Ecological Archives E095-098-A1

John W. Laundré, Lucina Hernández, Perla López Medina, Andrea Campanella, Jorge López-Portillo, Alberto González-Romero, Karina M. Grajales-Tam, Anna M. Burke, Peg Gronemeyer, Dawn M. Browning. 2014. The landscape of fear: the missing link to understand top-down and bottom-up controls of prey abundance?. Ecology 95:1141–1152. http://dx.doi.org/10.1890/13-1083.1

Appendix A. Development of predictions for the impact of the landscape of fear for parallel effects between competing species.

Impacts on top-down, bottom-up, and parallel effects: competing species

The following is the development of the predictions from Table 1 (in main article, reproduced here below as Table A1) for top-down, bottom-up, and parallel effects when a species coexists on the landscape with a competitor. In these examples, the initial species is considered the dominant competitor and the second species the subordinate one. For these predictions we assume first that the two species share risky and safe patches, i.e., risky and safe patches are the same for both species. We then consider the scenario where they have distinct patches, i.e., risky patches for one species are safe patches for the other. In these examples we consider the simplest case where the dominant competitor is dominant in both risky and safe patches and in both forms of competition, scramble and interference. It is understood that in some species, competitive dominance can switch with patch type and that dominance can switch with the form of competition. Although not treated here for space reasons, these scenarios could be added to develop appropriate predictions.

The first scenario is where there is 75% safe patches. For the dominant competitor, there will be little difference than when it existed alone. It will dominate in the extensive safe habitat and bottom-up and top down responses will remain strong and weak, respectively. The reaction of the subordinate competitor and the resulting parallel effects, however, will depend if its landscape of fear overlaps or not with the dominant one. If it is shared, then the safe patches are shared. Under these conditions we would predict first that the subordinate species would be rare on the landscape and second that bottom-up effects on this species would also be weak (Table 1). Because the subordinate species has to compete with the dominant one within the extensive safe habitat, its numbers will normally be lower than predicted if based on resource availability. In the same manner, its population size will not respond as strongly to changes in resource levels because of being out competed by the dominant species. In common with the dominant species, we would predict that top-down effects would be weak and that the removal of common predators would not affect the population size. The subordinate competitor will continued to be suppressed by competition in the original safe patches and the now safe, risky patches. Upon removal of the dominant competitor, we predict strong parallel effects (Table 1) as the subordinate species could then expand within the abundant safe patches. In this case, interspecific competition would appear to be the major factor affecting the subordinate's abundance. However, it is important to note that interspecific competition may be the proximate cause but the ultimate cause is due to the makeup of the landscape of fear. This will be seen in the next example when landscapes of fear do not overlap.

If both competing species have different anti-predator capabilities, their landscapes of fear will be distinct from each other. In this case, the large amount of safe patches of the dominant species may represent risky patches for the subordinate one and vice versa. Again, nothing will change for the dominant species (Table 1) For the subordinate species, it can be relatively free of competition in its corresponding safe patches, which are risky for the dominant one. However, because these patches are fewer on the landscape, the subordinate species will still be rare. Even though it could respond to increases in resources, its response to bottom-up effects in population size increase would still appear as weak (Table 1). Because it is rare on the landscape, there will be little relationship between its abundance and that of its predators. Removal of common predators would not cause an increase and could result in decreases in population size. Because of the presence of the dominant species in the larger former risky patches, the subordinate one would not be able to expand in these areas. However, freed from predation risk, the dominant one could expand into the safe habitat of the subordinate one, at the expense of the subordinate's abundance. Top-down effects then would appear weak. In contrast to the previous example, parallel effects, in the presences of predators, would be weak (Table 1). Removal of the dominant species frees up the abundant patches that were safe for it but because these are risky for the subordinate, an expansion into these areas would be dampened by fear of using them and increased predation when they do.; If, however, both competitor and predators were removed, then the subordinate species could expand into the larger former risky areas. This would lead to the conclusion of a strong complex interaction between predation and competition. However the basis of that interaction is the makeup of the corresponding landscapes of fear of the competitors and how that affects the tradeoffs for food and safety and optimal patch use that are made.

