that these swarm patterns will directly change human behaviors, just enough to reduce the likelihood for west nile infection taking place in numbers proportional to the swarm density.
In this study, a fairly straight roadway just over a mile long, traversing the flood plain, was inspected and a series of traps were set at fairly equal intervals along the road edge, taking care to select areas that are well-forested and demonstrate protective effects on mosquitoes related to the limited light penetration of the canopy, and there had to be a nearby small area wetland, pond, or other fairly stagant floodplain water formation, where presumably adults could rest, eggs could be lain, and new larvae could hatch.
The results of this study concentrated on just four species found commonly throughout the county (other species were captured, just not included in the population density change modeling analysis due to limited spatial distribution). The studies of their numbers and density within a swarm show that specific mathematical relationships existed in terms of their distribution across the floodplain. Species density (numbers) decreased at a fairly consistent rate across space; species numbers also reduced but in a more linear fashion. Mosquito species richness is greatest at the creek edge, thereby increasing the potential for competition and prevention of potentially infected vector-human interactions. The likelihood of infection taking place at the human level in fact becomes greater as one traverses the flood plains away from the river edge. This is because competition decreases as the density of swarms decrease exponentially, and the numbers of competitive species decrease as well. If the appropriate infectious species is present, it is more likely to be able to infect people, especially if available animal hosts also decrease in numbers and types as one travels away from the flood plain peririparian setting.
In the following representation of these results, notice how the potential vector Cx. pipiens-restuans (PRE) remains fairly constant in numbers across the entire flood plain, whereas the remaining three non-infective species reduce in counts as one travels away from the creek edge. This makes the percentage of potential carriers greater as distance from the creek edge increases. Given the right ecological conditions, Cx. pipiens-restuans may be able and willing to bite human hosts as we move away from the creek edge, a condition assisted by any reduction in other natural ecological conditions normally required by the other more natural-ecology (non-human ecology) based species.
- West Nile Surveillance