Echoes of Fear (9 Comments)


In 1999, Brown et al. introduced an interesting concept to the world of predator-prey interactions.  Previously, ecologists modeling these interactions treated predators as clever and prey species as sessile and rather stupid.  Brown et al. posited that predators and prey interacted through a game of stealth and fear.  Rather than a predator simply reducing numbers of prey directly (by gobbling them up), they reduced the numbers of prey by making them fearful.  This fear, in turn, altered the prey’s behavior.  Prey species became more vigilant, foraged differently, and moved around more.  Incorporating fear into the predator-prey equation, it was thought, could help ecologists better predict population dynamics of predators and prey alike.

As it turns out, including fear into the ecological calculus has helped ecologists to understand much more than how predators and prey interact.  Soon after the idea of the “Ecology of Fear” was introduced, Ripple and Bescheta (2004) found that the fear-based interactions between wolves and elk in Yellowstone National Park could influence far more than the number of predator and prey.   Changes in grazing behavior in response to increased wolf numbers led the elk to abandon riparian corridors, since these areas made detecting wolves difficult.  In turn, willow and cottonwood, no longer intensely grazed by fearless elk, recovered in number, size, and general health to levels recorded before wolves were extirpated from the park.  The increased vigor of these trees led to an increase in beaver colonies in the park, a species which had become rare in the park since wolves were removed.  All for fear of wolves.

Brown et al. developed the ecology of fear based on ecosystems dominated by “large, fierce” predators like mountain lions or wolves.  But recent evidence suggests that fear can act on much smaller scales, creating echoes that can impact critical ecosystem functions.

The recent (2012) paper by Hawlena et al. is one of those papers that simultaneously blew my mind, excited me, and induced academic despair regarding how ecologists are ever going to get a handle on how ecosystems work.  Their research reveals that the interactions between organisms in an ecosystem are incredibly complex and potentially more far reaching that this grassland ecologist had ever really considered.

You see, it is pretty well accepted that the quality and quantity of plant litter is one of the most important determinants of how nutrients are returned to soils, simply because there is so much of it.  Plants shed leaves and stems, those leaves and stems decompose through the actions of microbes and fungi, and the carbon and nitrogen in those tissues are returned to the soil to feed future plants.  The rate at which a community or ecosystem cycles nutrients through the process of litter decomposition is considered a major and important ecosystem function.

But what Hawlena et al. wondered was whether or not creatures a little higher up the food chain might have some measurable impact on rates of nutrient cycling.  Insect herbivores are thought to control litter decomposition primarily by eating leaves and stems, and preventing them from ever hitting the soil as litter.  But herbivores themselves can become part of the “litter” in an ecosystem when they die, if their body is not directly consumed by a predator.  And an herbivore’s body is packed with nutrients rare in plant litter and very necessary for the ecosystem…nutrients like nitrogen.

As it turns out, spiders and grasshoppers are the small and not-so-fierce analog to wolves and elk.  Grasshoppers alter their grazing, behavior, and metabolism in response to spider presence, which impacts the amount of carbon (C) and nitrogen (N) present in their bodies.  Stressed out grasshoppers should have less nitrogen and more carbon in their bodies, and make poorer-quality litter than fat, fearless grasshoppers.  Based on this assumption, Hawlena et al. set out to determine:

  1. Whether spider presence measurably alters grasshopper C:N ratios through fear
  2. Whether scared (but dead) grasshoppers or calm (but dead) grasshoppers change the rates at which microbes cycle nutrients when added as litter
  3. Whether this change in nutrient cycling impacts the rate at which plant litter is decomposed
  4. How long such a change might last in the ecosystem. 

Perhaps unsurprisingly they found that, yes indeed, terrified grasshoppers are significantly more nitrogen poor than their calm counterparts (Side note: this portion of the experiment involved rearing grasshoppers in cages filled with spiders that had their mouths glued shut!).  When they brought the carcasses of two groups of grasshoppers into the lab and added them to soil, they found that both carcasses stimulated the microbial community in a similar way (Figure 1A; again not surprising, as grasshoppers make much better litter than plant material.  They are like a microbe smorgasbord).

