Research articleSpecial delivery: scavengers direct seed dispersal towards ungulate carcasses
Abstract
Cadaver decomposition islands around animal carcasses can facilitate establishment of various plant life. Facultative scavengers have great potential for endozoochory, and often aggregate around carcasses. Hence, they may disperse plant seeds that they ingest across the landscape towards cadaver decomposition islands. Here, we demonstrate this novel mechanism along a gradient of wild tundra reindeer carcasses. First, we show that the spatial distribution of scavenger faeces (birds and foxes) was concentrated around carcasses. Second, faeces of the predominant scavengers (corvids) commonly contained viable seeds of crowberry, a keystone species of the alpine tundra with predominantly vegetative reproduction. We suggest that cadaver decomposition islands function as endpoints for directed endozoochory by scavengers. Such a mechanism could be especially beneficial for species that rely on small-scale disturbances in soil and vegetation, such as several Nordic berry-producing species with cryptic generative reproduction.
1. Introduction
Animal carcasses are inherent parts of the natural environment, yet their functional role in ecosystem processes has long been underappreciated in ecology [1,2]. However, recent studies have shown that carcasses can have substantial impacts on ecosystems by affecting soil biogeochemistry and community dynamics of microbiota, vascular plants, invertebrates and vertebrates [3]. Such impacts may emerge at local and landscape scales (e.g. nutrient redistribution) through various direct and indirect interactions among several functional groups and kingdoms [2,4].
During decomposition of larger terrestrial animals such as many ungulates, plant life in the immediate vicinity of a carcass turns chlorotic and dies from abrupt shifts in soil nutrient concentrations and acidity, turning the affected patch into a ‘cadaver decomposition island’ [2,3]. These nutrient-rich, bare patches can subsequently facilitate recruitment for various plant life that could otherwise not establish [5] and support vigorous plant growth during secondary succession [6]. Carcass decomposition can eventually modulate local vegetation community structure and diversity [6,7].
Many vertebrates are facultative scavengers [4], implying that they have generalist diets which can include carrion, vertebrate and invertebrate prey, and plant material. Such diverse diets give them great potential to function as dispersal vectors for endozoochory (i.e. plant seed dispersal after ingestion by animals) for seeds that can be ingested across their entire home range. Scavengers often concentrate their activity around carcasses, which likely results in spatial concentrations of their faeces, and thus the potential for endozoochorous plant seed deposition. This implies that scavengers can provide a mechanism for directed endozoochory towards cadaver decomposition islands.
Here, we assess the potential of cadaver decomposition islands as endpoints of endozoochorous plant seed dispersal by scavengers. We expected (H1) that the spatial distribution of scavenger faeces was positively correlated to a density gradient of reindeer (Rangifer tarandus) carcasses. We also expected (H2) that scavenger faeces at cadaver decomposition islands regularly contained viable plant seeds, using crowberry (Empetrum nigrum), an alpine keystone species with cryptic generative reproduction, as a focal species.
2. Material and methods
(a) Study site
Our study site was located at 1220 m.a.s.l. in an alpine tundra ecosystem at the Hardangervidda plateau in southeast Norway, where lightning killed nearly an entire herd (N = 323) of wild tundra reindeer on 26 August 2016. The carcasses are distributed over an area of about 240 × 100 m, with their highest concentration in approximately 50 × 50 m (figure 1a). The authorities removed all carcass heads to screen for chronic wasting disease but left the retaining carcass biomass at the site. The plant community is relatively species poor. The field layer is dominated by dwarf birch (Betula nana), ericaceous species and graminoids, and the ground layer is dominated by mosses and lichens (electronic supplementary material, table S1).
Figure 1. (a) Delineation of the study site in Hardangervidda, Norway. Grey and black dots represent reindeer carcasses and 1 × 1 m survey plots, respectively. The predicted number of mesopredator (b) and bird (c) faeces was positively related to carcass density and/or proximity. (d) Rodent faeces occurred more frequently further away from the carcass concentration. ‘H’ and ‘L’ represent high and low scat detection probability/number, respectively. (Online version in colour.)
The community of scavengers and carcass users includes corvids (e.g. raven Corvus corax and hooded crow C. cornix), golden eagle (Aquila chrysaetos), foxes (red fox and the arctic fox V. lagopus), wolverine (Gulo gulo) and several rodents (e.g. Arvicolidae). We observed several bird species foraging for blowflies and their larvae at the carcasses (e.g. the meadow pipit Anthus pratensis).
(b) Study design
We spaced 75 1 × 1 m survey plots in a semi-regular 10 × 10 m grid covering the study site (figure 1a). We divided each plot into four 50 × 50 cm subplots. Two observers systematically searched each subplot for 30 s (i.e. 4 min total search time per plot) and registered the number of (i) mesopredator faeces (i.e. red and arctic fox), (ii) bird faeces, and (iii) the presence of rodent faecal pellet groups. We used ArcGIS 10.4 to generate carcass density layers and a layer comprising the distance to the nearest carcass for each 1 × 1 m pixel in the study area. We generated carcass density kernels with different search radii (i.e. 1–10, 15, 20, 30, 40, 50, 100 and 200 m). These different radii allowed us to evaluate the most appropriate spatial scale for faecal deposition rates of each scavenger group.
