Thieving rodents as substitute dispersers of megafaunal seeds
Edited* by John W. Terborgh, Duke University, Durham, NC, and approved June 14, 2012 (received for review April 3, 2012)
Abstract
The Neotropics have many plant species that seem to be adapted for seed dispersal by megafauna that went extinct in the late Pleistocene. Given the crucial importance of seed dispersal for plant persistence, it remains a mystery how these plants have survived more than 10,000 y without their mutualist dispersers. Here we present support for the hypothesis that secondary seed dispersal by scatter-hoarding rodents has facilitated the persistence of these large-seeded species. We used miniature radio transmitters to track the dispersal of reputedly megafaunal seeds by Central American agoutis, which scatter-hoard seeds in shallow caches in the soil throughout the forest. We found that seeds were initially cached at mostly short distances and then quickly dug up again. However, rather than eating the recovered seeds, agoutis continued to move and recache the seeds, up to 36 times. Agoutis dispersed an estimated 35% of seeds for >100 m. An estimated 14% of the cached seeds survived to the next year, when a new fruit crop became available to the rodents. Serial video-monitoring of cached seeds revealed that the stepwise dispersal was caused by agoutis repeatedly stealing and recaching each other’s buried seeds. Although previous studies suggest that rodents are poor dispersers, we demonstrate that communities of rodents can in fact provide highly effective long-distance seed dispersal. Our findings suggest that thieving scatter-hoarding rodents could substitute for extinct megafaunal seed dispersers of tropical large-seeded trees.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
The Neotropics are rich in woody plant species with large-seeded fruits that are structurally similar to fruits of trees dispersed by megafauna (mammals >1,000 kg) in the Paleotropics (1, 2). Guimarães et al. (2) operationally defined these “megafaunal” fruits as oversized (4–10 cm in diameter) fleshy fruits that contain one to five extremely large seeds. Because such fruits and seeds are typically too large to be swallowed by extant wildlife (1, 2), Janzen and Martin (1) argued that these plants are in fact adapted to seed dispersal by Pleistocene megafauna, which ingested entire fruits and defecated the seeds intact away from the source. This dispersal mutualism would have been disrupted at the end of the Pleistocene, when nearly all megafauna of the Neotropics became extinct (3). Whether these plants are really “megafaunal” is controversial, partly because it is unclear how these species could have persisted >10,000 y without their dispersal mutualists (4–7). Seed dispersal is a critical requirement for plants to persist in the face of predators and pests that attack offspring (8–10). Additionally, plants need long-distance seed dispersal to maintain gene flow between populations, colonize new sites, and move their range as climate changes (10, 11). Plant species that fail to disperse are predicted to be outcompeted by others and ultimately to go extinct. The question remains as to how megafaunal species can still be widespread, and often locally common, throughout the Neotropics.
One hypothesized mechanism by which any megafaunal tree species may have persisted is substitute seed dispersal by scatter-hoarding rodents (2, 12–14). These rodents bury seeds as food reserves in shallow caches, each containing one or a few seeds. The buried seeds are protected against invertebrates and sit in ideal conditions to germinate and establish into seedlings if the animals fail to recover them (15). Scatter-hoarding rodents are already known to disperse seeds that they collect from dung piles (12, 16, 17). These rodents could have then become primary dispersers of megafaunal plant species after Pleistocene megafauna disappeared (1), an example of “evolutionary partner switching” by the plant (18).
The substitution hypothesis has met skepticism because rodents generally are not considered effective seed dispersers. First, they are believed to move seeds only over small areas. Documented seed dispersal distances produced by scatter-hoarding rodents are usually small (<25 m) (15), which may be inadequate to help seeds escape the intense competition and pest pressure near parent trees. The maximum distance at which a rodent might disperse seeds would be within an individual’s small home range; therefore, seeds would never travel far enough to colonize new sites. Second, scatter-hoarding rodents are believed to depredate most of the seeds they cache; documented survivorship of seed caches is extremely low, and primary caches almost never persist until a seedling has established (19). Thus, available data would suggest that rodents cannot be substitutes for megafaunal seed dispersers.
However, it is conceivable that the effectiveness of seed dispersal by scatter-hoarding rodents has been underestimated because reliably tracking seed movements and seed fate is notoriously difficult (20). Most seed tracking studies use traditional passive tags that are practical only for tracking seeds over short distances. The inability to track far-dispersed seeds and the categorization of those seeds as “missing” results in a heavy bias against longer dispersal distances (21). That rodents are capable of dispersal over longer distances is demonstrated by the occasional observance of seeds dispersed >100 m (e.g., refs. 22–24). Additionally, although most studies of seed dispersal only consider “primary” movement, the initial movement of seeds from the parent tree, some have shown that seeds recovered from caches are sometimes recached rather than eaten and that sequences of multiple secondary movements occur (23–27). These limitations suggest that the dispersal effectiveness of scatter-hoarding rodents may have been underestimated owing to methodological limitations.
