Trophic Cascades in a Multicausal World: Isle Royale and Yellowstone

In the paper

“In the next two years, the wolf population crashed by 80% owing to the synergistic effects of food stress and disease (canine parvovirus) that humans had inadvertently introduced (Peterson et al. 1998).During this time, the population also likely suffered increasingly from inbreeding depression (Adams et al. 2011).”

What the sources say

• ?Not checked — no full text of the sourcecites Peterson et al. 1998 source
• Partly supported

  Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites Adams et al. 2011 sourceExperimental — accuracy not yet measured

In the paper

“First, estimates collected during 1971-2014 indicate that wolf predation rate (the proportion of prey killed annually by predation) is a strong predictor of growth rate for the moose population (R 2 = 0.67, p < 10 -3 ) (Vucetich et al. 2011).The estimated slope of this relationship is -1.8 ( ±0.21 SE).”

What the sources say

• Partly supported

  The Results report that predation rate predicted Isle Royale prey growth rate with R² = 0.67, but give P < 0.01 rather than P < 0.001. The Isle Royale methods describe observations through 2011, not 2014, and the passages found do not report a slope of −1.8 ± 0.21 SE. Page numbers were unavailable in our copy. (LLM, conf=0.89)

  cites Vucetich et al. 2011 sourceExperimental — accuracy not yet measured

In the paper

“In 1996, a severe winter interacted with food stress, causing the moose population to decline by ∼60%.In 1998, genetic rescue by one immigrant wolf bolstered the fitness of the wolf population (Adams et al. 2011).”

What the sources say

• Partly supported

  Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites Adams et al. 2011 sourceExperimental — accuracy not yet measured

In the paper

“For further context, predation rates <12% correspond with strong population growth for the IRNP moose population.More specifically, for years of predation rate <12% (n = 33), mean growth rate of the IRNP moose population was 0.06 ± 0.02 (SE) and growth was positive in 79% of those years (see Vucetich et al. 2011, figure 4D,E).”

What the sources say

• Partly supported

  The Results (page unavailable) report that predation rate was a good predictor of Isle Royale prey growth rate (Fig. 4d), supporting the general association. Our copy does not state the <12% cutoff, the mean growth rate of 0.06 ± 0.02 SE, or positive growth in 79% of the 33 years; searches for those figures did not find them. (LLM, conf=0.85)

  cites Vucetich et al. 2011 sourceExperimental — accuracy not yet measured

In the paper

“The distance beavers will forage from water depends on their preference for specific forage species (they will travel farthest for aspen); slope (Shelton 1962, Jordan et al. 2000); and the presence/absence of predators, which beavers detect by olfaction (Severud et al. 2011).”

What the sources say

• ?Not checked — no full text of the sourcecites Shelton 1962 source
• ?Not checked — no full text of the sourcecites Jordan et al. 2000
• Partly supported

  The Abstract reports that beavers avoided trails treated with wolf urine, supporting olfactory detection of predator cues. The Discussion reports that beavers foraged closer to shore on an island with bears than on one without bears. The paper does not establish that olfactory detection of predator presence caused the difference in foraging distance; the other factors in the citing sentence have separate citations. (LLM, conf=0.83)

  cites Severud et al. 2011 sourceExperimental — accuracy not yet measured

In the paper

“Because of advanced age, canopy trees in the western portion have been dying rapidly and will soon be functionally absent (Frelich et al. 2012).”

What the sources say

• Partly supported

  Section 2, “Herbivore dynamics in a changing climate” (page unavailable), says balsam fir is disappearing from the canopy at Isle Royale’s southwest end as older trees die without replacement. It does not say canopy trees generally are dying rapidly because of advanced age or predict when they will be functionally absent; it emphasizes moose-restricted fir regeneration. (LLM, conf=0.87)

  cites Frelich et al. 2012 sourceExperimental — accuracy not yet measured

In the paper

“Following these events (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011), mean predation rate increased to 13.8% ( ±1.6%), moose abundance declined to its lowest level (Figure 3), and balsam fir responded with increased growth”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Vucetich & Peterson 2014

In the paper

“Mean predation rate was high (10.7% ± 0.1%, SE) and had an important influence on moose abundance between 1971 and 1980.”

