https://orcid.org/0000-0002-3773-1734
Figures
Abstract
The study of plant-frugivore interactions is essential to understand the ecology and evolution of many plant communities. However, very little is known about how interactions among frugivores indirectly affect plant reproductive success. In this study, we examined direct interactions among vertebrate frugivores sharing the same fruit resources. Then, we inferred how the revealed direct interspecific interactions could lead to indirect (positive or negative) effects on reproductive success of fleshy fruited plants. To do so, we developed a new analytical approach that combines camera trap data (spatial location, visitor species, date and time, activity) and tailored null models that allowed us to infer spatial-temporal interactions (attraction, avoidance or indifference) between pairs of frugivore species. To illustrate our approach, we chose to study the system composed by the Mediterranean dwarf palm, Chamaerops humilis, the Iberian pear tree, Pyrus bourgaeana, and their shared functionally diverse assemblages of vertebrate frugivores in a Mediterranean area of SW Spain. We first assessed the extent to which different pairs of frugivore species tend to visit the same or different fruiting individual plants. Then, for pairs of species that used the same individual plants, we evaluated their spatial-temporal relationship. Our first step showed, for instance, that some prey frugivore species (e.g. lagomorphs) tend to avoid those C. humilis individuals that were most visited by their predators (e.g. red foxes). Also, the second step revealed temporal attraction between large wild and domestic frugivore ungulates (e.g. red deer, cows) and medium-sized frugivores (e.g. red foxes) suggesting that large mammals could facilitate the C. humilis and P. bourgaeana exploitation to other smaller frugivores by making fruits more easily accessible. Finally, our results allowed us to identify direct interaction pathways, that revealed how the mutualistic and antagonistic relations between animal associates derived into indirect effects on both plants seed dispersal success. For instance, we found that large-sized seed predators (e.g. ungulates) had a direct positive effect on the likelihood of visits by legitimate seed dispersers (e.g. red foxes) to both fleshy fruited plants. Then, seed predators showed an indirect positive effect on the plants’ reproductive success. Our new analytical approach provides a widely applicable framework for further studies on multispecies interactions in different systems beyond plant-frugivore interactions, including plant-pollinator interactions, the exploitation of plants by herbivores, and the use of carcasses by vertebrate scavengers.
Citation: Selwyn M, Garrote PJ, Castilla AR, Fedriani JM (2020) Interspecific interactions among functionally diverse frugivores and their outcomes for plant reproduction: A new approach based on camera-trap data and tailored null models. PLoS ONE 15(10): e0240614. https://doi.org/10.1371/journal.pone.0240614
Editor: Pedro G. Blendinger, CONICET - Universidad Nacional de Tucumán, ARGENTINA
Received: June 4, 2020; Accepted: September 29, 2020; Published: October 16, 2020
Copyright: © 2020 Selwyn et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data is within the manuscript and its Supporting Information files.
Funding: This study has been funded by a grant from the Ministerio de Ciencia, Innovación y Universidades, PGC2018-094808-B-I00, to Jose M. Fedriani. Pedro J. Garrote was supported by a fellowship from Fundação para a Ciência e a Tecnologia, SFRH/BD/130527/2017. Antonio R. Castilla was supported by a fellowship from Fundação para a Ciência e a Tecnologia, Fundação para a Ciência e Tecnologia. Miriam Selwyn was supported by a felowship from Consejería de Educación, Junta de Castilla y León. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
A major long-established goal in community ecology and evolutionary biology is to understand how interspecific interactions influence population density, distribution, phenotypes, and genotypes [1], crucial to the selection and evolution of life-history traits. During the few last decades, the study of interspecific interactions has experienced an outstanding progress [2] such as, for example, moving from a traditional pair-wise perspective [3–6] to a more realistic and complex multispecific approach, where multiple species interact with each other [2, 7, 8]. However, most of these studies have focused on a single type of interaction at a time, usually studying either mutualistic or antagonistic interactions among species of particular taxonomic groups [9]. Examples of such interspecific interactions that have been most often investigated separately are competitive interactions among vertebrates [e.g., 10, 11] facilitative interactions among plants [e.g., 12, 13] and mutualistic/antagonistic plant-animal interactions [e.g., 8, 14]. Many of these interaction types take place within the same habitats [15–17] and thus, species involved in one interaction type (e.g. predation, competition) can also participate in other interactions (e.g. plant-animal interactions). However, to our knowledge, studies integrating these interaction types are still scarce [but see 18].
New analytical approaches have emerged allowing coping with the methodological challenge that represents studying systems comprising different species that interact among them in variable ways. For instance, multilayer networks [19], spatially explicit agent-based simulation modeling [20] and tailored null models based on resampling techniques [21] represent powerful tools to investigate such complex systems. In many situations, however, a basic challenge remains as to how to monitor subtle and mixed interactions under suboptimal field conditions, especially when target species are nocturnal, secretive, or otherwise elusive animals (e.g. many vertebrates). Importantly, during the last few decades camera traps have revolutionized wildlife research, enabling the collection of precise photographic evidence of rarely seen species, with relatively little cost [22–24]. Camera traps record very accurate data while barely disturbing the photographed animal, operate continually and silently, providing proof of individuals at a spatially restricted spot (e.g. a fruiting plant, a carcass, a water point), their precise time of visit, and their activity (foraging, perching, fighting, etc.) among other relevant data. Hundreds of thousands, if not millions, of photographs of large numbers of vertebrate species are indeed being systematically collected in diverse habitats worldwide [see review 25]. Surprisingly, until very recently camera trap surveys have been seldom used as a quantitative tool to thoroughly measure direct interactions among individuals of different vertebrate species [but see 26–28].
Here we propose a combination of camera-trap survey data and tailored null models as a unifying methodological framework to investigate whether and how interactions between foraging animals (predation, facilitation) alter subsequent plant-animal interactions (pollination, frugivory, herbivory). To illustrate the usefulness of our approach, we evaluate whether interspecific interactions among vertebrate frugivores alter the likelihood of their subsequent interactions with fleshy-fruited plants. Our analytical approach comprises two basic steps. First, we evaluated if different frugivore species within a study area visited the same fruiting individuals. Then, we used date-time data of photographs and resampling techniques to tailor null models that allowed us to evaluate whether different species of frugivores showed attraction, aversion, or indifference in their timing of visiting (and interacting) with such fruiting plants. Finally, we inferred the indirect consequences of these direct frugivore responses (attractive or aversive) on the plant’s reproductive success [e.g., 18].
To illustrate the value of our approach, we chose as study systems the Mediterranean dwarf palm Chamaerops humilis and the Iberian pear tree Pyrus bourgaeana as well as their shared community of vertebrate frugivores (mostly mammals) at Doñana National Park (SW Spain) [e.g., 18, 29]. This diverse assemblage of frugivores that consume C. humilis and P. bourgaeana’s fruits can be grouped into five functional frugivore groups: carnivores, wild ungulates, domestic ungulates, pulp feeders (i.e. lagomorphs and rodents) and birds, which may act as seed disperses or seed predators [18, 29–33]; thus, providing a wide spectrum of antagonistic and mutualistic interactions among multiple species across trophic levels [32, 34, 35].