(c) Statistical approach
We evaluated the relationship between the faecal count of mesopredators and birds at each plot in relation to carcass distance and density, using generalized linear models (GLMs) with a Poisson error structure. We used GLMs with a binomial error structure to assess the relationship between the occurrence of rodent faeces and carcass distance and density. We considered distance to the nearest carcass and carcass density with single, additive or interactive effects as potential explanatory variables. Because distance to the nearest carcass and carcass densities based on search radii more than or equal to 40 m were collinear (Pearson correlation coefficients |r| ≥ 0.6), these variables were not combined as explanatory variables in the same candidate models (electronic supplementary material, table S2). We accounted for spatial autocorrelation by including an autocovariate (a distance-weighted function of neighbouring response values, electronic supplementary material, S2) as an explanatory variable in all models [8]. We selected the most parsimonious model based on the Akaike information criterion corrected for small sample sizes (AICc). We considered the candidate model with the lowest AICc value and least degrees of freedom within an AICc range of 0–2 as most parsimonious [9]. We used R 3.4.0 [10] for all statistical analyses.
(d) Viability of crowberry seed in scavenger faeces
To assess the potential for directed endozoochory, we collected corvid faeces that contained crowberry seed at cadaver decomposition islands. We chose crowberry as a focal species, as it has easily recognizable seeds, cryptic sexual reproduction and functions a keystone species of the alpine tundra [11]. All faecal samples were air-dried at the time of collection. We extracted 5–20 crowberry seeds per sample for viability testing. The proportion of extracted seeds never exceeded 50%, as we retained material for parallel projects. We used a standard tetrazolium test procedure to assess seed viability [12].
3. Results
We registered 22 and 87 mesopredator and bird faeces at the 75 survey plots, respectively, and detected faecal pellet groups of rodents at 37 plots. The number of mesopredator scats detected at survey plots was positively related to carcass density (search radius 50 m; β = 56.835, se = 14.267, p-value < 0.001) (figure 1b; electronic supplementary material, tables S3 and S4). Counts of bird faeces were in strong, positive association with carcass density (search radius 7 m; β = 4.653, s.e. = 1.450, p-value = 0.001) and proximity (β = −0.144, s.e. = 0.049, p-value = 0.003) (figure 1c; electronic supplementary material, tables S3 and S4). The probability of detecting rodent pellets was negatively related to carcass density (search radius 50 m, β = −32.601, s.e. = 14.324, p-value = 0.023) (figure 1d; electronic supplementary material, tables S3 and S4). The autocovariates did not have significant effects on the response variables (p-values > 0.101, electronic supplementary material, table S4). We found viable crowberry seeds in 21 out of 24 (87.5%) faecal samples of corvids. The proportion of viable crowberry seed per sample averaged 0.385 (electronic supplementary material, table S5).
4. Discussion
Our results indicate that cadaver decomposition islands can act as endpoints for directed endozoochory, as faecal deposition rates of birds and mesopredators were closely associated with cadavers, and corvid faeces at the cadaver decomposition islands often contained viable plant seeds. Faecal pellet groups of rodents were negatively associated with carcasses, which could be explained by a scavenger-induced landscape of fear or by the lack of food resources (vegetation) at cadaver decomposition islands.
Our study provides novel insight into how scavengers may have landscape-level effects on plant distribution through directed endozoochory towards cadaver decomposition islands. Directed endozoochory towards cadaver decomposition islands could be especially beneficial for plant species that take advantage of small-scaled disturbances for successful generative reproduction, such as many clonal species of the boreal and subarctic biomes [13]. In addition to potentially affecting the distribution of dispersed plants, the role of scavengers may have landscape-wide consequences for the genetic structure of dispersed plant species, as seeds are likely ingested across their dispersers' home ranges [14]. To what extent, however, such seeds germinate and establish under field conditions remains to be tested. Our study took advantage of a rare event, an ungulate mass die-off, to illustrate directed endozoochory by scavengers towards cadaver decomposition islands. It remains unexplored, however, how this mechanism operates at cadaver decomposition islands from animals that die from more common causes (e.g. hunting, predation), and that are less concentrated in space and time [15].
As anthropogenic impacts on ecosystems reach unprecedented levels [16], humans likely disrupt the mechanism of directed endozoochorous dispersal by scavengers towards cadaver decomposition islands by modulating the availability and the spatio-temporal distribution of carcasses. For example, policies for disease prevention may require full carcass removal (e.g. free-ranging livestock) from their natural place of death for disposal in authorized facilities [17], which also implies the removal of their ecological functions. Negative human attitudes towards scavengers and large carnivores can result in severe population reductions or even extirpation [18], which likely inhibits their roles as seed dispersers (i.e. scavengers) or as carcass providers (i.e. large carnivores). In addition, anthropogenic hunting is an important provider of animal carcasses, but deviates in space and time from other causes of wildlife mortality [15]. This implies that hunting can modulate the spatio-temporal distribution of endpoints for directed endozoochory by scavengers. We suggest that further elucidating this mechanism will provide novel insight into the regeneration and genetic structure of certain plant species and communities.
Ethics
Landowner association ‘Sameiet lisetfjell’ and Telemark County provided fieldwork permission.
Data accessibility
Data used in this study are deposited in the Dryad Digital Repository (http://dx.doi.org/10.5061/dryad.h3c55cc) [19].
Authors' contribution
S.M.J.G.S. and S.C.F. conceived the study. S.M.J.G.S., S.C.F., R.B. interpreted the data and drafted the article. All authors assisted with fieldwork and data collection, provided valuable comments to earlier drafts, approved the manuscript, and agreed to be accountable for its content.
Competing interests
The authors have no competing interests.
Funding
The REINCAR project is currently self-funded.
Acknowledgements
REINCAR is hosted at the LEBE! lab at the University of South-Eastern Norway. We thank the Norwegian Environment Agency for field guidance. Seed viability was tested at the ISTA-accredited Kimen Seed Laboratory.
