Patrick A. Jansen

Seed mass and mast seeding enhance dispersal by a neotropical scatter-hoarding rodent

Many tree species that depend on scatter-hoarding animals for seed dispersal produce massive crops of large seeds at irregular intervals. Mast seeding and large seed size in these species have been explained as adaptations to increase animal dispersal and reduce predation. We studied how seed size and seed abundance simultaneously influenced seed dispersal and predation by scatter-hoarding rodents in the large-seeded rain forest tree Carapa procera (Meliaceae) in French Guiana. We individually tracked the fates of 3000 seeds, using remote video monitoring and thread-marking. Seed size was manipulated by broadly varying intraspecific seed mass, whereas effects of seed abundance were examined by tracking seeds in three seed-rich years and two seed-poor years. The hypotheses, that seed mass and seed abundance both enhance dispersal success and that seed abundance reinforces the effect of seed mass, were supported by the results. Most seeds were removed by the scatter-hoarding rodent red acouchy (Myoprocta acouchy) and subsequently were buried in scattered, single-seeded caches up to distances >100 m. Seeds that were not removed failed to establish seedlings. Seed removal was slower, pre-removal seed predation was greater, and seed dispersal was less far in seed-rich years than in seed-poor years, suggesting poorer dispersal under seed abundance. However, this was more than counterbalanced by a disproportionally greater survival of cached seeds in seed-rich years. The per capita probability of seed survival and seedling establishment was at least 4½ times greater under seed abundance. Large seeds were removed faster, were more likely to be scatter-hoarded, and were dispersed farther away than smaller ones, resulting in a higher probability of seedling establishment for larger seeds. Size discrimination was greater under seed abundance, albeit only during seed removal. Overall, large seeds shed in rich years had the highest probability of seedling establishment. Hence, both larger seed size and greater seed abundance stimulate rodents to act more as dispersers and less as predators of seeds. We conclude that scatter-hoarding rodents can select for both large seed crops and large seeds, which may reinforce mast seeding.

Thieving rodents as substitute dispersers of megafaunal seeds

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.

Multiple recruitment limitation causes arrested succession in mediterranean cork oak systems

Lack of tree regeneration and persistency of species-poor shrublands represent a growing problem across Mediterranean evergreen oak forests. What constrains forest regeneration is poorly understood, and restoration attempts have been largely unsuccessful. We assessed the contribution of four different mechanisms of tree recruitment limitation (that is, source, dispersal, germination, and establishment) in a cork oak (Quercus suber) system in southern Portugal. Using a combination of field studies and experiments, we quantified seed production, seed removal and dispersal, seed survival and germination, seedling establishment and survival, as well as cork oak natural regeneration for the three dominant vegetation types in this system (Cistus ladanifer shrubland, oak forest, and oak savanna). We found that all four forms of cork oak recruitment limitation were significantly more severe in shrublands than in oak forests and savannas, so that oak seedling recruitment in shrubland was impeded in multiple ways. Our results explain why transitions from shrublands to oak savannas and forests are extremely difficult, and that the release from arrested succession in this system requires the simultaneous relief of multiple constraints on recruitment limitation in the early life history of oaks. These results have important implications for the restoration and conservation of Mediterranean oak systems.

Directed seed dispersal towards areas with low conspecific tree density by a scatter-hoarding rodent.

Scatter-hoarding animals spread out cached seeds to reduce density-dependent theft of their food reserves. This behaviour could lead to directed dispersal into areas with lower densities of conspecific trees, where seed and seedling survival are higher, and could profoundly affect the spatial structure of plant communities. We tested this hypothesis with Central American agoutis and Astrocaryum standleyanum palm seeds on Barro Colorado Island, Panama. We radio-tracked seeds as they were cached and re-cached by agoutis, calculated the density of adult Astrocaryum trees surrounding each cache, and tested whether the observed number of trees around seed caches declined more than expected under random dispersal. Seedling establishment success was negatively dependent on seed density, and agoutis carried seeds towards locations with lower conspecific tree densities, thus facilitating the escape of seeds from natural enemies. This behaviour may be a widespread mechanism leading to highly effective seed dispersal by scatter-hoarding animals.

