Trends can be explored through the interactive App below.
Toggle between exploring trends for SRMS Regions and Natural Heritage Zones via the buttons in the left-hand corner. Select a species, parameter and up to two regions of interest to be plotted and explore the data table by filtering on the appropriate fields. N.B. When interpreting these data it is important to consider any associated caveats indicated in the caption below the plot or in the ‘caveats’ field of the table. Further descriptions of these caveats are provided below the App.
Caveats:
- Expanding population – population of a recently re-introduced species, known to be rapidly expanding. This means that traditional approaches to raptor monitoring (focusing on known home ranges or discrete study areas) are likely to underestimate rates of population growth, and bias measures of productivity towards older, more experienced pairs.
- Nest box based – a large proportion of monitored individuals are based in nest boxes. If either nest boxes tend to be preferred over natural sites or vice versa, numerical trends may not be representative unless a high proportion of pairs nesting in natural sites are also found and monitored. Moreover, because only a small population of any raptor species is based in nest boxes, if any measures of productivity differ between nesting attempts in boxes and those in natural sites, estimates of and trends in productivity may also be unrepresentative.
- No home range random effect – inclusion of the home range as a random effect in a productivity trend model caused the results of that model to depart unrealistically from the observed range of variation for that trend, so this variable was removed from the model. This could make the trend more prone to being unduly influenced by variation between individual home ranges; particularly when the home ranges contributing to the trend changed over time.
- Obvious gaps in coverage – some regions holding a substantial proportion of the population contribute few or no data to a national trend.
- Sample size small – mean annual sample size is less than 20. This is likely to decrease the precision of annual estimates, and to increase the influence of ‘noise’ (random variation) on apparent change from one year to the next. This is not based on any formal power analyses but simply highlights that trends based on samples of more than 20 home ranges are likely to be more robust/representative than those based on smaller samples.
- Variable effort – variation in sample size between years suggests that variable monitoring effort could result in inter-annual variation in the location and nature of home ranges that are monitored, or in the effort put into collecting data from these. Such variable effort could result in ‘noise’ (random variation between years) or, if effort increases or decreases over time, introduce bias into the trend.