In plant communities, their functional traits vary within environmental gradients and among species occupying similar conditions, raising a challenge for the synthesis of functional and community ecology [36]. The coexistence of consumer species is fostered by resource-use and niche differences, leading to greater resource use in communities with higher number of species [37]. The ecological niche of species is multi-dimensional, including three axes of particular importance that explicitly impact diets and spatial-temporal patterns of abundance: trophic interactions, habitat use and their temporal variability [38]. As a result of interspecific facilitative and competitive interactions, resource partitioning, and niche differentiation, our general hypothesis is that visitation of a fruiting plant by an individual of a given species of vertebrate frugivore could alter (either increasing or lessening) the likelihood of subsequent visits by individuals of other vertebrate frugivore species. Specifically, we expected that (i) different frugivores species would tend to visit different fruiting individual plants lessening thus interspecific competition or predation risk, (ii) when using the same fruiting plant, prey frugivore species (e.g. small mammals) would tend to temporally avoid plants being foraged by their predators (e.g. carnivores) to reduce predation risk, and (iii) large-bodied frugivores (e.g. ungulates) would facilitate the plant exploitation to smaller frugivores by making the fruit more easily accessible (e.g. by giving up ripe fruits in the plants’ immediacy).
Materials and methods
Study area and sites
The study was carried out during the fruiting seasons (September to December) of 2018 and 2019 in Doñana National Park, SW of the Iberian Peninsula (37° 9′ N, 6° 26′ W, Fig 1a). The region is characterized by sub-humid Mediterranean climate, with hot and dry summers (June-September) and mild, wet winters (November-February [39]). The average annual temperature ranges between 15.4 and 18.7°C (mean ± SE = 16.9 ± 1°C; n = 35; period 1978–2017 and annual precipitation is highly variable, ranging between 170 and 1028 mm (mean ± SE = 542.6 ± 32.8 mm; n = 35; period 1978–2017), with most rainfall during winter (271.4 ± 27.1 mm) and extreme drought during summer (33.3 ± 5.3 mm) (data from Monitoring Team of Natural Process of Doñana Biological Station; http://icts.ebd.csic.es/en/web/icts-ebd/monitoring-program-physical-environment).
a) Location of Doñana National Park in Europe and the Iberian Peninsula. b) Location of the study sites: Matasgordas (area = 274 ha) and Martinazo (area = 16.7 ha).
We selected two C. humilis populations (10 km apart) (Matasgordas and Martinazo sites, Fig 1b) and one population of P. bourgaeana located at Matasgordas’ site (as density of P. bourgaeana is extremely low in Martinazo). Both sites have suffered consequences derived from human activities which have made them differ in vegetation and physiographic characteristics. In Matasgordas, Mediterranean shrub species (C. humilis, Pistacia lentiscus, Halimium halimifolium, Cistus spp.) as well as some scattered Pyrus bourgaeana, Quercus suber, Fraxinus angustifolia and Olea europaea var. sylvestris trees are slowly recolonizing the old-field area [40–42]. In Martinazo site, the area has been recolonized mainly by early-successional species (H. halimifolium, Ulex spp., and Stauracanthus genistoides) and animal-dispersed native plants (C. humilis, Rubus ulmifolius, Phillyrea angustifolia or Asparagus aphyllus [43]).
Study species
Chamaerops humilis (Mediterranean dwarf palm) and Pyrus bourgaeana (Iberian pear tree) are endemic species to the Mediterranean region, specifically distributed in southern Europe and northern Africa [30, 44]. Chamaerops humilis is a dioecious palm that blooms from March to May and is mostly pollinated by insects [45–48]. Its fruits are ‘polydrupes’ (usually 1–3 drupes, which vary from 1–3 cm in diameter [49]) attached to infrutescences of up to 30 cm long (7–120 fruits per infructescence) and usually located at ~10–30 cm above the ground level. The ripening occurs in autumn (September-November). Seedlings emerge during spring and early summer, experiencing high mortality due to summer droughts and herbivory [29]. Pyrus bourgaeana is a small (3–6 m height) deciduous monoecious tree which flowers during February and March and is pollinated by numerous Hymenoptera, Diptera, and Coleoptera species [50, 51]. Each tree typically produces 200–450 fruits [29] which are non-dehiscent globose pomes (2–3 cm diameter), with a sugary water-rich pulp, that ripe and drop to the ground from September to December [18]. Seedlings emerge during early spring and experience high mortality due to summer droughts and herbivory [52].
In Doñana National Park, C. humilis and P. bourgaeana’s seed dispersal is mainly accomplished by medium-size carnivores such as Eurasian badgers (Meles meles) and red foxes (Vulpes vulpes) [28, 29]. These two carnivores intensively prey on lagomorphs (Oryctolagus cuniculus) and up to five rodent species (mostly Apodemus sylvaticus and Mus spretus) [32] that feed on C. humilis and P. bourgaeana’s pulp and seeds [29, 30, 33]. Lagomorphs and rodents act mostly as pulp feeders, although may occasionally act as short-distance seed dispersers for C. humilis [29]. Whereas for P. bourgaeana, lagomorphs act mostly as pulp feeders and may occasionally disperse some seeds at short distances, rodents damage all seeds eaten [18, 30]. Fruit predators such as red (Cervus elaphus) and fallow deer (Dama dama) grind all ingested seeds [29, 30], although can disperse viable seeds of C. humilis by regurgitation [31]. Wild boars (Sus scrofa) act mostly as seed predators though may occasionally disperse some viable seeds of both species [30, P.J. Garrote unpublished data]. Free-ranging cows and horses are abundant in some areas of Doñana and are known to consume and usually predate both C. humilis and P. bourgaeana seeds. Large mammals (e.g. deer, cows) seem to be better adapted in their foraging, either by breaking many of the defensive needle-like spines in fruiting C. humilis or by plucking P. bourgaeana’s fruits still attached to the branches (Authors personal observations, see S1A File). Hence, ungulates remove many fruits of both plant species and often leave some of them whole or partially consumed around mother plants (Authors unpublished data). These left fruits are more easily accessible to medium-sized frugivores which hardly cope with the palm’s spines and cannot reach the fruits attached to P. bourgaeana branches. Therefore, we predict that the foraging of smaller frugivores could be facilitated by larger ones [e.g., 18, 53–55]. Also, birds such as azure-winged magpie Cyanopica cyanus and blackcap Silvia atricapilla consume P. bourgaeana’s fruits mostly acting as pulp-predators rather than seed dispersers [30]. As a result, this diverse assemblage of frugivores can be grouped into the above-mentioned five functional groups (carnivores, wild ungulates, domestic ungulates, pulp feeders and birds).
Data collection
During 2018 fruiting season, we collected data for 29 fruiting C. humilis individuals in both study sites (N = 9 for Matasgordas and N = 20 for Martinazo). As for 2019 fruiting season, data was collected for 15 C. humilis and 21 P. bourgaeana fruiting individuals, all located at Matasgordas site. Data on crop size was recorded for all selected individuals during both study seasons, using visual survey methods (VSM) [56–58]. Camera traps (LTL ACORN 5310A, detection range = 18 m) were installed to collect data regarding to frugivore species visits and use of fruits. Cameras were placed from three to five meters distance from the focal plants and were automatically activated any time a movement occurred, taking a three-photo sequence every second. Camera sampling effort for each individual plant (i.e. camera active days) was carried out during the time period since the fruits were ripe until there were not any fruits left or these were musty. For a given camera and vertebrate species, we considered successive visits separated by more than 5 minutes between them. Although 5 minutes might lead to recording the same individuals several times, we were interested in quantifying successive visits by the same individuals since such accumulated number of visits is likely to alter the behavior of other frugivores. Also, to test our results, we used a stricter criterion, considering independent species visits when the time window was higher than 30 minutes, and therefore individuals were less likely to be resampled [59–61]. For all considered visits, the following data was recorded: date, time, day number since camera was active, frugivore species and their use of the fruiting individual plant. We classified the use of individual plants by each frugivore visitor into two types: (i) recorded mammals and birds that were not physically interacting with the target plant (value = 0) and (ii) mammals or birds clearly interacting with the plant (most likely eating fruit; value = 1) (see S1B and S1C File). Accordingly, we described as total number of visit data all types of plant use (i.e. value = 0 and 1), whilst total number of interaction data refers only to those in which visitors were clearly interacting with the plant (i.e. value = 1). Data on sampling effort days and total number of photographs used in the analysis can be found in S1 Table.