The role of seed size in dispersal by a scatter-hoarding rodent

Patrick A. Jansen,1 Martijn Bartholomeus,2 Frans Bongers,1 Jelmer A. Elzinga,2,* Jan den Ouden1 and Sipke E. Van Wieren2 1Silviculture and Forest Ecology Group, Department of Environmental Sciences, Wageningen University, PO Box 342, NL-6700 AH Wageningen, The Netherlands; 2Tropical Nature Conservation and Vertebrate Ecology Group, Department of Environmental Sciences, Wageningen University, Bornsesteeg 69, NL-6708 PD Wageningen, The Netherlands

Quantifying levels of animal activity using camera-trap data

Summary 1. Activity level (the proportion of time that animals spend active) is a behavioural and ecological metric that can provide an indicator of energetics, foraging effort and exposure to risk. However, activity level is poorly known for free-living animals because it is difficult to quantify activity in the field in a consistent, cost-effective and non-invasive way. 2. This article presents a new method to estimate activity level with time-of-detection data from camera traps (or more generally any remote sensors), fitting a flexible circular distribution to these data to describe the underlying activity schedule, and calculating overall proportion of time active from this. 3. Using simulations and a case study for a range of small- to medium-sized mammal species, we find that activity level can reliably be estimated using the new method. 4. The method depends on the key assumption that all individuals in the sampled population are active at the peak of the daily activity cycle. We provide theoretical and empirical evidence suggesting that this assumption is likely to be met for many species, but may be less likely met in large predators, or in high-latitude winters. Further research is needed to establish stronger evidence on the validity of this assumption in specific cases; however, the approach has the potential to provide an effective, non-invasive alternative to existing methods for quantifying population activity levels.

Scatterhoarding Rodents and Tree Regeneration

Tree seeds that drop down to the rainforest floor, either naked or embedded in fruit pulp, often disappear after a while. Many authors refer to this phenomenon as seed predation, implicitly assuming that the seed is destroyed by seedeaters such as rodents or ungulates (Forget et al. 1998). However, not all seed removal is by seedeaters. Many species of animals remove seeds while feeding on what is around seeds rather than on the seeds themselves. Terrestrial birds, mammals and reptiles ingest seeds while feeding on fallen fruit (e.g. Fragoso 1997, Erard & Sabatier 1988), tortoises ingest seeds while feeding on dung of frugivorous mammals (B. Josseaume unpublished data.), and dung beetles take and bury seeds with dung (Shephard & Chapman 1998, Andresen 1999, Feer 1999). All these animals may bring viable seeds into favourable conditions.

Automated identification of animal species in camera trap images

Image sensors are increasingly being used in biodiversity monitoring, with each study generating many thousands or millions of pictures. Efficiently identifying the species captured by each image is a critical challenge for the advancement of this field. Here, we present an automated species identification method for wildlife pictures captured by remote camera traps. Our process starts with images that are cropped out of the background. We then use improved sparse coding spatial pyramid matching (ScSPM), which extracts dense SIFT descriptor and cell-structured LBP (cLBP) as the local features, that generates global feature via weighted sparse coding and max pooling using multi-scale pyramid kernel, and classifies the images by a linear support vector machine algorithm. Weighted sparse coding is used to enforce both sparsity and locality of encoding in feature space. We tested the method on a dataset with over 7,000 camera trap images of 18 species from two different field cites, and achieved an average classification accuracy of 82%. Our analysis demonstrates that the combination of SIFT and cLBP can serve as a useful technique for animal species recognition in real, complex scenarios.

