Authors: Aaron C. Greenville, Vuong Nguyen, Glenda M. Wardle and Chris R. Dickman

Published in: Australian Zoologist

Abstract:

Field work doesn’t always go to plan and can lead to gaps in your datasets.

Long-term field-based monitoring is essential to develop a deep understanding of how ecosystems function and to identify species at risk of decline. However, conducting field-based research poses some unique challenges due to the frequently harsh environmental conditions or extreme weather events that may be encountered. Fieldwork issues can arise from vehicle breakdowns, wildfires and heavy rainfall events, all of which can delay or even cancel data collection. In addition, long-term monitoring often requires multiple observers, which may add observation bias to estimates of measured parameters. Thus there is an increasing need to develop new statistical techniques that take advantage of the power of long time-series datasets that also are incomplete. Here we introduce researchers to multivariate autoregressive state-space (MARSS) modelling; a new statistical technique for modelling long-term time-series data. MARSS models allow users to investigate incomplete datasets caused by missing values. In contrast to traditional modelling techniques, such as generalised linear models that only estimate error from environmental stochasticity (process error), MARSS models estimate both process and observation errors. By estimating observation errors, researchers can incorporate bias from different observers and methods into population or other parameter estimates. To illustrate the MARSS technique we interrogate long-term animal and plant datasets from central Australia that contain missing values and were collected by multiple observers. We then discuss the findings from the MARSS models and their implications for management. Lastly, we provide future applications that this new technique could be used for, such as studies of animal movements and food webs.

Media:

#FieldWorkFail? Making the most of incomplete long-term datasets, TERN Newsletter, September 2018.

Reference: Greenville, A. C, Nguyen, V.,  Wardle, G. M. & Dickman, C. R. (2018). Making the most of incomplete long-term datasets: the MARSS solution. Australian Zoologist, In-Press.