Analytical approach
To explore our data, we estimated the total number of visits and interactions (as defined above) for each functional group of frugivores (carnivores, wild ungulates, domestic ungulates, pulp feeders and birds) for every focal individual plant. We considered, for each study site, the same pool of vertebrate species due to their high mobility and therefore we did not predict changes in frugivore community structure within each site. Environmental conditions, related to each plant’s micro or meso habitat (e.g. crop size), may affect the visitation patterns of different frugivore species [62]. Thus, we assessed the relation between fruit availability (crop size) and total number of visits and interactions by each functional group using Pearson’s correlation coefficient applying the function corrplot implemented in the R package corrplot [63].
Because the same frugivore individual could have been photographed several times within a short time period, the analyses described below (i.e. null models) were carried out considering photographs of a given frugivore species at a given individual plant separated by at least 5 minutes (or 30 minutes in a second set of analyses). Our approach was first applied to the whole set of visit data and then to the subset of data corresponding to physical interactions between frugivores and C. humilis and P. bourgaeana individuals separately.
We assessed the potential interactions between frugivore species by means of three spatially explicit null models that allow coping with the effect of each individual plant’s spatial location:
Null model 1.
We compared the average number of visits by a frugivore species (sp1) to individual plants visited (PV) and not visited (PNV) by a second frugivore species (sp2). Under the null hypothesis of no interaction between sp1 and sp2, and all else being equal, the average number of visits by sp1 to PV and PNV by sp2 should not differ. If for a given pair of frugivore species the average number of visits by sp1 to PV was significantly larger or smaller than to PNV by sp2 we assumed spatial attraction or avoidance, respectively, between the pair of frugivore species (i.e. sp1 and sp2). For each pair of frugivore species, we examined potential significant differences between the number of visits by one particular species to individual plants visited and not visited by the paired species by fitting Generalized Linear Models with Poisson distribution and log link function [64]. In a second set of analyses, we added “crop size” as a co-variable in the models. Since the new results did not vary regarding the models without the covariate, for major simplicity, we omitted these results.
Null models 2 and 3.
To examine temporal interactions among different frugivore species, we compared the observed mean values of time differences between successive visits by different frugivore species (OMTD) with the 95% percentile of the expected time difference value (95%ETD) based on two different null models [21]. We considered aversive or attractive responses between pairs of frugivore species when OMTD was above or below the 95%ETD limits, respectively. When OMTD was within 95%ETD limits we concluded no significant response between pairs of frugivore species [21].
The metric used to determine observed time differences was the time elapsed (in hours) between the visit of one frugivore species (sp1) until the first visit of a second frugivore species (sp2; i.e. minimum time elapsed between successive interspecific visits). Because we considered the possibility that, for example, sp1 altered the temporal pattern of the plants visits and interactions by sp2 but, for example, sp2 did not alter the temporal pattern of the plants visits and interactions by sp1, the order of species occurrence was taken into account. Thus, for each pair of species, we calculated the time elapsed between a specific species to another (e.g. sp1-sp2) and vice versa (e.g. sp2-sp1). This metric was determined considering the all camera traps with valid data (two camera traps at Martinazo site were discarded) and for all sampling effort days. For all analyses, we only considered visits and interactions of frugivore pairs separated by < 36 hours, as we assume there would not be any species interactions in a greater time window. Interactions between pair of species were only contemplated when the number of observations for OMTD and 95%ETD was ≥4.
For null model 2, the expected minimum time elapsed between successive interspecific visits was calculated for each target pair of frugivore species as for OMTD but considering their occurrences at different and spatially independent (i.e. >100m or >200m apart) plants (S1 Fig). By considering frugivore occurrences at two plants substantially separated we ensured the condition of no potential direct interactions between such frugivore visitors.
For null model 3, calculation of expected time differences between pairs of interspecific frugivore visits was based on the randomization of frugivore occurrence (and timing of occurrence) for each individual plant (S2 Fig). To do so, we run one thousand iterations that assigned occurrence (i.e. visits or interactions) and their timing to a randomly chosen different individual plant. Our procedure preserved not only the observed number of occurrences of each frugivore species, but also their observed circadian rhythms as the observed times linked to each frugivore occurrence (i.e. recorded picture) were not altered. After each iteration, we calculated the time difference between such simulated co-occurrences.
Because results for both C. humilis fruiting seasons were consistent, so as not to be redundant, analyses were carried out uniting data of both seasons. Our analytical approach was performed using free software R 3.5.0 [65]. The implemented packages used to tailor the null models were plyr [66], dplyr [67] and lubridate [68]. See S2 File for the main functions used in the analysis.
Results
Chamaerops humilis
Do pairs of frugivore species visit the same fruiting C. humilis? (null model 1).
For the complete set of visits in a time window higher than 5 minutes, we found that, in decreasing order, the functional groups with higher number of visits were wild ungulates, domestic ungulates, carnivores, birds and pulp feeders. However, for interaction data the most frequent frugivores were wild ungulates, carnivores, pulp feeders, birds and domestic ungulates (S2 Table).
For the complete visit data, only total number of carnivores’ visits was positively correlated with crop size (r5’ = 0.75, n5’ = 44, P5’ <0.001). Regarding to data corresponding to total number of interactions with C. humilis we found that crop size was positively correlated with total number of interactions (r5’ = 0.57, n5’ = 44, P5’ <0.001), as well as with carnivores (r5’ = 0.78, n5’ = 44, P5’ <0.001) and wild ungulates (r5’ = 0.44, n5’ = 44, P5’ <0.01).
In relation to all visits, we found evidence of spatial attraction and segregation in the use of C. humilis by some predator-prey species (Fig 2). For instance, on average, badgers visited 2.28 times more often those palms visited by lagomorphs (Fig 2a). Also, lagomorph’s visits to palms not visited by red foxes were 8.6 times higher as compared to the mean number of visits to palms visited by red foxes (Fig 2b). Our results showed that palms visited by large ungulates, were more often visited by medium-sized frugivores. For example, red foxes visited palms which were visited by cows 3.34 times more often than those that were not (Fig 2c) and also, visited palms which were visited by horses 7.90 times more often than those that were not (Fig 2d).
The observed values (black bars) represent the mean number of visits by a frugivore species (sp1) to individual plants visited (PV) by a second frugivore species (sp2). The expected values (grey bars) represent the mean number of visits by sp1 to plants not visited (PNV) by sp2. (* P<0.05, ** P<0.01, *** P<0.001).
Are there differences between observed and expected minimum time elapsed between successive interspecific visits to C. humilis? (null model 2).
