Smoothing and trend detection in waterbird monitoring data using structural time-series analysis and the Kalman filter

Abstract

Many wildlife-monitoring programmes have long time series of species abundance that cannot be summarized adequately by linear trend lines. To describe long time series better, generalized additive models may be used to obtain a smooth trend line through abundance data. We describe another approach to estimate a smoothed trend line through time series consisting of one observation per time point, such as year or month. This method is based on structural time-series models in combination with the Kalman filter and is computerized in the TrendSpotter software. One of its strengths is the possibility to test changes in smoothed abundances between years, taking into account serial correlation. The trend method is applied in the Dutch Waterbird Monitoring Scheme (DWMS), a monitoring scheme for migrating and overwintering waterbirds. Taking the numbers of overwintering Greater Scaup (Aythia marila) in the Netherlands as an example, we demonstrate three applications of the method: (1) trend calculation and classification for each year in the time series, (2) assessing alerts for alarming population declines and (3) testing yearly abundance against a population threshold. We discuss the situations where TrendSpotter is to be preferred over other methods.

Long-term population trends reveal increasing importance of a Mediterranean wetland complex (Messolonghi lagoons, Greece) for wintering waterbirds

“…Long-term population trends of the 28 waterbird species were assessed with the statistical software TrendSpotter 6.4 (Visser 2004a, b;Soldaat et al 2007). TrendSpotter uses a smoothing method for analysing environmental time series, based on structural time series analysis in combination with the Kalman filter, and it is the method of choice to smooth time series that contain only one value per time point, as in this study ).…”

“…the TCR expressed as a mean change rate per year, and confidence intervals could then be used to classify the trends per year in six categories: strong increase (lower confidence limit (CL) > 1.05), moderate increase (1 < lower CL ≤ 1.05), stable (confidence interval (CI) contains 1.00 and lower CL ≥ 0.95 and upper CL ≤ 1.05), moderate decline (0.95 ≤ upper CL < 1.00), steep decline (upper CL < 0.95) and uncertain (CI contains 1.00 and [lower CL < 0.95 or upper CL > 1.05]). For further details on TCR and YCR calculations and trend classification, see Soldaat et al (2007). All data counts were ln-transformed prior to the analysis to achieve normality.…”

Viability of meta-populations of wetland birds in a fragmented landscape: testing the key-patch approach

“…1 2 ) and a coverage of at least 15 years since 1990. Census data are collected by numerous volunteers and professionals using a standard protocol and a central qualitycontrol at SOVON (see also Soldaat et al 2007). We selected 15 species (Table 1) out of 30 potentially useful wetland species because (a) maps of variable density were available (some species are only recorded as present/absent, such as the reed bunting), (b) abundance should be distinctly higher inside the wetlands studied than outside (this was not the case for the hen harrier), and (c) species distribution should not be limited to a few colonies (as in the purple heron).…”

The introduction of subsidies for grassland conservation in the Italian Alps coincided with population decline in a threatened grassland species, the CorncrakeCrex crex

“…We used the program TRIM (TRends & Indices for Monitoring data) 3.54 to assess population trends by means of a generalized estimating equations approach which takes into account overdispersion and serial correlation (Pannekoek & Van Strien 2001, Soldaat et al 2007, Ludwig et al 2008. The proportion of missing data within our counts was relatively low (Pannekoek & Van Strien 2001).…”