Here we investigated seed dispersal of a reportedly megafaunal tree species by one of the Dasyproctidae, scatter-hoarding rodents that occur throughout the Neotropics. Our study species was Astrocaryum standleyanum (Arecaceae), a cocosoid palm that produces hard stony endocarps with large seeds, enclosed in brightly colored, sweet fruits that typify the megafaunal syndrome of oversized fleshy fruits with one or few extremely large seeds (2). It has been hypothesized that such fruits evolved to attract megafaunal seed dispersers that are now extinct, such as elephant-like gomphotheres (1, 2), which would have ingested the fruits without cracking the hard stones and defecated the seeds intact, as is still seen today in African elephants (28). Today, A. standleyanum palm is primarily dispersed by Central American agoutis (Dasyprocta punctata) that scatter-hoard the seeds across their 2- to 3-ha home ranges as food reserves for the low-fruit season (29). Janzen and Martin (1) suggested that the agouti was “once a trivial dispersal agent [that] figured primarily as seed predator” of seeds that it removed from the megafaunal dung piles, but then took the role as the principal seed disperser after the extinction of the Pleistocene megafauna.
Our goal was to obtain unbiased estimates of seed dispersal distance and survivorship produced by scatter-hoarding rodents to determine how effective they are as seed dispersers. To overcome the problem of bias against long-distance dispersal we affixed motion-sensitive telemetric thread tags (30) to seeds and used manual and automated telemetry (31) to continuously track their movement. We tracked the seeds regardless of how far and how often they were moved, until they were finally eaten or left to germinate. We also assessed the mechanism causing cache dynamics and multistep movement, distinguishing between cache theft and cache recovery by “owners.” For this aspect of our study, we tagged agoutis for individual identification and monitored cached seeds with camera traps.
Results
Camera-trap monitoring of 589 radio-tagged seeds placed at 52 experimental stations across Barro Colorado Island (BCI) showed that seeds were quickly removed (Fig. S1A). Seeds were almost exclusively removed by scatter-hoarding rodents, and mostly by agoutis (83% of 423 removed seeds). Radio-tracking allowed us to retrieve nearly all (97%) of the removed seeds and record their fate, regardless of how far they traveled. The majority (85%) of the 409 seeds that we retrieved were buried in caches, and only 3% were eaten. Agoutis cached 88% of the seeds they removed. The median initial dispersal distance of seeds removed from the stations was just 8.75 m, which is similar to prior studies of seed dispersal by rodents, yet the tail of the distribution ranged as far as 159 m (Fig. 1A). Studies using traditional tagging methods such as thread tags would typically have missed the one fifth of the seeds that were carried >25 m (21).
Fig. 1.
The survivorship of 224 first-order agouti-made caches that we monitored was short. Most seeds (57%) were recovered by an animal within 1 wk after caching, and 99% of the caches were ultimately recovered (Fig. 1B). However, only 13% of the seeds removed from the primary seed cache were actually eaten; the majority were recached. Moreover, most seeds were dug up and recached multiple times (median eight times; Fig. S1B), sometimes twice in a single day. This repeated recaching produced stepwise movement along paths of up to 892 m long [median 131 m (confidence interval [CI] 100–188 m)]; Fig. S1C) and gradually dispersed seeds farther away from the point of release [median 68 m (CI 51–86); Figs. 1A and 2 and Movie S1]. One seed was cached as many as 36 times, traveled more than 749 m, and ended 280 m from its starting point, when it was dug up and eaten by an agouti 209 d after initial dispersal. This recaching behavior resulted in an estimated 87% of seeds moving past the immediate vicinity of the parent tree (<15 m), the primary zone of Janzen-Connell effects. An estimated 35% of seeds attained a net dispersal distance >100 m, which is often used as a threshold for long-distance seed dispersal (32–34). Many seed movements extended well beyond the size of the typical agouti home range, suggesting the involvement of other individuals besides the initial cache owner.
Fig. 2.
Remote cameras placed at a subset of the agouti-made caches revealed that most cache recovery was by conspecifics [94%, vs. 4% by spiny rat (Proechimys semispinosus), 1% by squirrel (Sciurus granatensis), and 1% by land crab Potamocarcinus richmondi; n = 134]. For 116 agouti-made caches, we were able to distinguish whether agoutis recovering the seed were the owner or a thief, because 16 agoutis in the study area were individually recognizable by tags or marks. We found that just 16% of the seeds were recovered by their owner; the majority (84%) were stolen by other agoutis. Theft was strongly reciprocal: individuals that were robbed also stole cached seeds from others (Fig. 3). Thus, the stepwise dispersal of seeds across home range boundaries was driven by reciprocal theft.