When considering the whole set of visits in a time window higher than 5 minutes, null model 2 showed significant interactions between some frugivore pairs. Specifically, the observed mean difference between the time of cow and red fox visits was lower (0.16 times for 100m and 0.14 times for 200 m) than the expected mean difference, therefore indicating strong temporal attraction between both species (Fig 3A and 3B). Also, the observed mean difference between the time of cow and bird was between 0.93 and 0.90 times lower than expected (depending on the criteria used), indicating a moderate attraction between them (Fig 3A and 3B). Conversely, the observed mean differences between the visitation times of horse and badger was 1.94 times greater than the expected mean differences for this frugivore pair, indicating temporal aversion (Fig 3B). As to the subset of data corresponding to interactions between frugivores and C. humilis, and separated at least 100 or 200 meters, we found a significant and strong attractive response for the boar-badger pair, which was 0.35 times lower than the expected value for both criterions (Fig 3C and 3D). Also, we found that for a separation of at least 200 meters, that the fox-horse pair showed a significant and attractive response, with the observed mean time difference between the visitations being only 0.62 times the expected value (Fig 3D).
Black dots indicate the observed mean time differences (OMTD). Black lines and grey bars represent the expected 95% and 90% intervals of time elapsed between visits, respectively. (A) Interactions derived from the whole set of visit data, using C. humilis individuals at a minimum distance of 100 m. (B) Interactions derived from the whole set of visit data, using C. humilis individuals at a minimum distance of 200 m. (C) Interactions derived from data corresponding to frugivores physical interactions with C. humilis at a minimum distance of 100 meters. (D) Interactions derived from data corresponding to frugivores physical interactions with C. humilis at a minimum distance of 200 meters.
Are there differences between the observed and the expected time difference between interspecific frugivore visits to C. humilis? (null model 3).
When considering all visits within a time window higher than 5 minutes, the difference between the observed and the expected time differences between interspecific visits by frugivore pairs was significant in a few cases. Specifically, the observed mean difference between the pairs cow-red fox and horse-lagomorph visits was much lower (0.17 and 0.30 times, respectively) than the expected mean difference, thus indicating temporal attraction (Fig 4A). Besides, the observed mean differences between the visit times by the lagomorph-badger and lagomorph-rodent pairs was 1.40 and 1.15 times greater, respectively, than the expected mean differences, indicating temporal aversion (Fig 4A). Furthermore, when considering physical interaction data, we found an attractive response for the bird-boar pair, which observed mean difference was 0.33 times lower than the expected one (Fig 4B). Finally, we found a significant aversive response for the lagomorph-rodent pair, which was 1.33 times greater than the expected mean time difference, thus indicating aversion (Fig 4B).
Black dots indicate the observed mean time differences (OMTD). Black lines and grey bars represent the expected 95% and 90% intervals of time elapsed between visits, respectively. (A) Interactions derived from the whole set of visit data. (B) Interactions derived from data corresponding to frugivore physical interactions with C. humilis.
Pyrus bourgaeana
Do pairs of frugivore species visit the same fruiting P. bourgaeana? (null model 1).
For the complete set of visits in a time window higher than 5 minutes, we found that, in decreasing order, the functional frugivore groups with higher number of visits were wild ungulates, pulp feeders, birds and carnivores. However, for interaction data the most frequent frugivores were wild ungulates, pulp feeders, carnivores and birds (S2 Table).
For the complete visit data, total number of carnivores’ and wild ungulate visits were positively correlated with crop size (r5’ = 0.54, n5’ = 21, P5’ <0.01 and r5’ = 0.52, n5’ = 21, P5’ <0.001, respectively). Regarding to data corresponding to total number of interactions with P. bourgaeana, we found that crop size was positively correlated with total number of interactions (r5’ = 0.52, n5’ = 21, P5’ <0.001), as well as with carnivores (r5’ = 0.55, n5’ = 21, P5’ <0.01) and wild ungulates (r5’ = 0.52, n5’ = 21, P5’ <0.001).
In relation to all visits (Fig 5), we found significant evidence of spatial avoidance by lagomorphs which visited 0.98 times less often those P. bourgaeana trees visited by badgers (Fig 5a). Also, we observed that trees visited by large ungulates, were more often visited by medium or small-sized frugivores. For example, lagomorphs only visited trees that were visited by boars (Fig 5b) or red deer (Fig 5c).
The observed values (black bars) represent the mean number of visits by a frugivore species (sp1) to individual plants visited (PV) by a second frugivore species (sp2). The expected values (grey bars) represent the mean number of visits by sp1 to plants not visited (PNV) by sp2. (* P<0.05, ** P<0.01, *** P<0.001).
Are there differences between observed and expected minimum time elapsed between successive interspecific visits to P. bourgaeana? (null model 2).
When taking into account the whole set of visit data, within a time window higher than 5 minutes, null model 2 showed only one significant response regardless of the spatial independence distance criterion. Specifically, the observed mean difference between the time of fallow deer and red deer visits was lower (0.13 times for 100 and 200 meters) than the expected mean difference for this frugivore pair, therefore indicating temporal attraction.
Are there differences between the observed and the expected time difference between interspecific frugivore visits to P. bourgaeana? (null model 3).
When considering all visits within a time window higher than 5 minutes, the difference between the observed and the expected time differences between interspecific visits by frugivore pairs was significant in only a few cases. Specifically, the observed mean time difference between the pairs red deer-red fox, lagomorph-fallow deer and fallow deer-red deer visits was always lower (0.73, 0.66 and 0.09 times, respectively) than the expected mean difference, thus indicating temporal attraction (Fig 6A).
Black dots indicate the observed mean time differences (OMTD). Black lines and grey bars represent the expected 95% and 90% intervals of time elapsed between visits, respectively. (A) Interactions derived from the whole set of visit data.
Similar patterns were found when applying the 3 null models for visit and interaction data within a time window higher than 30 minutes. Thus, regardless of the criteria used to select the dataset, for most cases, frugivore pairs used the same C. humilis and P. bourgaeana fruiting individuals. Most of the resulting interactions were consistent when restricting the analysis to a narrower time window (30 minutes). Further details on the complete outcome, figures and statistical summaries (for both 5 and 30 minutes) can be found in S3 File and S3 Table, respectively.
Indirect effects on plant reproductive success
As a result of summing up the outcomes of null models 2 and 3, we assessed both direct and indirect interaction pathways among C. humilis and P. bourgaeana frugivores’ in relation to these plants reproductive success. For instance, we found that large ungulates such cows or red deer, had a direct (mostly) negative effect on seed dispersal as they grind most ingested seeds (Fig 7A). However, cows had a direct positive effect for legitimate seed dispersers, such as red foxes which were attracted to palms visited by this large ungulate and thus, lead to an indirect positive effect on C. humilis’ reproductive success (Fig 7A). Our results also showed a similar situation for the interactions between red deer, red foxes, and P. bourgaeana (Fig 7A). Finally, our study also disclosed how, for instance, lagomorphs which mainly had a direct negative effect on C. humilis reproductive success acting as pulp feeders, also had a direct negative effect on seed predators such as red deer. Thus, these direct interspecific interactions lead to an indirect positive effect on C. humilis reproductive success (Fig 7B).
Direct and indirect interaction pathways between pairs of frugivore species in relation to Chamaerops humilis and Pyrus bourgaeana’s reproductive success based on interaction patterns obtained from null models 2 and 3.
Discussion
This study represents one of the first assessments of whether and how functionally diverse vertebrate frugivores alter fruiting plant interactions with subsequent frugivore associates [but see 18]). Specifically, our results revealed, for example, how some frugivore prey (e.g. lagomorphs) avoid those C. humilis individuals most often visited by one of its major predators (e.g. red foxes [32]). Additionally, we revealed temporal attraction between large-sized frugivores (e.g. cows, red deer) and medium-sized frugivores (e.g. red foxes) suggesting that ungulates facilitate C. humilis and P. bourgaeana exploitation to other frugivores by making fruits more accessible. Our analytical approach, based on camera trap data and tailored null models, has also allowed us to identify a number of interactions among frugivores with potential indirect effects on C. humilis and P. bourgaeana’s reproductive success. Our approach can be widely applied by taking advantage of the large amount of data generated by numerous (on-going and completed) camera-traps surveys done all over the world [25]. This will certainly foster a most comprehensive understanding of the direct and indirect effects of interspecific interactions on ecosystem functioning within and across spatial scales.