Establishment limitation of holm oak (Quercus ilex subsp. ballota (Desf.) Samp.) in a Mediterranean savanna – forest ecosystem

Abstract• Tree recruitment in Mediterranean savannas is generally hampered, in contrast with the original oak forests where these savannas are derived from. We asked whether this difference in recruitment success can be explained by differential post-dispersal survival. For one year we monitored experimentally cached holm oak acorns in a savanna — forest ecosystem in Central Spain, and recorded cache pilferage, type of pilferer, boar rooting, seedling emergence, seedling survival and the cause of mortality.• Cache pilferage was significantly lower in savanna (8%) than in forest (21%). However, the higher cache survival was more than offset by lower seedling emergence and, particularly, by nine times higher seedling mortality in savanna, mainly due to desiccation. Wild boar rooting did not differ between experimental caches and controls without acorns, indicating that individual cached acorns do not trigger rooting activity.• Our results indicate that the difference in post-dispersal survival between savanna and forest is due to lower emergence and, primarily, higher seedling mortality in savanna, not to higher cache pilferage. Absence of safe sites such as shrubs, abundantly present in the forest, may well explain the lack of recruitment in the savanna. Management measures appear necessary for long-term persistence of Mediterranean savannas in general.Résumé• Le recrutement des arbres dans les savanes méditerranéennes est généralement entravé, ce qui contraste avec les forêts de chênes originelles dont ces savanes sont issues. Nous nous sommes demandé si cette différence dans le succès du recrutement pouvait être expliqué par une différence de survie post-dissémination. Pendant une année, nous avons suivi expérimentalement des glands de chêne vert cachés dans un écosystème de savane forestière en Espagne centrale, et nous avons enregistré les chapardages des caches, les types de chapardage, la fouille des sangliers, l’émergence des semis, la survie des semis et les causes de la mortalité• Le chapardage dans les caches a été significativement plus faible dans la savane (8 %) que dans la forêt (21 %). Toutefois, la survie plus élevée dans les caches été plus que compensée par une baisse de l’émergence des semis et, en particulier, par une mortalité des semis neuf fois plus élevée dans la savane, principalement due à la dessiccation. La fouille des racines par les sangliers n’est pas différente entre les caches expérimentales et les témoins sans glands, indiquant que les mises en cache des glands ne déclenchent pas activité de fouille des racines.• Nos résultats indiquent que la différence de survie post-dissémination entre la savane et la forêt est due à la baisse d’émergence et en premier lieu, à une mortalité plus élevée des semis dans la savane, et non pas à des chapardages des caches plus élevés. L’absence de sites abris tels que les arbustes, très présents dans la forêt, peut expliquer l’absence de recrutement dans la savane. Des mesures de gestion apparaissent nécessaires pour une persistance à long terme des savanes méditerranéennes.

Standardized Assessment of Biodiversity Trends in Tropical Forest Protected Areas: The End Is Not in Sight

Extinction rates in the Anthropocene are three orders of magnitude higher than background and disproportionately occur in the tropics, home of half the world’s species. Despite global efforts to combat tropical species extinctions, lack of high-quality, objective information on tropical biodiversity has hampered quantitative evaluation of conservation strategies. In particular, the scarcity of population-level monitoring in tropical forests has stymied assessment of biodiversity outcomes, such as the status and trends of animal populations in protected areas. Here, we evaluate occupancy trends for 511 populations of terrestrial mammals and birds, representing 244 species from 15 tropical forest protected areas on three continents. For the first time to our knowledge, we use annual surveys from tropical forests worldwide that employ a standardized camera trapping protocol, and we compute data analytics that correct for imperfect detection. We found that occupancy declined in 22%, increased in 17%, and exhibited no change in 22% of populations during the last 3–8 years, while 39% of populations were detected too infrequently to assess occupancy changes. Despite extensive variability in occupancy trends, these 15 tropical protected areas have not exhibited systematic declines in biodiversity (i.e., occupancy, richness, or evenness) at the community level. Our results differ from reports of widespread biodiversity declines based on aggregated secondary data and expert opinion and suggest less extreme deterioration in tropical forest protected areas. We simultaneously fill an important conservation data gap and demonstrate the value of large-scale monitoring infrastructure and powerful analytics, which can be scaled to incorporate additional sites, ecosystems, and monitoring methods. In an era of catastrophic biodiversity loss, robust indicators produced from standardized monitoring infrastructure are critical to accurately assess population outcomes and identify conservation strategies that can avert biodiversity collapse.