Fig. 3.
Because seeds recovered from caches were usually recached rather than eaten, seeds ultimately survived more than 10 times longer than caches, on average [half-life, 45 d (CI 35–83 d) vs. 3.9 d; Fig. 1B]. Although the majority of seeds (58%) were found eaten or moved into tree cavities, an estimated 14% survived buried underground to the next year (CI 9–22%). This estimation is based on 4% of the seeds surviving, accounting for 38% censored observations due to severed tags (72) or failing transmitters (14), the ultimate fate of which remained unknown. By that time, a next generation of fresh seed crops had ripened, drawing the attention of the rodents away from the older caches and potentially allowing the seeds to germinate and establish into seedlings.
Discussion
Our results indicate that rodents may disperse seeds over much longer distances than was ever anticipated, and with much higher seed survivorship. Although initial dispersal distances were mostly short, repeated movements resulted in seeds being dispersed outside the immediate vicinity of the parent, the primary zone of Janzen-Connell effects (35). One third of the seeds attained a net dispersal distance >100 m, the often-used threshold for long-distance seed dispersal (32–34), which is far enough to colonize new habitats and shift ranges over generations in the tropics, where climatic shifts are much smaller than in the temperate zone. Finally, one seventh of the seeds survived to the next year.
The extremely fast recovery of caches found in this study is typical for rodent-mediated dispersal (19). Similar low cache survivorship in previous studies has fueled the argument that scatter-hoarding rodents are essentially seed predators rather than seed dispersers (36). Our study, however, demonstrates that low cache survivorship is not equivalent to seed survivorship. In fact, most seeds dug up by animals from caches were recached, usually at further dispersal distances, and potentially in more suitable environments. Although previous studies had found high proportions of recaching (23–27), no study had been able to estimate ultimate survivorship because traditional tracking methods were biased against seeds that moved more often or traveled longer distances. Our improved tracking method makes our estimate of 14% survival of cached seeds to the next year robust, regardless of the extent of movements.
The multistep dispersal and massive reciprocal cache robbery we discovered resulted in scatter-hoarding agoutis moving seeds much farther than expected. Home range sizes of agoutis are small relative to that of many other vertebrate seed dispersers (37), but robbing agoutis gradually moved seeds across home range boundaries. The ultimate dispersal distances were far enough for most seeds to escape Janzen-Connell effects around parent trees and colonize new sites. Previous studies that tracked secondary seed movement also found multiple movements of individual seeds (23–27), but the maximum number of movements per seed in these studies (five movements) pales in comparison with the median of eight movements per seed in our study (range, 1–36). Some seeds in our study were moved more than once per day. Such detail would have gone unseen if we had censused caches only at regular time intervals, as is typical in seed-tracking studies.
The fact that seeds were frequently moved beyond the typical home-range size of agoutis suggested that multiple individuals were involved. Our unique combination of radio tracking of seeds, individual tagging of agoutis, and camera monitoring of caches revealed that the stepwise dispersal of seeds was driven by reciprocal theft. All monitored agoutis engaged in stealing also lost caches to conspecific theft. The resulting theft rates were high: less than one out of six caches were recovered by the owner. Reciprocity has been put forward as an explanation for the evolutionary persistence of scatter-hoarding behavior in the face of high theft (19). However, never before had reciprocal pilferage been demonstrated so convincingly in wild populations.
Multistep seed dispersal as observed in this study is driven by competition for food among conspecifics. Rates of cache pilferage, as well as distances over which scatter-hoarding rodents move seeds, are both known to increase with seed scarcity, which can be calculated as the ratio of rodents to seeds (23, 38, 39). Although our study was conducted during a year of average fruit abundance, the extent of multistep seed dispersal observed could have been influenced by the high density of agoutis on BCI. In years of high fruit abundance or in locations with low agouti abundance due to hunting, pilferage rates and dispersal distances are predicted to be smaller. This relationship with abundance implies that hunting may affect seed dispersal by agoutis even more than believed (40), as seeds are predicted to be moved less often and over shorter distances.