Plant crop size variation and frugivore visitation
Our results revealed that frugivores respond to crop sizes of fruiting plants, by visiting more frequently those plants bearing more fruits. These results are in line with previous studies that have also found a positive relationship between frugivore visitation and plant resource availability [see 69 and references therein]. Interestingly, the positive relationship between crop size and frugivore visitation was mostly driven by two functional groups of frugivores (i.e. carnivores and wild ungulates). Carnivores (i.e. badgers and foxes) are the most effective seed dispersers of C. humilis and P. bourgaeana [29, 30], whereas wild ungulates can behave as seed predators or seed dispersers [29–31]. Therefore, a greater dispersal success of individuals bearing more fruits due to higher visitation of the most effective dispersers can be limited depending on whether ungulates behave mostly as dispersers or as seed predators [70]. Other environmental conditions derived from the micro (e.g. plant’s location, herbaceous cover, distance to shrub and light availability) and meso habitat (e.g. elevation and type of forest cover) related to each individual plant as well as each animal’s niche differences, that allow the coexistence of consumer species, could also exert an effect on the spatial pattern of frugivore visitation [62, 71]. In our system, most of the species (carnivores and ungulates) show high mobility, being able to travel several kilometers within the same day. Therefore, and given the relatively small spatial scale of our study sites, it can be assumed that different species of frugivores had access to all fruiting plants. This scenario however differs for rodents and lagomorphs since they are less mobile; thus, a few cases of spatial avoidance by small mammals identified (null model 1) should be considered with caution.
Competitive, predatory and facilitative interactions among functionally diverse frugivores
Decrease in fruit display due to exploitation by frugivores is expected to lessen subsequent visitation rates by other fruit consumers [e.g., 72]. In this line, null model 1 results showed that lagomorphs visited less often palms visited by boars, and that red deer visited less often palms visited by rodents, likely because they lessened available ripe fruits. Predator-prey interactions may also affect the spatial and temporal foraging patterns by frugivores [73–75]. Results from null model 1 revealed that lagomorphs tended to avoid C. humilis individuals visited by red foxes, likely to reduce predation risk. Also, null models 2 and 3 suggest that badgers seemed to temporally avoid palms visited by lagomorphs. This unexpected result turns less surprising if we consider the fact that badgers do not generally prey upon adult lagomorphs (i.e. the ones that feed on C. humilis fruit) but upon small juveniles by digging them out from their burrows [32].
Facilitative interactions in the acquisition of food are highly prevalent among many terrestrial vertebrates and have critical importance in structuring and function of many communities [53–55, 76, 77]. In contrast with Carreira et al.’s [27] findings, we detected that medium-sized frugivores tended to be temporally attracted by larger ungulates. For instance, null models 2 and 3 indicate that red foxes visited those palms previously visited by cows more often than would be expected under the null hypothesis of no interaction between both mammals. This pattern could be a result of ungulates giving up ripe fruits in the palm’s immediacy or reducing the number of defensive needle-like spines in fruiting C. humilis (Authors personal observations) making its fruits more easily accessible to such medium-sized frugivores. Also, null model 3 showed that red foxes seem to be attracted to P. bourgaeana individuals previously visited by red deer. This could relate to red deer plucking the fruits from the trees leaving fallen ripe fruits within the tree’s immediacy (Authors personal observations; see S1C File) which might attract smaller frugivores. Our study shows how this functionally diverse assemblage of frugivores leads to interspecific facilitation in foraging, a widespread pattern in other temperate and tropical habitats [55, 76, 77]. Positive interactions may be an important mechanism linking high diversity to high productivity under stressful environmental conditions [54, 78]. Identifying the specific mechanisms by which frugivore species sense each other is out of the scope of the present investigation, although, signals such odors and trail and territory marks, amongst others, are frequently perceived by numerous species of terrestrial vertebrates [79–83] and may be involved in our system.
Community level and indirect effects on C. humilis and P. bourgaeana reproductive success
Our approach allowed us to identify direct interactions among frugivores of C. humilis and P. bourgaeana but also potential indirect effects on the reproductive success of plant populations. These indirect effects can either be positive or negative, depending on whether and how they change subsequent visits by antagonistic or mutualistic animal associates. This is an important consideration since the fruits of most plant species are consumed by functionally diverse assemblages of frugivores such as legitimate seed dispersers, seed predators, and pulp feeders [52, 84–86]. As predicted, seed predators and pulp feeders can alter the interaction frequency between fleshy-fruited plants and their mutualistic and antagonistic animal associates [18]. In the first case, ungulate seed predators increased the interaction frequency with legitimate seed dispersers, for both C. humilis and P. bourgaeana yielding thus an indirect positive effect on seed survival (Fig 7A). Secondly, pulp feeders, that often lessen long-distance dispersal [52], decreased C. humilis’ foraging by ungulate seed predators, enhancing thus seed survival and fruit removal by mutualistic animal associates (Fig 7B). Therefore, both, mutualistic and antagonistic partners that shape the interaction outcomes among individual plants lead to contrasting seed dispersal success [18, 70], appearing to shape a complex web of direct and indirect effects which’s net effect is most likely dependent on the community context [52]. This outcome may be strongly dependent on initial population densities that would most likely alter the dynamics of the system, thus, the net effect of dispersal success, among other factors, will probably be a result of the abundance of different functional guild of frugivores (legitimate seed dispersers, pulp feeders and seed predators) [87–89].
Conclusions
The combined use of camera trap data and tailored null models has proved to be an effective tool to assess positive and negative interactions between functionally diverse frugivore in relation to a shared resource and infer direct and indirect interaction pathways that influence the reproductive success of fleshy-fruited plants. This new approach may provide advantages and complementarities to ongoing research and methodologies for multispecies interactions such as multilayer networks or spatially explicit agent-based simulation modeling [19–20, 89]. For instance, by relaying on data from camera-trapping, our approach can be applied to uncover interspecific interactions among understudied nocturnal or otherwise elusive animals such as many mammals and other secretive vertebrates. Our analytical approach enabled us to evaluate the specific hypothesis and predictions that we had projected for our particular study system and would hardly have been able to achieve using other similar approaches [see 26–28] that impose a number of requirements and assumptions that did not always fit our specific system.
This study may also foster a greater understanding of the implications of multispecies plant-animal interactions on ecosystem function and restoration by means of direct and indirect interaction pathways that influence seed dispersal and recruitment in disturbed areas [90]. It can also enrich a great potential of interaction network researches predicting indirect novel interactions or changes in functional network structures under global changing conditions (e.g. land use changes, human disturbance and climate change) [35]. Indeed, although we have focused here on interactions among frugivores and two fruiting plants, our approach can be easily expanded at a community level by increasing the sampling effort (i.e. number of camera traps). Finally, it can also be applied to other varied systems and species such as, for example, plant-herbivore systems [54], plant-pollinator systems [91], or carcass-scavenger systems [92], taking advantage of the large amount of camera-trap surveys that are collecting data all over the world [25].
Supporting information
S1 File. Classified frugivore plant use.
