Inventory and Survey Methods for Nonindigenous Plant Species (PDF)

Adaptive Sampling DesignChapter 7Timothy S. PratherBackgroundIncreased global trade has great potential for bringingnew plant species to our doorstep (Mack and Lonsdale2001), some of which may disperse and establish,usually along transportation routes (MacDonald et al.1989). Because an increase in nonindigenous plantspecies (NIS) in an area is related to the density ofroads in that area (Kalin et al. 2000), conducting anNIS inventory/survey biased towards roads makessense. Rivers and streams are also important to plantmovement (Nilsson et al. 2002; Parendes and Jones(2000). However, wind-dispersed plant species andspecies dispersed by birds will have less associationwith roads, trails, and rivers. The seeds of winddispersedspecies such as Russian thistle (Salsolaiberica) may move as little as 60 m and as much as 4km, depending on wind speed and duration (Stallingset al. 1995). Topography and thermal updrafts wereconsidered more important than wind speed in astudy involving mouse-ear hawkweed (Hieraciumpilosella) dispersal. As topographic relief increased,thermal updrafts increased in strength, resulting inlong-distance dispersal of at least 250 m (Tackenberg2003). Dispersal by birds is not as likely to be relatedto road and trail systems, but rather to perches suchas fencerows. For example, dispersal of seeds byhornbills in tropical West Africa measured 1.1 to 2.0km on average, with maximal distances of up to 6.9 km(Tackenberg 2003).A strategy for NIS survey that focuses on transportationroutes should increase the efficiency of the survey,especially as the size of the area managed increases.Furthermore, when rivers and streams are in proximityto roads or trails, areas downstream from infestationsshould be examined. Because a survey biased towardtransportation routes may not accurately assessspecies dispersed by wind or birds, transects perpendicularto the transportation route should be included.ObjectiveAreas of disturbance and human activity (roads, trails,etc.) provide both the conditions and the opportunityfor NIS establishment. The adaptive sampling methodprioritizes areas to survey by focusing on roads, trails,campgrounds, and waterways. Biasing the survey to areasmore likely to contain NIS improves efficiency, much likeusing a stratified sampling technique increases samplingefficiency. The primary objective of adaptive survey is toefficiently sample areas for NIS, particularly areas with littleor no prior inventory/survey data. An additional objectiveis to record all the areas sampled, rather than just locationsof NIS. Knowledge of areas where NIS do not reside is ofconsiderable importance for management and these data areuseful to better understand the potential distribution of anNIS.MethodThe adaptive sampling method presented in this chapteruses areas of human activity as a starting point. Forexample, the starting point may be a trail, campground,or beaching area for rafts. The area is then traversed, andlack of NIS are recorded until a target NIS is located. Oncea target NIS is located, the crew spaces themselves at twicethe detection distance (Figure 1) of the targeted species. Thecrew then moves away from the known location into thesurrounding area, either in concentric circles in gentle terrain,or by walking the elevational contour from the knownlocation. As the target NIS are found, the survey crewcontinues away from the new individual/patch, in circles oron contours, until no target NIS are found for a predefineddistance.NIS are mapped as point features and the radius of thepopulation is recorded. If infestations are large (radiusgreater than 40 m) and irregularly shaped, the perimeter ofthe infestation is walked and recorded directly as a polygon.NIS populations are recorded as separate points or polygonsif 20 to 40 m separate the individual plants or patches. Theseparation distance between patches is defined before thesurvey begins and depends on the visibility of the targetedspecies. Identifying every plant is not important; rather, it isimportant to identify the outer boundary of the infestation.The outer boundary is considered to be located when notarget NIS have been found for 400 m (species not winddispersed) or 800 m (wind-dispersed species). A surveydistance of 800 m has been adopted for wind-dispersedspecies such as hawkweeds (Hieracium spp.) because surveycrews have found that wind-dispersed species could befound up to 650 m from the last known plant. The radiusdata for each population is used later within the geographicinformation system (GIS) software to extend points tocircular polygons to represent the size of each infestation.Perpendicular TransectsOnce an outer boundary of an adaptive survey is determined(i.e., no target NIS have been recorded for 400 to 800m depending on dispersal type), the survey crew returns tothe trail (or other area of interest) and continues to survey.If NIS are not found for the next 4 km, a transect is taken56Inventory and Survey Methods for Nonindigenous Plant Species