Dispersing seeds are notoriously difficult to track (20), and the low estimates of seed dispersal distances and survivorship that dominate the literature on seed dispersal by rodents are likely an artifact of using tagging systems that cannot reliably record the fate of seeds that are moved multiple times and over longer distances. We overcame these limitations by tracking seeds with telemetric thread tags (30). Radio tags have previously been used to track primary dispersal of Quercus acorns by jays in Spain (41) and by mice in Japan (42), and dispersal of Juglans walnuts by squirrels and mice in Japan (43, 44), yielding dispersal distances up to 168 m (squirrels) and 550 m (for jays). However, these studies relied on prescheduled physical censuses for monitoring seed location and fate and had limited duration due to battery constraints. We introduced motion triggering, so that tags started transmitting only when the seeds were moved, (30) and an automated radio telemetry system (31) to continuously listen for transmitter signals of any seeds that moved. This combination of techniques allowed us to record seed fate pathways in great detail and for a full year, despite the battery limitations inherent to miniature transmitters.
Conclusion
Our findings demonstrate that scatter-hoarding rodents can be effective seed dispersers. Communities of stealing rodents can provide long-distance dispersal comparable to other vertebrates that are considered good seed dispersers (45, 46). Although this seed movement is not as far as provided by large-bodied frugivores, such as megafauna that can transport seeds over multiple kilometers (47, 48), the survival potential of scatter-hoarded seeds is likely superior to that of hundreds of seeds deposited together in megafauna dung piles (48–50). Our results make it conceivable that secondary seed dispersal by rodents can be sufficiently effective to substitute for primary dispersal of megafaunal seeds by large mammals. Because Dasyproctids, such as agoutis and the similarly behaving acouchis, are common throughout the Neotropics, this may explain how many Neotropical tree species that once relied on large megafaunal seed dispersers have persisted in the wake of the Pleistocene extinctions (1). However, the relationship between rodents and large nuts is much older than the Pleistocene, appearing as early as the Late Paleocene (>55 MyBP), which is the same period as when large terrestrial frugivores abounded (51). These plants may have a long history of seed dispersal by rodents, and perhaps never depended on megafauna in the first place.
Materials and Methods
We radio-tracked the movement of megafaunal palm seeds in tropical moist forest of Central Panama to determine the seed dispersal distances and survivorship that scatter-hoarding rodents produced. We individually tagged agoutis and monitored cached seeds with camera traps to assess the role of theft in cache dynamics.
Site and Species.
Field work was carried out on BCI, Panama (9°10′N, 79°51′W), a 1,560-ha island covered with tropical moist forest, administered by the Smithsonian Tropical Research Institute (52). Our study area consisted of 50 ha of late-secondary forest in the central part of BCI. We tracked seeds of A. standleyanum, a large-fruited and large-seeded palm that is presumed to be adapted for seed dispersal by megafauna now extinct (1, 2). A. standleyaum trees produce 10-g stony endocarps with large seeds, enclosed in bright-orange, sweet-fleshy fruits (53). Astrocaryum on BCI has a seasonal fruiting peak during May–June, when each tree produces up to 1,500 fruits. This palm is now dispersed primarily by the Central American agouti (D. punctata), a 2- to 4-kg caviomorph rodent that scatter-hoards large seeds across its 2- to 3-ha home range as food reserves (29, 54). Agoutis feed almost exclusively on these reserves during times of fruit scarcity (54–56).
Radio-Tracking of Seeds.
Ripe Astrocaryum fruits were collected using seed traps suspended below haphazardly selected fruiting trees. Seeds were defleshed using a knife to resemble natural defleshing by rodents (cf. 53), air dried, and given a 55-cm-long “telemetric thread tag” (30) consisting of a 30-cm black nylon-coated stainless steel leader wire (Surflon 1 × 7 black coating; American Fishing Wire) and a 4.1-g cylindrical very high frequency (VHF) transmitter (Advanced Telemetry Systems) with 20-cm antenna. Affixing thread tags to seeds is the widely accepted standard method for tracking seed dispersal by rodents (57), because rodents will bury the seed but leave the thread above ground, allowing researchers to locate the seed.
The telemetric thread tag was attached to the seed by tying the leader wire to a 7-mm screw eye that was inserted in the basal end of the stone (away from the embryo) via a 1.5- to 3-mm drilled hole. A 7-cm piece of pink flagging with a unique code was attached to the wire near the transmitter. Experimental tests indicated that agoutis do not discriminate between telemetric and standard thread tags (30). To prolong battery life, the transmitter was shut down during intervals without movement. We did this by placing the transmitter on a small magnet taped to the head of a 10-cm nail that was pushed into the ground 25 cm away from the seed. The radio was reactivated when a seed was moved, thereby pulling the transmitter off the magnet (30).