In the paper

“Mean harvest rate had been ∼6% before wolf reintroduction.For the first decade following wolf reintroduction, harvest was managed in a manner that led harvest rate to increase to an average of 9% (as high as 17%).”

In the paper

“For the first 13 years following wolf reintroduction (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008), the average predation rate was 6.7%, and most recently (2005)(2006)(2007)(2008)(2009)(2010), the annual predation rate by wolves was <12% in all but one year (Figure 1b).”

In the paper

“From 2012 to 2014, predation rate dropped to its lowest observed levels (mean = 2.7% ± 0.3%).In this three-year period, moose abundance doubled (see below).”

In the paper

“On average, the plots exhibited no change between 1991 and 2006 in their density of recruitment stems.Individual watersheds, however, exhibited considerable variation, with 42% showing significant increases, 37% exhibiting significant decreases, and 21% exhibiting no significant change.”

What the sources say

• ?Not checked — no full text of the sourcecites Kimble et al. 2011 source

In the paper

“Shorter fir trees exhibited increased growth, and by 2013, many were approaching ∼3 m, a height at which fir begin to produce seeds and escape moose herbivory by growing into the canopy (Supplemental Sidebar 2) (see also Vucetich & Peterson 2014).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Vucetich & Peterson 2014

In the paper

“In the eastern portion, fir is present in every size class from seedling to canopy tree and browsing pressure is minimal (except during the early 1990s; see below).”

In the paper

“A year later, a wolf immigrated from the mainland.”

In the paper

“The moose population collapsed in 1996 with the most severe winter on record.”

In the paper

“During 2012-2014, predation rate dropped to its lowest observed level (<2.7%), moose abundance doubled (Figure 3), and the abundance of beaver colonies increased by ∼60% to a level where in 2012 approximately half of the >300 sites occupied historically by beavers were again active (Peterson & Romanski 2012;M. Romanski, personal communication).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Peterson & Romanski 2012

In the paper

“Prior to 2003, predation rate by NR wolves was low largely because the wolf-to-elk ratio was low.From some time around 2003 to 2011, wolf predation was higher, in large part because the ratio of wolves to elk was higher (Figure 1b).”

In the paper

“An estimated slope that is significantly less than -1 could indicate undetected synergy between predation and winter severity or may result from population growth that is influenced by the reduced foraging efficiency of moose during years of high predation risk.”

In the paper

“In the NR, the density of grizzly bears rose and may have tripled between the early 1980s and late 1990s (Hamlin et al. 2009), after which their density has remained high.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Hamlin et al. 2009

In the paper

“The theoretical origins of trophic cascades suggest they are communitywide phenomena involving changes in the biomass of primary producers, in response to changes in herbivore biomass due, in turn, to changes in predator biomass (e.g., Hairston et al. 1960, Oksanen et al. 1981, Fretwell 1987).”

What the sources say

• ?Not checked — no full text of the sourcecites Hairston et al. 1960 source
• ?Not checked — no full text of the sourcecites Oksanen et al. 1981 source
• ?Not checked — no full text of the sourcecites Fretwell 1987 source

In the paper

“In the decade following wolf reintroduction, drought and recreational harvest of elk outside the Park boundaries almost certainly influenced the abundance of NR elk.Those factors also appear capable of explaining the observed decline in elk abundance without invoking the influence of wolf predation (Vucetich et al. 2005; see also Eberhardt et al. 2007).”

What the sources say

• ?Not checked — no full text of the sourcecites Vucetich et al. 2005 source
• ?Not checked — no full text of the sourcecites Eberhardt et al. 2007 source

In the paper

“Such conceptualization is tantamount to equating a tropic cascade with nearly any indirect ecological effect of predation.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Silliman & Angelini 2012

In the paper

“Cougar abundance in the NR also increased during the 1990s from 18 to 32 individuals (Ruth et al. 2014).Nevertheless, little is known about the influence of cougar predation on elk dynamics (Ruth & Murphy 2010); it could be minor or it could be significant.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ruth et al. 2014
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ruth & Murphy 2010

In the paper

“The existence of a trophic cascade in IRNP and NR also depends greatly on how it is defined.”