A) Red deer plucking the fruits from Pyrus bourgaeana. B) Frugivores not interacting with the plant (value = 0). C) Frugivores physically interacting with the plant, most likely eating its fruits (interaction value = 1).
https://doi.org/10.1371/journal.pone.0240614.s001
(PDF)
S2 File. R functions and packages.
Main functions for tailored models 1, 2 and 3 used in the analysis.
https://doi.org/10.1371/journal.pone.0240614.s002
(PDF)
S3 File. Complete outcome, figures and statistical summaries (for both 5 and 30 minutes).
A) Difference in Chamaerops humilis visit patterns for frugivore pairs of species within a time window higher than 5 minutes (null model 1). B) Difference in Chamaerops humilis visit patterns for frugivore pairs of species within a time window higher than 30 minutes (null model 1). C) Zoom of difference in Chamaerops humilis visit patterns for frugivore pairs of species within a time window higher than 5 minutes (null model 1). D) Difference in P. bourgaeana visit patterns for frugivore pairs of species within a time window higher than 5 minutes (null model 1). E) Difference in P. bourgaeana visit patterns for frugivore pairs of species within a time window higher than 30 minutes (null model 1). F) Zoom of difference in P. bourgaeana visit patterns for frugivore pairs of species within a time window higher than 5 minutes (null model 1). G) Significant interactions between frugivores of Chamaerops humilis resulting from applying null model 2 within a time window higher than 30 minutes. H) Significant interactions between frugivores of Chamaerops humilis resulting from applying null model 3 within a time window higher than 30 minutes. I) Significant interactions between frugivores of Pyrus bourgaeana resulting from applying null model 3 within a time window higher than 30 minutes.
https://doi.org/10.1371/journal.pone.0240614.s003
(PDF)
S1 Fig. Null model 2.
Dwarf palms (Chamaerops humilis) A and B are spatially independent individuals, in this case, separated by 100 meters from each other (the set of analysis was also run for palm trees separated 200 meters for a stricter spatial independence criterion). For this example, a badger has been recorded at dwarf palm 1 at the time bt0. For dwarf palm 2, two species have been recorded twice each: as for the first species, a fox has been recorded at time ft1 and time after, another fox has been recorded at time ft2. Whilst the second species was a horse recorded at time ht3 and after a period of time, another horse has been recorded at time ht4. To calculate expected time differences between pairs of interspecific species we have only considered the first successive visit of a different species. Therefore, for this example we have calculated time elapsed between b t0—ht3 and bt0- ft1. This scheme also applies to the Iberian pear (Pyrus bourgaeana) tree system.
https://doi.org/10.1371/journal.pone.0240614.s004
(TIF)
S2 Fig. Null model 3.
A, B and C represent Chamaerops humilis individuals at which one or both target frugivore groups have been recorded. For this example, a badger has been recorded at dwarf palms A (at time bt1) and B (at time bt2), and a red fox has been recorded at dwarf palms A (at time ft1), B (at time ft2) and C (at time ft3). To calculate expected time differences between the pair of interspecific species, we randomly assigned occurrence of frugivore groups (and the timing of the occurrences) by shuffling them 1000 times. When, by chance, both frugivore species (badger and red fox in this case) meet at the same C. humilis individual, we calculated time differences between such co-occurrence. In this example we have calculated time elapsed between bt1 and ft2 for C. humilis A and between bt2 and ft1 for C. humilis B. This scheme also applies to the Iberian pear tree (Pyrus bourgaeana) system.
https://doi.org/10.1371/journal.pone.0240614.s005
(TIF)
S1 Table. Sampling effort days and total number of photographs used in the analysis for both C. humilis and P. bourgaeana.
https://doi.org/10.1371/journal.pone.0240614.s006
(PDF)
S2 Table. Overall patterns.
Total number of visits and interactions for each frugivore functional group and plant species for both 5 and 30 minutes between successive species records.
https://doi.org/10.1371/journal.pone.0240614.s007
(PDF)
S3 Table. Statistical summary for null models 2 and 3.
Table 1) Chamaerops humilis 5 minutes statistical summary for null model 2. Table 2) Chamaerops humilis 30 minutes statistical summary for null model 2. Table 3) Chamaerops humilis 5 minutes statistical summary for null model 3. Table 4) Chamaerops humilis 30 minutes statistical summary for null model 3. Table 5) Pyrus bourgaeana 5- and 30-minutes statistical summary for null model 2. Table 6) Pyrus bourgaeana 5- and 30-minutes statistical summary for null model 3.
https://doi.org/10.1371/journal.pone.0240614.s008
(PDF)
Acknowledgments
We are sincerely grateful to the staff of Doñana National Park for their assistance and to Angelo Santana who kindly helped us with the statistical analyses. Pedro Blendinger, Juliano Bogoni, and Luc Tedonzong provided numerous constructive comments and suggestions that improved a previous version of the manuscript.
References
- 1. Strauss SY, Irwin RE (2004) Ecological and evolutionary consequences of multispecies plant-animal interactions. Annu. Rev. Ecol. Evol. Syst., 35, 435–466.
- 2.
Thompson JN (2014) Interaction and coevolution. University of Chicago Press.
- 3. Maddox GD, Root RB (1987) Resistance to 16 diverse species of herbivorous insects within a population of goldenrod, Solidago altissima: genetic variation and heritability. Oecologia, 72, 8–14. pmid:28312889
- 4. Maddox GD, Root R B (1990) Structure of the encounter between goldenrod (Solidago altissima) and its diverse insect fauna. Ecology, 71, 2115–2124.
- 5. Rauscher MD, Simms EL (1989) The evolution of resistance to herbivory in Ipomoea purpurea. I. Attempts to detect selection. Evolution (USA).
- 6. Juenger T, Bergelson J (1998) Pairwise versus diffuse natural selection and the multiple herbivores of scarlet gilia, Ipomopsis aggregata. Evolution, 52, 1583–1592. pmid:28565335
- 7. Polis GA, Strong DR (1996) Food web complexity and community dynamics. The American Naturalist, 147, 813–846.
- 8. Bascompte J, Jordano P (2007) Plant-animal mutualistic networks: the architecture of biodiversity. Annual Review of Ecology Evolution and Systematics, 38, 567–593.
- 9. Sauve AM, Fontaine C, Thébault E (2014) Structure-stability relationships in networks combining mutualistic and antagonistic interactions. Oikos, 123, 378–384.
- 10. Fedriani JM, Fuller TK, Sauvajot RM, York EC (2000) Competition and intraguild predation among three sympatric carnivores. Oecologia, 125, 258–270. pmid:24595837
- 11. Arsenault R, Owen-Smith N (2002) Facilitation versus competition in grazing herbivore assemblages. Oikos, 97, 313–318.
- 12.
Callaway RM (2007) Positive interactions and interdependence in plant communities. Springer Science & Business Media.
- 13. Garrote PJ, Castilla AR, Fedriani JM (2019) The endemic Mediterranean dwarf palm boosts the recolonization of old-fields: Implications for restoration. Journal of environmental management, 250, 109478. pmid:31493700
- 14. Eldridge DJ, Delgado-Baquerizo M, Travers SK, Val J, Oliver I, Hamonts K, et al (2017) Competition drives the response of soil microbial diversity to increased grazing by vertebrate herbivores. Ecology, 98, 1922–1931. pmid:28459141
- 15. Palomares F, Ferreras P, Fedriani JM, Delibes M (1996). Spatial relationships between Iberian lynx and other carnivores in an area of south-western Spain. Journal of Applied Ecology, 5–13.