A total of 589 radio-tagged seeds were placed at 52 seed stations scattered across the study area during May 16–July 16, 2010. Each seed station contained five seeds arrayed in a line with 3-cm interspacing. We checked each station daily. After 8 d, or when all seeds were removed from the station, we either replaced the seeds or discontinued placing seeds at that location. Stations received 5–25 seeds in total (mean ± SD, 11.3 ± 5.9). Removed seeds were located by sight if possible, or otherwise with hand-held radio-telemetry equipment (Yaesu-VR500), to record their fate and movement distance. If the seed was found within 20 m of the seed station, the dispersal distance was measured with measuring tape, and the direction was recorded using a precision compass (Suunto KB-14). If the seed moved >20 m, the new location was recorded using a global positioning system receiver (Garmin 60CSx). Seed removal from all stations was also monitored with remote cameras (RC55 or PC800; Reconyx) that took a continuous series of photographs at 1 s−1 when animals triggered the motion sensor. These cameras allowed us to determine which species removed the seeds and the exact date and time of removal (cf. 58).
We monitored 224 radio-tagged cached seeds during 12 mo by checking for radio signals every day in the first 6 mo and at least once per month thereafter, which yielded a total of 1,453 movements. Ultimately, 129 seeds (58%) were found eaten. For 86 seeds (24%), movement paths and fates were censored because rodents cut off the tag (4.8% per movement) or because the radio failed to initiate (1.1% per movement). Tag cutting typically happens when agoutis carry out their routine of meticulously cleaning seeds before caching (53). Nine cached seeds were still alive and being monitored at the last census (May 2011).
Animal Tagging.
In our study area, a total of 16 agoutis were individually recognizable. These agoutis were captured with live traps (Tomahawk Live Trap) that were baited with banana and checked twice daily. Adult individuals (>2.3 kg) (54) were fitted with a VHF radio transmitter that had a unique pattern of reflective tape affixed to the collar. Subadult animals were individually marked with freeze brands (59). The trapping and marking procedures were approved by the Institutional Animal Care and Use Committee of the Smithsonian Tropical Research Institute and conducted under research permits authorized by the Barro Colorado Nature Monument.
Video Surveillance of Cache Theft.
To determine whether cached seeds were recovered by owners or thieves, we monitored 154 agouti-made caches (multiple caches for each of 59 seeds) with remote cameras. Footage allowed us to identify the species for 134 cached seeds that were retrieved on camera. For 116 of 126 caches retrieved by agoutis, we could determine whether seed removal was theft, because 16 agoutis were individually recognizable. Reciprocal theft of seeds among individual agoutis was visualized with the R package bipartite (60).
Analyses.
All analyses were conducted in R 2.12.2 (61). Seed removal and cache survival were analyzed as waiting times using Kaplan-Meier survival analysis with the R package survival (62). Seeds that had not experienced the focal event (removal from the station or recovery from a cache) before monitoring ended were included as censored observations. Dispersal distance was also analyzed using Kaplan-Meier survival analysis, essentially treating distance as time (cf. 21, 23, 63). Here, seeds that could no longer be followed because their tag failed or was cut off from the seed were included as observations censored at the last recorded distance. Likewise, the number of movements per seed was analyzed using Kaplan-Meier survival analysis, whereby the number of movements per seed was treated as time, and seeds that could no longer be followed were included as censored observations. Reciprocal theft of seeds among individual agoutis was visualized with the R package bipartite (60).
Data Availability
Data deposition: The data reported in this paper has been deposted in the Movebank database, www.movebank.org (doi: 10.5441/001/1.9t0m888q).
Acknowledgments
We thank Tom Garin (Advanced Telemetry Systems), Daniel Obando, Alejandro Ortega, and Meg Crofoot for technical support; Lieneke Bakker, Reyna Bell, Chris Carson, Matt McElroy, Veronica Pereira, Torrey Rodgers, Sumana Serchan, Michiel Veldhuis, and Brian Watts for field assistance; Joe Wright for the original idea of automated radio-tracking of seeds; John Terborgh and two anonymous reviewers for comments; and the Smithsonian Tropical Research Institute for facilities and administrative support. Financial support was provided by National Science Foundation Grant NSF-DEB-0717071 (to R.K. and M.W.), Netherlands Foundation for Scientific Research Grants NWO-WOTRO W85-239 and NWO-ALW 863-07-008 (to P.A.J.), the Leiden University Fund, the Huygens Scholarship program, a Leiden University Lustra Scholarship, the “Fonds voor Onderzoek ten behoeve van het Natuurbehoud” (to V.Z.-G.), the Smithsonian Tropical Research Institute, and the Frank Levinson Family Foundation.
Supporting Information
Supporting Information (PDF)
Supporting Information
- Download
- 208.22 KB
sm01.wmv
- Download
- 4.84 MB
References
1
DH Janzen, PS Martin, Neotropical anachronisms: The fruits the gomphotheres ate. Science 215, 19–27 (1982).
2
PR Guimarães, M Galetti, P Jordano, Seed dispersal anachronisms: Rethinking the fruits extinct megafauna ate. PLoS ONE 3, e1745 (2008).