In the paper

“For the period ending around 2005 (i.e., before the cessation of significant rates of harvest by humans), elk decline would have likely occurred in the absence of wolves.Wolves are also unlikely to have had a significant demographic influence on elk during the decade that followed their reintroduction, given the low rates of wolf predation during that period.”

In the paper

“By that broad conceptualization, detecting the existence of a trophic cascade would be far less illuminating than characterizing its nature, i.e., the extent and strength of indirect effects.”

In the paper

“Questions of whether trophic cascades occur in Isle Royale National Park (IRNP) or Yellowstone National Park's northern range (NR) cannot lead to simple, precise, or definitive answers.Such answers are limited especially by multicausality in the NR and by complex temporal variation in IRNP.”

In the paper

“In Isle Royale, wolves are the only predator and primary cause of death for moose, which represent ∼90% of wolves' diet (Peterson 1977).Humans do not hunt wolves or moose (for additional background, see Vucetich & Peterson 2004a,b).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Peterson 1977

In the paper

“Overall, the decline of NR elk is the result of multiple causes, including harvest, drought, bears, wolves, and, probably, cougars.”

In the paper

“Increased bear abundance in the NR (and a concomitant increase in calf mortality) is the result of grizzly bear recovery efforts mandated by the US Endangered Species Act (Schwartz et al. 2006, Interag. Grizzly Bear Study Team 2013).”

What the sources say

• ?Not checked — no full text of the sourcecites Schwartz et al. 2006 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Interag. Grizzly Bear Study Team 2013

In the paper

“The causes of elk decline have shifted from harvest to predation, first by grizzly bears and cougars, then by wolves.This shift has also been corroborated by analyses of the causes of mortality for radio-collared elk (Evans et al. 2006, Hamlin et al. 2009).”

What the sources say

• ?Not checked — no full text of the sourcecites Evans et al. 2006 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Hamlin et al. 2009

In the paper

“Recent estimates indicate the NR is inhabited by 54-69 grizzly bears per 1,000 km 2 (Barber-Meyer et al. 2008, Hamlin et al. 2009), which is comparable to the average density of wolves in the NR during 2004-2013.”

What the sources say

• ?Not checked — no full text of the sourcecites Barber-Meyer et al. 2008 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Hamlin et al. 2009

In the paper

“This study suggested that the rate of consumption (but not overall productivity), as well as net nitrogen mineralization, had declined.”

What the sources say

• ?Not checked — no full text of the sourcecites Frank 2008 source

In the paper

“Because reports of aspen recovery derive from four watersheds that were selected a priori because they had shown growth (Ripple & Beschta 2007), not unlike 42% of the watersheds exhibiting growth just outside the park boundary, a concern is that a species-specific trophic cascade may require an observation of widespread growth.By contrast, a number of these reports may represent only localized”

What the sources say

• ?Not checked — no full text of the sourcecites Ripple & Beschta 2007 source
• ?Not checked — no full text of the sourcecites Kauffman et al. 2010 source
• ?Not checked — no full text of the sourcecites Winnie 2012 source

In the paper

“If, however, an aspen trophic cascade is an indirect effect of predation on some aspect of aspen demography that limits the maintenance of aspen clones, then some of the evidence indicating an aspen trophic cascade in the NR may be viewed with circumspection.”

In the paper

“Nevertheless, a concern with attributing this growth to reduced aspen herbivory is that it coincided with the alleviation of a multiyear drought, which may have contributed to increased growth (Figure 1c).”

In the paper

“The multicausal nature of aspen demography has also disqualified simple statements about the causes of aspen demography anywhere in western North America (e.g., Sankey 2012, Kulakowski et al. 2013), including the Greater Yellowstone Ecosystem (Hessl 2002, Brown et al. 2006; see also Eisenberg et al. 2013, Seager et al. 2013).”

What the sources say

• ?Not checked — no full text of the sourcecites Sankey 2012 source
• ?Not checked — no full text of the sourcecites Kulakowski et al. 2013 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Hessl 2002

In the paper

“Both patterns could have been the result of a decline in elk preference for aspen as sapling height increased.”

In the paper

“Thus, observed growth could be attributable to another factor, with reduced herbivory a subsequent consequence and not the ultimate cause of growth.”