- 16. Barasona JA, Latham MC, Acevedo P, Armenteros JA, Latham ADM, Gortazar C, et al. (2014). Spatiotemporal interactions between wild boar and cattle: implications for cross-species disease transmission. Veterinary research, 45, 122. pmid:25496754
- 17. Triguero-Ocaña R, Barasona JA, Carro F, Soriguer RC, Vicente J, Acevedo P (2019). Spatio-temporal trends in the frequency of interspecific interactions between domestic and wild ungulates from Mediterranean Spain. PloS one, 14, e0211216. pmid:30682123
- 18. Fedriani JM, Delibes M (2013) Pulp feeders alter plant interactions with subsequent animal associates. Journal of Ecology, 101, 1581–1588.
- 19. Pilosof S, Porter MA, Pascual M, Kéfi S (2017) The multilayer nature of ecological networks. Nature Ecology & Evolution, 1, 0101.
- 20.
Railsback SF, Grimm V (2019) Agent-based and individual-based modeling: a practical introduction. Princeton university press.
- 21.
Manly BF (2018) Randomization, bootstrap and Monte Carlo methods in biology. Chapman and Hall/CRC.
- 22.
Wearn OR, Glover-Kapfer P (2017) Camera trapping for conservation: A guide to best practices. WWF Conservation Technology Series, 1,1. Woking, UK: WWF-UK.
- 23. Apps P, McNutt JW (2018) Are camera traps fit for purpose? A rigorous, reproducible and realistic test of camera trap performance. African journal of ecology, 56, 710–720.
- 24. Bahaa-el-din L, Cusack JJ (2018) Camera trapping in Africa: Paving the way for ease of use and consistency. African Journal of Ecology, 56, 690–693.
- 25.
O’Connell AF, Nichols JD, Karanth KU (2010) Camera traps in animal ecology: methods and analyses. Springer Science & Business Media.
- 26. Karanth KU, Srivathsa A, Vasudev D, Puri M, Parameshwaran R, Kumar NS (2017) Spatio-temporal interactions facilitate large carnivore sympatry across a resource gradient. Proceedings of the Royal Society B: Biological Sciences, 284, 20161860. pmid:28179511
- 27. Carreira DC, Brodie JF, Mendes CP, Ferraz KMP, Galetti M (2020) A question of size and fear: competition and predation risk perception among frugivores and predators. Journal of Mammalogy.
- 28. Monterroso P, Díaz-Ruíz F, Lukacs PM, Alves PC, Ferreras P (2020) Ecological traits and the spatial structure of competitive coexistence among carnivores. Ecology.
- 29. Fedriani JM, Delibes M (2011) Dangerous liaisons disperse the Mediterranean dwarf palm: Fleshy-pulp defensive role against seed predators. Ecology, 92, 304–315. pmid:21618910
- 30. Fedriani JM, Delibes M (2009) Seed dispersal in the Iberian pear, Pyrus bourgaeana: a role for infrequent mutualists. Ecoscience, 16, 311–321.
- 31. Castañeda I, Fedriani JM, Delibes M (2018) Potential of red deer (Cervus elaphus) to disperse viable seeds by spitting them from the cud. Mammalian Biology, 90, 89–91.
- 32. Fedriani JM, Palomares F, Delibes M (1999) Niche relations among three sympatric Mediterranean carnivores. Oecologia, 121, 138–148. pmid:28307883
- 33. Muñoz-Gallego R, Fedriani JM Traveset A, (2019) Non-native mammals are the main seed dispersers of the ancient Mediterranean palm Chamaerops humilis L. in the Balearic Islands: rescuers of a lost seed dispersal service? Frontiers in Ecology and Evolution, 7, 161.
- 34.
Valverde JA (1967) Estructura de una comunidad de vertebrados terrestres. Estación Biológica de Doñana. Monografia, 1.
- 35. Kissling WD, & Schleuning M. (2015) Multispecies interactions across trophic levels at macroscales: retrospective and future directions. Ecography, 38(4), 346–357.
- 36. Ackerly DD, & Cornwell WK (2007) A trait-based approach to community assembly: partitioning of species trait values into within-and among-community components. Ecology letters, 10(2), 135–145. pmid:17257101
- 37. Finke DL, & Snyder WE (2008) Niche partitioning increases resource exploitation by diverse communities. Science, 321(5895), 1488–1490. pmid:18787167
- 38.
Chase JM, & Leibold MA (2003) Ecological niches. Linking classical and contemporary approaches. University of Chicago Press.
- 39. Żywiec M, Muter E, Zielonka T, Delibes M, Calvo G, Fedriani JM (2016) Long-term effect of temperature and precipitation on radial growth in a threatened thermo-mediterranean tree population. Trees 31,1–11.
- 40. Fedriani JM, Wiegand T (2014) Hierarchical mechanisms of spatially contagious seed dispersal in complex seed-disperser networks. Ecology, 95, 514–526. pmid:24669744
- 41. Fedriani JM, Wiegand T, Delibes M (2010) Spatial pattern of adult trees and the mammal-generated seed rain in the Iberian pear. Ecography, 33, 545–555.
- 42. Fedriani JM, Wiegand T, Ayllón D, Palomares F, Suárez-Esteban A, Grimm V (2018) Assisting seed dispersers to restore old-fields: an individual-based model of the interactions among badgers, foxes, and Iberian pear trees. J Appl Ecol 55,600–611.
- 43. Granados M, Martin A, García Novo F (1988) Long-term vegetation changes on the stabilized dunes of Doñana National. Vegetatio, 75, 73–80.
- 44. Guzmán B, Fedriani JM, Delibes M, Vargas P (2017) The colonization history of the Mediterranean dwarf palm (Chamaerops humilis L., Palmae). Tree Genetics & Genomes, 13, 24.
- 45. Anstett MC (1999) An experimental study of the interaction between the dwarf palm (Chamaerops humilis) and it’s floral visitor Derelomus chamaeropsis throughout the life cycle of the weevil. Acta Oecologica, 20, 551–558.
- 46. Jácome-Flores ME, Delibes M, Wiegand T, Fedriani JM (2018) Spatio-temporal arrangement of Chamaerops humilis inflorescences and occupancy patterns by its nursery pollinator, Derelomus chamaeropsis. Annals of botany, 121, 471–482. pmid:29300822
- 47. Dufaÿ M (2010) Impact of plant flowering phenology on the cost/benefit balance in a nursery pollination mutualism, with honest males and cheating females. Journal of evolutionary biology, 23, 977–986. pmid:20345818
- 48. García Y, Clara Castellanos M, Pausas JG (2018) Differential pollinator response underlies plant reproductive resilience after fires. Annals of botany, 122, 961–971. pmid:29982299
- 49.
Valdés-Castrillón B, Talavera-Lozano S, & Fernández-Galiano FE (1987) Flora vascular de Andalucía Occidental, 3 volúmenes. Ketres Editora SA, Barcelona, España.
- 50. Herrera J (1988) Pollination relationships in southern Spanish Mediterranean shrublands. J Ecol 76, 274–287
- 51. Żywiec M, Delibes M, Fedriani JM (2012) Microgeographical, inter-individual, and intra-individual variation in the flower characters of Iberian pear Pyrus bourgaeana (Rosaceae). Oecologia, 169(3), 713–722. pmid:22200854
- 52. Fedriani JM, Zywiec M, Delibes M (2012) Thieves or mutualists? Pulp feeders enhance endozoochore local recruitment. Ecology, 93, 575–587.
- 53. Hernandez A (2008) Cherry removal by seed-dispersing mammals: mutualism through commensal association with frugivorous birds. Polish Journal of Ecology, 56(1), 127.
- 54.