3
PS Martin, RG Klein Quaternary Extinctions: A Prehistoric Revolution (Univ of Arizona Press, Tucson, AZ, 1984).
4
HF Howe, Gomphothere fruits: A critique. Am Nat 125, 853–865 (1985).
5
DH Janzen, Enterolobium cyclocarpum seed passage rate and survival in horses, Costa Rican Pleistocene seed dispersal agents. Ecology 62, 593–601 (1981).
6
JR Hunter, Seed dispersal and germination of Enterolobium cyclocarpum (Jacq.) Griseb.(Leguminosae: Mimosoideae): Are megafauna necessary? J Biogeogr 16, 369–378 (1989).
7
TK Pratt, Pleistocene seed dispersal. Science 216, 6 (1982).
8
DH Janzen, Herbivores and the number of tree species in tropical forests. Am Nat 104, 501–528 (1970).
9
JH Connell, On the role of natural enemies in preventing competitive exclusion in some marine animals and in rain forest trees. Dynamics of Populations: Proceedings of the Advanced Study Institute on Dynamics of Numbers in Populations, eds PJ den Boer, GR Gradwell (Pudoc, Oosterbeek, The Netherlands), pp. 298–312 (1971).
10
R Nathan, HC Muller-Landau, Spatial patterns of seed dispersal, their determinants and consequences for recruitment. Trends Ecol Evol 15, 278–285 (2000).
11
R Nathan, et al., Mechanisms of long-distance seed dispersal. Trends Ecol Evol 23, 638–647 (2008).
12
DH Janzen, Removal of seeds from horse dung by tropical rodents: Influence of habitat and amount of dung. Ecology 63, 1887–1900 (1982).
13
W Hallwachs Agoutis (Dasyprocta punctata): The Inheritors of Guapinol (Hymenaea courbaril: Leguminosae). Frugivores and Seed Dispersal, Tasks for Vegetation Science, eds A Estrada, TH Fleming (Dr. W. Junk, Dordrecht, The Netherlands) Vol 15, 285–304 (1986).
14
PM Forget, et al., Seed allometry and disperser assemblages in tropical rainforests: A comparison of four floras on different continents. Seed Dispersal: Theory and Its Application in a Changing World, eds AJ Dennis, RJ Green, EW Schupp, DA Westcott (CABI, Wallingford, UK), pp. 5–36 (2007).
15
SB Vander Wall Food Hoarding in Animals (Univ of Chicago Press, Chicago), pp. 445 (1990).
16
SB Vander Wall, KM Kuhn, MJ Beck, Seed removal, seed predation, and secondary dispersal. Ecology 86, 801–806 (2005).
17
SB Vander Wall, WS Longland, Diplochory: Are two seed dispersers better than one? Trends Ecol Evol 19, 155–161 (2004).
18
E Toby Kiers, TM Palmer, AR Ives, JF Bruno, JL Bronstein, Mutualisms in a changing world: An evolutionary perspective. Ecol Lett 13, 1459–1474 (2010).
19
SB Vander Wall, SH Jenkins, Reciprocal pilferage and the evolution of food-hoarding behavior. Behav Ecol 14, 656–667 (2003).
20
BC Wang, TB Smith, Closing the seed dispersal loop. Trends Ecol Evol 17, 379–385 (2002).
21
BT Hirsch, MD Visser, R Kays, PA Jansen, Quantifying seed dispersal kernels from truncated seed-tracking data. Meth Ecol Evol 3, 595–602 (2012).
22
W Hallwachs, The clumsy dance between agoutis and plants: Scatterhoarding by Costa Rican dry forest agoutis (Dasyprocta punctata: Dasyproctidae: Rodentia). PhD thesis. (Cornell University, Ithaca, NY, 1994).
23
PA Jansen, F Bongers, L Hemerik, Seed mass and mast seeding enhance dispersal by a Neotropical scatter-hoarding rodent. Ecol Monogr 74, 569–589 (2004).
24
R Perea, AS Miguel, L Gil, Acorn dispersal by rodents: The importance of re-dispersal and distance to shelter. Basic Appl Ecol 12, 432–439 (2011).
25
PA Jansen, PM Forget, Scatterhoarding and tree regeneration. Nouragues: Dynamics and Plant-Animal Interactions in a Neotropical Rainforest, eds F Bongers, P Charles-Dominique, PM Forget, M Théry (Kluwer Academic, Dordrecht, The Netherlands), pp. 275–288 (2001).
26
JMT Haugaasen, T Haugaasen, CA Peres, R Gribel, P Wegge, Seed dispersal of the Brazil nut tree (Bertholletia excelsa) by scatter-hoarding rodents in a central Amazonian forest. J Trop Ecol 26, 251–262 (2010).