In the paper

“Nevertheless, only one study has assessed changes in grassland communities since wolves were reintroduced, reporting data only during 1995-2000 when the level of wolf predation was relatively insignificant (Frank 2008).”

What the sources say

• ?Not checked — no full text of the sourcecites Frank 2008 source

In the paper

“Insofar as aspen demography is manifest on a relatively slow timescale, an evaluation of the indirect effects of wolves on aspen may be impossible for some time.”

In the paper

“One concern with attributing the growth of NR aspen to reduced herbivory is that elk abundance had previously declined below approximately 5,000 for an 8-year period in the 1960s (Eberhardt et al. 2007).Yet, this period of low abundance was not low enough or long enough to stimulate significant aspen growth.As such, significant aspen growth should not be expected until after elk have declined to”

What the sources say

• ?Not checked — no full text of the sourcecites Eberhardt et al. 2007 source

In the paper

“Localized recovery of aspen in the presence of relatively high elk density might be evidence of a behaviorally mediated trophic cascade (BMTC) (e.g., Beschta & Ripple 2012, 2013).However, spatial heterogeneity in aspen demography may not be, by itself, convincing evidence for a BMTC because aspen demography is spatially heterogeneous even when risk-sensitive foraging by elk is not implicated (e.g.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Beschta & Ripple 2012
• ?Not checked — no full text of the sourcecites Brown et al. 2006 source

In the paper

“However, this criticism is also misplaced for failing to appreciate that no aspen (or willow or cottonwood) had grown beyond a height of ∼1 m in the NR for decades, with the exception of aspen that had been inaccessible to elk, such as those growing in exclosures or on steep talus (see references cited in Beschta 2005).Yet, by 2006, some aspen saplings living where elk could browse had grown to ∼1”

What the sources say

• ?Not checked — no full text of the sourcecites Beschta 2005 source
• ?Not checked — no full text of the sourcecites Ripple & Beschta 2012 source

In the paper

“Further complicating the answer, the demography of aspen involves many components (i.e., growth, density, and survival of stems across a range of size classes).Additionally, multiple factors influence each component, one effect of which is to create remarkable spatial heterogeneity.”

In the paper

“Aspen is one of several NR vegetative communities reported to have been involved in a trophic cascade (Ripple et al. 2001, Ripple & Beschta 2007, Ripple et al. 2013).Those reports have been celebrated as well as criticized (e.g., Kauffman et al. 2010, Eisenberg et al. 2013), in particular, for being an artifact of biased sampling (Winnie 2012, Kauffman et al. 2013).”

What the sources say

• ?Not checked — no full text of the sourcecites Ripple et al. 2001 source
• ?Not checked — no full text of the sourcecites Ripple & Beschta 2007 source
• ?Not checked — no full text of the sourcecites Winnie 2012 source

In the paper

“Wolf predation has also influenced spatiotemporal patterns in the distribution of wolf-killed moose carcasses, which creates small-scale spatial heterogeneity in available soil nutrients, with belowground effects on soil microbial communities and aboveground feedbacks that change leaf tissue quality (Bump et al. 2009).”

What the sources say

• ?Not checked — no full text of the sourcecites Bump et al. 2009 source

In the paper

“Although herbivory has been a key influence, other factors have also almost certainly been important.”

In the paper

“Drought would, for example, qualify.”

In the paper

“Each new period appears to have been triggered by some exogenous random event, including introduced disease, a catastrophically severe winter, and an unlikely immigration event that mitigated inbreeding depression among wolves.”

In the paper

“Nevertheless, top-down influences are also reflected in both large-scale spatial variation (Figure 2) and fine-scale temporal variation within single study sites (Figure 5b,c).”

In the paper

“Although a trophic cascade is a coherent concept in such a world, the greatest intellectual contribution of asking”

In the paper

“Those circumstances give reason to recall Bayes' theorem: Given the observed data, the probability a trophic cascade (or any event) has occurred is proportional to one's prior belief about the probability that a trophic cascade has occurred.”