Owen-Smith RN (2002) Adaptive herbivore ecology: from resources to populations in variable environments. Cambridge University Press.
- 55. Severo-Neto F, Yule TS, & Costa-Pereira R (2011) Parakeets as facilitators of fish frugivory in the Southern Pantanal. Atualidades Ornitológicas, 169, 15.
- 56. Perry RW, & Thill RE (1999) Estimating mast production: an evaluation of visual surveys and comparison with seed traps using white oaks. Southern Journal of Applied Forestry, 23(3), 164–169.
- 57. Koenig WD, Knops JM, Carmen WJ, Stanback MT, & Mumme RL (1994) Estimating acorn crops using visual surveys. Canadian Journal of Forest Research, 24(10), 2105–2112.
- 58. Nakajima A, Masaki T, Koike S, Yamazaki K, & Kaji K (2015) Estimation of tree crop size across multiple taxa: generalization of a visual survey method. Open Journal of Forestry, 5(07), 651.
- 59. O’Brien TG, Kinnaird MF, Wibisono HT (2003) Crouching tigers, hidden prey: Sumatran tiger and prey populations in a tropical forest landscape. Animal Conservation forum, 6, 131–139.
- 60. Kitamura S, Suzuki S, Yumoto T, Wohandee P, Poonswad P (2009) Evidence of the consumption of fallen figs by Oriental Pied Hornbill Anthracoceros albirostris on the ground in Khao Yai National Park, Thailand. Ornithological Science, 8, 75–79.
- 61. Haurez B, Daïnou K, Tagg N, Petre CA, Doucet JL (2015) The role of great apes in seed dispersal of the tropical forest tree species Dacryodes normandii (Burseraceae) in Gabon. Journal of Tropical Ecology, 31, 395–402.
- 62. Ameztegui A, & Coll L (2015) Herbivory and seedling establishment in Pyrenean forests: influence of micro-and meso-habitat factors on browsing pressure. Forest Ecology and Management, 342, 103–111.
- 63. Wei T, Wei MT (2013) Package “corrplot”. Statistician, 56, 316–324.
- 64. Bolker BM, Brooks ME, Clark CJ, Geange SW, Poulsen JR, Stevens MHH, et al. (2009) Generalized linear mixed models: a practical guide for ecology and evolution. Trends in ecology & evolution, 24(3), 127–135.
- 65.
R Development Core Team. (2018). R: A Language and Environment for Statistical Computing. R Foundation for Stat. Comput., Vienna, Austria.
- 66.
Anderson S (2012) A quick introduction to plyr.
- 67.
Mailund T (2019) Manipulating Data Frames: dplyr. In R Data Science Quick Reference (pp. 109–160). Apress, Berkeley, CA.
- 68. Grolemund G, & Wickham H (2011) Dates and times made easy with lubridate. Journal of Statistical Software, 40, 1–25.
- 69. Palacio FX, Ordano M (2018) The Strength and Drivers of Bird-Mediated Selection on Fruit Crop Size: A Meta-Analysis. Frontiers in Ecology and Evolution, 6, 18.
- 70. Jácome-Flores ME, Jordano P, Delibes M, Fedriani JM (2020) Interaction motifs variability in a Mediterranean palm under environmental disturbances: the mutualism-antagonism continuum. Oikos, 129, 367–379.
- 71. García D, Zamora R, & Amico GC (2011) The spatial scale of plant–animal interactions: effects of resource availability and habitat structure. Ecological Monographs, 81(1), 103–121.
- 72. Moegenburg SM, Levey DJ (2003) Do frugivores respond to fruit harvest? An experimental study of short-term responses. Ecology, 84, 2600–2612.
- 73. Howe HF (1979) Fear and frugivory. The American Naturalist, 114, 925–931.
- 74. Fedriani JM, Boulay R (2006) Foraging by fearful frugivores: combined effect of fruit ripening and predation risk. Functional Ecology, 20, 1070–1079.
- 75. Garrote PJ, Calvo G, Żywiec M, Delibes M, Suárez-Esteban A, Fedriani JM (2018) Strong among population variation in frugivory strength by functional diverse frugivores: a ‘reciprocal translocation’experiment. Oecologia, 187, 143–154. pmid:29497835
- 76. García D, & Martínez D (2012) Species richness matters for the quality of ecosystem services: a test using seed dispersal by frugivorous birds. Proceedings of the Royal Society B: Biological Sciences, 279(1740), 3106–3113. pmid:22456879
- 77. Quitián M, Santillán V, Espinosa CI, Homeier J, Böhning-Gaese K, Schleuning M, et al. (2018) Elevation-dependent effects of forest fragmentation on plant–bird interaction networks in the tropical Andes. Ecography, 41(9), 1497–1506.
- 78. Mulder CPH, Uliassi DD, Doak DF (2001) Physical stress and diversity-productivity relationships: the role of positive interactions. Proceedings of the National Academy of Sciences, 98, 6704–6708.
- 79.
Whittaker RH (1970) “Biochemical ecology of higher plants,” in Chemical Ecology, eds, Sondheimer E. and Simeone J. B. (New York, NY: Academic Press), 43–70.
- 80. Nordlund DA, Lewis WJ (1976) Terminology of chemical releasing stimuli in intraspecific and inteHOLrspecific interactions. Journal of Chemical Ecology, 2, 211–220.
- 81. Parsons MH, Blumstein DT (2010) Familiarity breeds contempt: kangaroos persistently avoid areas with experimentally deployed dingo scents. PLoS One, 5, e10403. pmid:20463952
- 82.
Schaefer HM, & Ruxton GD (2011) Plant-animal communication. OUP Oxford.
- 83. Nielsen B. L., Rampin O., Meunier N., & Bombail V. (2015). Behavioral responses to odors from other species: introducing a complementary model of allelochemics involving vertebrates. Frontiers in neuroscience, 9, 226. pmid:26161069
- 84. Howe HF (1980) Monkey dispersal and waste of a neotropical fruit. Ecology, 61, 944–959.
- 85. Traveset A (1994) Influence of quality of bird frugivory on the fitness of Pistacia terebinthus L. Evolutionary Ecology, 8, 618–627.
- 86.
Herrera CM (2009) Seed Dispersal by Vertebrates. In: Herrera CM, Pellmyr O (eds.), Plant-Animal Interactions: An Evolutionary Approach, pp. 185–208. Blackwell Science, Oxford, United Kingdom.
- 87. Nathaniel Holland J, & DeAngelis DL (2009) Consumer-resource theory predicts dynamic transitions between outcomes of interspecific interactions. Ecology Letters, 12(12), 1357–1366. pmid:19807773
- 88. Wang Y., DeAngelis D.L. & Holland J.N. (2012) Uni-directional interaction and plant–pollinator–robber coexistence. Bulletin of Mathematical Biology, 74, 2142–2164. pmid:22806690
- 89. Fedriani JM, Ayllón D, Wiegand T, and Grimm V. (2020) Intertwined effects of defaunation, increased tree mortality, and density compensation on seed dispersal. Ecography, 43(9), 1352–1363.
- 90. Buckley YM, Anderson S, Catterall CP, Corlett RT, Engel T, Gosper CR, et al., (2006) Management of plant invasions mediated by frugivore interactions. Journal of Applied Ecology, 43(5), 848–857.
- 91.
Waser NM, & Ollerton J (2006) Plant-pollinator interactions: from specialization to generalization. University of Chicago Press.
- 92.
Benbow ME, Tomberlin JK, & Tarone AM (2015) Carrion ecology, evolution, and their applications. CRC press.