27
SB Vander Wall, JW Joyner, Recaching of Jeffrey pine (Pinus jeffreyi) seeds by yellow pine chipmunks (Tamias amoenus): Potential effects on plant reproductive success. Can J Zool 76, 154–162 (1998).
28
DY Alexandre, Le rôle disséminateur des éléphants en forêt de Taï, Côte d’Ivoire. La Terre et la Vie 32, 47–72 (1978).
29
N Smythe, Seed survival in the palm Astrocaryum standleyanum: Evidence for dependence upon its seed dispersers. Biotropica 21, 50–56 (1989).
30
BT Hirsch, R Kays, PA Jansen, A telemetric thread-tag for tracking seed dispersal by scatter-hoarding animals. Plant Ecol 213, 933–943 (2012).
31
R Kays, et al., Tracking animal location and activity with an automated radio telemetry system in a tropical rainforest. Comput J 54, 1931–1948 (2011).
32
K Holbrook, et al., Using toucan-generated dispersal models to estimate seed dispersal in Amazonian Ecuador. Seed Dispersal: Theory and Its Application in a Changing World, eds AJ Dennis, EW Schupp, RJ Green, DA Westcott (CABI, Wallingford, UK), pp. 300–321 (2007).
33
SE Russo, CK Augspurger, Aggregated seed dispersal by spider monkeys limits recruitment to clumped patterns in Virola calophylla. Ecol Lett 7, 1058–1067 (2004).
34
P Jordano, C García, JA Godoy, JL García-Castaño, Differential contribution of frugivores to complex seed dispersal patterns. Proc Natl Acad Sci USA 104, 3278–3282 (2007).
35
J Terborgh, Enemies maintain hyperdiverse tropical forests. Am Nat 179, 303–314 (2012).
36
PE Hulme Seed Eaters: Seed Dispersal, Destruction and Demography. Seed Dispersal and Frugivory: Ecology, Evolution and Conservation, Third Frugivores and Seed Dispersal Symposium, eds DJ Levey, WR Silva, M Galetti (CABI, Wallingford, UK), pp. 257–273 (2002).
37
W Jetz, C Carbone, J Fulford, JH Brown, The scaling of animal space use. Science 306, 266–268 (2004).
38
D Gálvez, B Kranstauber, RW Kays, PA Jansen, Scatter hoarding by the Central American agouti: A test of optimal cache spacing theory. Anim Behav 78, 1327–1333 (2009).
39
TC Theimer, Rodent scatterhoarders as conditional mutualists. Seed Fate: Predation, Dispersal and Seedling Establishment, eds PM Forget, JE Lambert, PE Hulme, SB Vander Wall (CABI, Wallingford, UK), pp. 283–295 (2005).
40
PM Forget, PA Jansen, Hunting increases dispersal limitation in the tree Carapa procera, a nontimber forest product. Conserv Biol 21, 106–113 (2007).
41
J Pons, JG Pausas, Acorn dispersal estimated by radio-tracking. Oecologia 153, 903–911 (2007).
42
K Sone, et al., Hoarding of acorns by granivorous mice and its role in the population processes of Pasania edulis (Makino) Makino. Ecol Res 17, 553–564 (2002).
43
N Tamura, Walnut hoarding by the Japanese wood mouse, Apodemus speciosus Temminck. J For Res 6, 187–190 (2001).
44
N Tamura, Application of a radio-transmitter for studying seed dispersion by animals. J Jpn Forestry Soc 76, 607–610 (1994).
45
JA Godoy, P Jordano, Seed dispersal by animals: Exact identification of source trees with endocarp DNA microsatellites. Mol Ecol 10, 2275–2283 (2001).
46
R Kays, PA Jansen, EMH Knecht, R Vohwinkel, M Wikelski, The effect of feeding time on dispersal of Virola seeds by toucans determined from GPS tracking and accelerometers. Acta Oecol 37, 625–631 (2011).
47
JMV Fragoso, KM Silvius, JA Correa, Long-distance seed dispersal by tapirs increases seed survival and aggregates tropical trees. Ecology 84, 1998–2006 (2003).
48
JP Dudley, Seed dispersal by elephants in semiarid woodland habitats of Hwange National Park, Zimbabwe. Biotropica 32, 556–561 (2000).
49
JMV Fragoso, JM Huffman, Seed-dispersal and seedling recruitment patterns by the last Neotropical megafaunal element in Amazonia, the tapir. J Trop Ecol 16, 369–385 (2000).
50
K LoGiudice, R Ostfeld, Interactions between mammals and trees: Predation on mammal-dispersed seeds and the effect of ambient food. Oecologia 130, 420–425 (2002).