In the paper

“The richer and warmer soils of the western portion favor deciduous growth (Linn 1957), and the thin soils of the eastern portion result in more frequent wind throws, which create canopy gaps and seedling substrates (rotten logs) that favor the growth of shade-intolerant fir (McLaren & Janke 1996).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Linn 1957
• ?Not checked — no full text of the sourcecites McLaren & Janke 1996 source

In the paper

“Too often, this discourse seems to imply that the appropriateness of conserving a top predator depends critically on demonstrating that its decline or recovery is the cause of a trophic cascade (however such may be defined).”

In the paper

“The logic of hypothesis testing (frequentism) is analogous: One's belief in a trophic cascade depends greatly on whether one is willing to believe so until proven otherwise, or unwilling to believe until definitively proven.”

In the paper

“Questions about trophic cascades have led to great insight about a much more fundamental question,”

In the paper

“Without satisfying answers to such questions, the standards for assessing trophic cascades will remain elusive.”

In the paper

“By that logic, any process that kills herbivores with ecological impacts counts as predation.”

What the sources say

• ?Not checked — no full text of the sourcecites Ripple et al. (2014) source

In the paper

“Inbreeding depression among wolves caused the impaired predation, and canine parvovirus, which reappeared in 2007 for the first time in 17 years, may have exacerbated it (see Supplemental Sidebar 4).”

In the paper

“By contrast, most ecosystems are uncontrolled and entail complicated, multicausal food webs governed by nonequilibrium dynamics (Botkin 2012).”

What the sources say

• ?Not checked — no full text of the sourcecites Botkin 2012 source

In the paper

“Pleistocene glaciers deposited till that contributed to deep soils in the western portion of Isle Royale but scoured the eastern portion, leaving shallow soils (typically <30 cm) and frequent bedrock exposure (Huber 1973).”

What the sources say

• ?Not checked — no full text of the sourcecites Huber 1973 source

In the paper

“When data are difficult to gather and interpret, the influence of one's a priori disposition is strong and unavoidable.”

In the paper

“Evaluating trophic cascades in large, complicated ecosystems requires data that are difficult to collect and inferences about causality that are difficult to make reliably.”

In the paper

“Judging where trophic cascades have occurred in a multicausal world, dominated by human influence, also depends greatly on what constitutes a trophic cascade.”

In the paper

“Being too permissive about what counts as predation reduces questions about trophic cascades merely to questions about the importance of herbivory.”

In the paper

“A trophic cascade, especially as the idea is associated with Hairston et al. (1960) and Oksanen et al. (1981), is an intellectual construct born in an imaginary world of simple food chains governed by equilibrium dynamics.”

What the sources say

• ?Not checked — no full text of the sourcecites Hairston et al. (1960) source
• ?Not checked — no full text of the sourcecites Oksanen et al. (1981) source

In the paper

“The answer is also a weapon in the social discourse on the value of conserving top predators.”

In the paper

“Spatial heterogeneity (McCann 2011) and contingency (Schmitz 2010) further complicate ecosystems.”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites McCann 2011
• ?Not checked — no full text of the sourcecites Schmitz 2010 source

In the paper

“The NPS recently announced its decision not to conduct genetic rescue (assisted immigration) in the short term, although doing so would have likely mitigated the effects of inbreeding depression.”

In the paper

“In those regions, moose browsing has prevented fir from growing into the canopy for the past century, and many short firs have a morphology reflecting decades of severe browsing ( Janke et al. 1978, McLaren 1996).”

What the sources say

• ?Not checked — no full text of the sourcecites Janke et al. 1978 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites McLaren 1996

In the paper

“Climate warming is implicated because ice bridges form far less frequently and now appear to have been critical for periodic gene flow throughout the history of IRNP wolves (Hedrick et al. 2014).”

What the sources say

• ?Not checked — no full text of the sourcecites Hedrick et al. 2014 source

In the paper

“Some short balsam firs in the western portion could have grown into the canopy in the late 1980s and early 1990s, were it not for the dramatic increase in moose abundance that resulted from the loss of wolves caused by a human-introduced disease (Figure 3), whose effect was likely exacerbated by inbreeding depression (Peterson et al. 1998).”

What the sources say

• ?Not checked — no full text of the sourcecites Peterson et al. 1998 source

In the paper

“Drought involves, essentially, no feedback; predation and harvest differ profoundly in the nature of their feedback.”