What was really interesting, though, was that despite a relatively small (~4%) but statistically significant difference in C:N ratios between the two groups of grasshoppers, being a stressed or calm grasshopper carcass dramatically impacted how microbes broke down plant material.  It seems that, when stimulated by a nitrogen-rich grasshopper carcass, the microbes go on a binge and just dig into the plant material, reducing it by 62% when compared with microbes stimulated by nitrogen-poor grasshoppers, a very significant change in this ecosystem process (Figure 1B).

And the effect of scary spiders is long-lasting (as anyone who has ever unexpectedly walked through a spider web knows).  When soil collected from field-based experiments where spiders had been excluded but grasshoppers were allowed to grow, die, and decompose was compared with soil where spiders (with their mouths glued shut) were allowed to scare grasshoppers, soil under calm grasshoppers decomposed litter more efficiently for up to four months after the grasshoppers had died.  This demonstrates that the effect of spiders is detectable over long time periods, even when other ecological processes are acting.

This paper is fantastic and the processes it reveals are humbling.  The processes controlling ecosystem function are much more nuanced that ecologists typically account for.  A grasshopper’s fear of a spider has echoing consequences for an entire ecosystem. Working in prairies and focused on plants, I never would have considered that the spider crawling up my pant leg might be what is driving the fertility and plant diversity of the patch I am standing in.


Brown J.S., Laundré J.W., Gurung M. & Laundre J.W. (1999). The Ecology of Fear: Optimal Foraging, Game Theory, and Trophic Interactions, Journal of Mammalogy, 80 (2) 385. DOI: 

Hawlena D., Strickland M.S., Bradford M.A. & Schmitz O.J. (2012). Fear of Predation Slows Plant-Litter Decomposition,Science, 336 (6087) 1434-1438. DOI: 

Ripple W.J. & Beschta R.L. (2004). Wolves and the Ecology of Fear: Can Predation Risk Structure Ecosystems?,BioScience, 54 (8) 755. DOI: 

6 November, 2013

November 6, 2013

Comments Leave a Comment

Please review our commenting policy.

  • There’s a paper I cite all the time by Kimbro et al. where they put “boxing gloves” (which I think amounted to mesh bags) on the crabs in the “fear” treatment so they couldn’t crush the snails. I suppose it’s like the bands on lobster claws – but somehow gluing spider mouthparts is so much more rediculous/awesome.

  • I once had a job as an undergraduate that involved painting really intricate patterns on the backs of ants for an experiment. I imagine gluing spider mouths shut is done in a similarly tedious way, but probably more thrilling/terrifying (depending on who you are) because we are talking about spiders.

  • I’m curious how different the levels of stress response and associated ecological effects are when the grasshoppers are allowed to be mobile. Surely it must be more terrifying to be surrounded by spiders *and have no means of escaping them* than to encounter spiders in general.

    In the wild, you can move to / live in patches of low spider density, flee when in the presence of a spider, etc. That’s not an option when you’re trapped in a cage with them.

    • Tor, I agree. That is a facet of this interaction that the authors didn’t really look into. They reared the grasshoppers with and without spiders present, and then brought their bodies into the lab to test effects in a microcosm set up. It would be really cool to recreate this experiment with differing densities of spiders to capture differing levels of stress, to look at the effect in situ and on the microbe and plant community, rather than in a microcosm, etc. There are always more questions raised by a study than are answered!

      • Yes, it’s really tricky trying to incorporate all of the potential for context dependence for these effects, but there is a pretty good body of work exploring how non-consumptive and trait-mediated interactions can change in different contexts. If anyone is interested, here are a few papers that come to mind. Tadpoles, snails, and daphnia (oh my!) are common model systems in this area:
        -Effects of temporal variation: Hamilton and Heithaus (2001) Proc B.
        -Effects of variation in predator density/risk cues: Schoeppner & Relyea (2008) Ecology
        -Effects of conspecific density & predator density: Relyea (2004) Ecology
        -Availability of refuge habitat: Trussell et al. (2006) Ecology Letters
        Orrock et al. (2013) Ecology

  • A thought that came to me that predator themselves could be responding to this change in prey-quality, if your surrounding prey items are all stressed, low quality individuals you will need to spend more time/energy hunting. In this context what strategy would they develop? Repulsion from patches with too high predator density, targeting of young naive individuals…
    Very nice post and this article was also a revelation to me and extremely relevant to my work!

Leave a Reply