51
ME Collinson, JJ Hooker, Fossil evidence of interactions between plants and plant-eating mammals. Philos Trans R Soc Lond B Biol Sci 333, 197–207, 207–208 (1991).
52
EG Leigh Tropical Forest Ecology: A View from Barro Colorado Island (Oxford Univ Press, New York), pp. 264 (1999).
53
PA Jansen, K Elschot, PJ Verkerk, SJ Wright, Seed predation and defleshing in the agouti-dispersed palm Astrocaryum standleyanum. J Trop Ecol 26, 473–480 (2010).
54
N Smythe, The natural history of the Central American agouti (Dasyprocta punctata). Smithson Contrib Zool 257, 1–52 (1978).
55
G Dubost, O Henry, Comparison of diets of the acouchy, agouti and paca, the three largest terrestrial rodents of French Guianan forests. J Trop Ecol 22, 641–651 (2006).
56
O Henry, Frugivory and the importance of seeds in the diet of the orange-rumped agouti (Dasyprocta leporina) in French Guiana. J Trop Ecol 15, 291–300 (1999).
57
PM Forget, DG Wenny, How to elucidate seed fate? A review of methods used to study seed caching and secondary seed dispersal. Seed Fate: Predation, Dispersal and Seedling Establishment, eds PM Forget, JE Lambert, PE Hulme, SB Vander Wall (CABI, Wallingford, UK), pp. 379–394 (2005).
58
PA Jansen, J Den Ouden, Observing seed removal: remote video monitoring of seed selection, predation and dispersal. Seed Fate: Predation, Dispersal and Seedling Establishment, eds PM Forget, JE Lambert, PE Hulme, SB Vander Wall (CABI, Wallingford, UK), pp. 363–378 (2005).
59
HH Hadow, Freeze-branding: A permanent marking technique for pigmented mammals. J Wildl Manage 36, 645–649 (1972).
60
CF Dormann, J Fründ, N Blüthgen, B Gruber, Indices, graphs and null models: Analyzing bipartite ecological networks. The Open Ecology Journal 2, 7–24 (2009).
61
; R Development Core Team R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, Vienna, 2011).
62
T Therneau, A Package for Survival Analysis in S. R package version 2.36–14. (2012).
63
PA Jansen, F Bongers, PJ van der Meer, Is farther seed dispersal better? Spatial patterns of offspring mortality in three rainforest tree species with different dispersal abilities. Ecography 31, 43–52 (2008).
Information & Authors
Information
Published in
Cover image: Pictured are open-air markets amidst heavy traffic in Hyderabad, India, showing the chaotic environments that populations in developing countries must contend with to obtain produce. Much of the food in these markets is transported from small farms outside the city and carried through clogged streets by a vast ad hoc network of small trucks, scooter trucks, and bicycle carts. Laurette Dubé et al. propose a roadmap for progress toward sustainable global nutrition security. Such roadmaps could help reduce hunger and poverty, and halt the spread of noncommunicable chronic diseases. See the article by Dubé et al. on pages 12294–12301, which introduces the Agriculture Development and Nutrition Security Special Feature. Image courtesy of Alex R. Freedman (Community Cooperative Farms, Sheffield, MA).
Classifications
Copyright
Freely available online through the PNAS open access option.
Data Availability
Data deposition: The data reported in this paper has been deposted in the Movebank database, www.movebank.org (doi: 10.5441/001/1.9t0m888q).
Submission history
Published online: July 16, 2012
Published in issue: July 31, 2012
Keywords
Acknowledgments
We thank Tom Garin (Advanced Telemetry Systems), Daniel Obando, Alejandro Ortega, and Meg Crofoot for technical support; Lieneke Bakker, Reyna Bell, Chris Carson, Matt McElroy, Veronica Pereira, Torrey Rodgers, Sumana Serchan, Michiel Veldhuis, and Brian Watts for field assistance; Joe Wright for the original idea of automated radio-tracking of seeds; John Terborgh and two anonymous reviewers for comments; and the Smithsonian Tropical Research Institute for facilities and administrative support. Financial support was provided by National Science Foundation Grant NSF-DEB-0717071 (to R.K. and M.W.), Netherlands Foundation for Scientific Research Grants NWO-WOTRO W85-239 and NWO-ALW 863-07-008 (to P.A.J.), the Leiden University Fund, the Huygens Scholarship program, a Leiden University Lustra Scholarship, the “Fonds voor Onderzoek ten behoeve van het Natuurbehoud” (to V.Z.-G.), the Smithsonian Tropical Research Institute, and the Frank Levinson Family Foundation.
Notes
*This Direct Submission article had a prearranged editor.
Authors
Competing Interests
The authors declare no conflict of interest.
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
Cite this article
109 (31) 12610-12615,
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.