Now let's turn our attention to an area of 75% risky habitat shared by the dominant and subordinate species (Table 1). We predict no differences in top-down and bottom-up effects for the dominant species from when it was considered alone. For the subordinate species, we first consider when the landscapes of fear overlap and patch risk levels are shared (Table 1). Both species have little safe habitat and will be rare on the landscape. As with the dominant species, the subordinate one would not react significantly to changes in resource levels. In fact it would react even less as it would compete with the dominant species for the limited available resources. Thus, bottom-up effects would also be weak for the subordinate species. In contrast to the dominant species, top-down effects would be weak (Table 1). The subordinate species would be rare and any expansion within and outside of the limited safe areas would be dampened by interspecific competition. Subsequently, we do not predict abundance changes in this species to correlate with predator numbers. Removal of common predators would, as indicated above, favor the dominant species. Subordinate would not be able to expand into previously common risky habitat because of competition with the dominant one. Consequently, under this scenario, interspecific competition would have impacts on the subordinate species.; Ironically, these parallel effects will appear weak in the presence of predators (Table 1). This is because we predict that the removal of the dominant species will not result in major increases in the subordinate one. The majority of the area is risky for both species and removal of the dominant one will not enable the subordinate to expand into these areas.

The last scenario is where landscapes of fear are distinct and the dominant species has mostly risky patches while the subordinate one has mostly safe patches. The two species will separate out on the landscape as seen above. However, in this case, the dominant species will paradoxically be less abundant than the subordinate one. As before, the dominant species will have weak bottom-up effects. The subordinate one, protected by predation risk from competition, will be able to take advantage of resource changes. Thus, we predict that it will exhibit strong bottom-up effects (Table 1). As most of its abundance changes occur in the ample safe habitat, predators will not respond to these changes. Predator removals will allow the dominant species to expand into its previously risky habitat at the expense of the subordinate one. This will result in actual reductions of the subordinate species upon predator removals. Consequently, top-down effects will appear to be mixed (Table 1). Parallel effects in this scenario will appear weak because the removal of the dominant competitor will only result in access to the limited risky habitat previously occupied by it (Table 1).

It is stressed that the predictions developed from these analyses represents the simplest case where spatial structure of rich and poor and safe and risky patches occur on the landscape. It may get more complicated if we have seasonal or daily pulses of resource or changes in predation risk, e.g. full vs. new moon. However, we argue that by incorporating these factors into the landscape of fear model, appropriate predictions of top-down, bottom-up, and parallel impacts could be developed and tested. We also stress that the examples presented here represent starting points in considering the impact of the landscape of fear on bottom-up, top;-down, and parallel factors. If field testing of these predictions proves promising, further development and refinement of predictions can be made to better incorporate the makeup of the landscape of fear into the population dynamics of a species.

Table A1 (Table 1 in main article). Predicted strengths of bottom-up, top-down, and parallel effects on population abundance relative to proportions of risky vs. safe habitat. For this example, we use high (75%) amounts of first safe and then risky habitat for a dominant competing species. Under each scenario, we add a subordinate competitor whose landscape of fear first overlaps with and then is opposite of the dominant species.

 

Bottom-up

Top-Down

Top-down

Parallel

 

 

Population

Removal

 

Dominant: 75% safe

Strong

Weak

Weak

 

Subordinate: 75% safe

Weak

Weak

Weak

Strong

Subordinate: 75% risky

Weak

Weak

Inverse

Weak

Dominant: 75% risky

Weak

Strong

Strong

 

Subordinate: 75% risky

Weak

Weak

Moderate

Weak

Subordinate: 75% safe

Strong

Weak

Inverse

Weak


[Back to E095-098]