Skip to main content
Dryad logo

Archive data for: Loss of predation risk from apex predators can exacerbate marine tropicalization caused by extreme climatic events


Nowicki, Robert et al. (2020), Archive data for: Loss of predation risk from apex predators can exacerbate marine tropicalization caused by extreme climatic events, Dryad, Dataset,


1. Extreme climatic events (ECEs) and predator removal represent some of the most widespread stressors to ecosystems. Though species interactions can alter ecological effects of climate change (and vice versa), it is less understood whether, when, and how predator removal can interact with ECEs to exacerbate their effects.  Understanding the circumstances under which such interactions might occur is critical because predator loss is widespread and ECEs can generate rapid phase shifts in ecosystems which can ultimately lead to tropicalization.

2. Our goal was to determine whether loss of predation risk may be an important mechanism governing ecosystem responses to extreme events, and whether the effects of such events, such as tropicalization, can occur even when species range shifts do not.  Specifically, our goal was to experimentally simulate loss of an apex predator, the tiger shark (Galeocerdo cuvier) effects on a recently damaged seagrass ecosystem of Shark Bay, Australia by applying documented changes to risk sensitive grazing of dugong (Dugong dugon) herbivores. 

3. Using a 16-month field experiment established in recently disturbed seagrass meadows, we used previous estimates of risk-sensitive dugong foraging behavior to simulate altered risk-sensitive foraging densities and strategies of dugongs consistent with apex predator loss, and tracked seagrass responses to the simulated grazing.

4. Grazing treatments targeted and removed tropical seagrasses, which declined. However, like in other mixed-bed habitats where dugongs forage, treatments also incidentally accelerated temperate seagrass losses, revealing that herbivore behavioral changes in response to predator loss can exacerbate ECE effects and promote tropicalization, even without range expansions or introductions of novel species. 

5. Our results suggest that changes to herbivore behaviors triggered by loss of predation risk can undermine ecological resilience to ECEs, particularly where long lived herbivores are abundant.  By implication, ongoing losses of apex predators may combine with increasingly frequent ECEs to amplify climate change impacts across diverse ecosystems and large spatial scales.


This dataset was collected over 16 months from 2013-2014 in degraded seagrass beds of Shark Bay, Western Australia.  Each row refers to a visit to an experimental plot, each measuring 3x3m and consisting of three main macrophyte types: the seagrasses Amphibolis antarctica and Halodule uninervis as well as benthic macroalgae.  At each plot, percent cover was assessed via diver by using a single 60 cm quadrat overlaid on the plot in a 5 x 5 pattern to cover the entire plot.  Because each quadrat measurement is strongly correlated in space to others within the same plot, the percent cover across all 25 quadrats was averaged to generate a single percent cover value for each macrophye group in the plot. As such, each row contains the average percent cover in that plot for that time point.  Later columns also group the data collected in a specific month by treatment group, and by treatment x bank subgroup.  The data are otherwise unprocessed; full details can be found in the methods of the manuscript.  There is also a column key on the second sheet.

Usage Notes

There are some missing values in the dataset at experiment start for benthic macroalgae, which were initially not measured because their cover was negligible.  Soon after experimental establishment, however, algae measurement began upon realizing that algae may exhibit strong seasonal cover effects and may be important later in the experiment. Likewise, plots that were destroyed by time a visit occured were not measured from that time point onward; one plot was removed entirely because it was destroyed too early in the experiment to warrant inclusion in the datset as a whole.


National Science Foundation, Award: OCE -1329408

National Science Foundation, Award: DGE-1038321

PADI Foundation, Award: Small Grant (RJN)

Florida International University, Award: Dissertation Year Fellowship (RJN)