Ungulate presence and predation risks reduce acorn predation by mice in dehesas.

Foraging decisions by rodents are key for the long-term maintenance of oak populations in which avian seed dispersers are absent or inefficient. Decisions are determined by the environmental setting in which acorn-rodent encounters occur. In particular, seed value, competition and predation risks have been found to modify rodent foraging decisions in forest and human-modified habitats. Nonetheless, there is little information about their joint effects on rodent behavior, and hence, local acorn dispersal (or predation). In this work, we manipulate and model the mouse-oak interaction in a Spanish dehesa, an anthropogenic savanna system in which nearby areas can show contrasting levels of ungulate densities and antipredatory cover. First, we conducted a large-scale cafeteria field experiment, where we modified ungulate presence and predation risk, and followed mouse foraging decisions under contrasting levels of moonlight and acorn availability. Then, we estimated the net effects of competition and risk by means of a transition probability model that simulated mouse foraging decisions. Our results show that mice are able to adapt their foraging decisions to the environmental context, affecting initial fates of handled acorns. Under high predation risks mice foraged opportunistically carrying away large and small seeds, whereas under safe conditions large acorns tended to be predated in situ. In addition, in the presence of ungulates lack of antipredatory cover around trees reduced mice activity outside tree canopies, and hence, large acorns had a higher probability of survival. Overall, our results point out that inter-specific interactions preventing efficient foraging by scatter-hoarders can reduce acorn predation. This suggests that the maintenance of the full set of seed consumers as well as top predators in dehesas may be key for promoting local dispersal.

Introduction

Scatter-hoarders are key seed dispersers in temperate and Mediterranean forests dominated by oaks [1, 2]. Nut dispersal by scatter-hoarders (synzoochory) is a classical plant-animal conditional mutualism. The outcome of the interaction may be either mutualistic (dispersal) or antagonistic (predation) depending on the proportion of seeds consumed vs. cached and not retrieved [3]. The balance between mutualism and antagonism is contingent on intrinsic properties of interaction partners (e.g. propensity of animals to store food) as well as on the ecological setting in which the interaction occurs [2]. As a result, the net effects of synzoochory can be highly dynamic in space and time, making it difficult to predict its outcomes along environmental gradients and ecological timescales [4, 5].

Acorn dispersal depends on scatter-hoarder corvids and rodents. Corvids disperse acorns tens to hundreds of meters [6], whereas rodents transport acorns locally and a high proportion of them are eventually predated [7, 8]. Nonetheless, several mouse species (Apodemus, Mus, Peromyscus) can become important acorn dispersers in landscapes where scatter-hoarding corvids are absent [9] or become inefficient [10]. Two main external factors modulate mouse foraging decisions: competition for seeds and predation risk [11-13]. Intraspecific competition and the presence of ungulates tend to encourage seed dispersal [3, 14–16]. Especially, when predating seeds in situ is more time-consuming than carrying them away, and hence, scatter-hoarding facilitates stockpiling seeds before they are depleted by competitors [12, 17]. The effects of risk perception on mouse foraging decisions depend on factors that affect exposure to predators (e.g. moonlight) as well as direct cues of their presence (e.g. scent) [18-23]. In general, intermediate risks can promote acorn removal over predation when mice carry away seeds to manipulate them in safer locations or when handling times of consuming seeds in situ are too long [12]. However, if lack of cover in the vicinity of trees triggers predation risks, acorn mobilization distances and caching rates can be significantly reduced [13, 16]. In general, suboptimal conditions for foraging mice (i.e. competition and predation risk) tend to favor scatter-hoarding over in situ predation. In the absence of stress, rodents usually act as efficient seed predators consuming, immediately or soon afterwards, seed crops under the canopy of mother trees [2].

Beyond the environmental conditions of plant-animal encounters, seed size can affect the initial outcomes of the interaction (selected, eaten or cached) as well as post-dispersal processes such as germination and seedling survival. Larger seeds are usually selected and preferentially cached because they provide higher food rewards [7, 24–28]. In addition, seed size enhances post-dispersal seedling survival and establishment [29], which is a key component of dispersal effectiveness [30] in scatter-hoarder animals [31]. Nonetheless, the strength and even sign of acorn size effects on mouse foraging decisions are not unequivocal, but context-dependent. Larger acorns are most preferred when food is scarce [32-34], but may be avoided when longer handling times [25] diminish their profitability [35, 36] or result in unaffordable predation risks during manipulation [11, 12]. Therefore, to have a full picture of mice role in acorn dispersal we need to account for seed size effects on scatter-hoarding decisions as well as the influence of competition and risk.

In this context, dehesas represent an excellent study system to assess the main factors modulating mouse foraging decisions, and hence, acorn dispersal. They are savanna-like habitats, simpler than natural forests but diverse enough to maintain all key elements influencing the mouse-oak conditional mutualism. Depending on the local intensity of management, nearby areas can have contrasting levels of shrub cover and competition with ungulates [18, 37]. In addition, the community of predators is simpler than in forested areas, facilitating the experimental manipulation of direct cues of risks [23]. In this work we take advantage of a large-scale experiment of ungulate exclosure in a Mediterranean dehesa to (1) quantify acorn size effects across different stages of the dispersal process (from seed selection to initial fates after transportation); and (2) evaluate if size effects are consistent across contrasting scenarios of predation risk and competition for seeds. In addition, we parameterized a transition probability model that assembled all scatter-hoarding decisions by mice to quantify and tease apart the net effect of competition and risk on acorn dispersal. We expected that suboptimal conditions for mice (i.e. competition and risk) would constrain their ability to forage efficiently, thus reducing acorn predation.

Methods

Study area and species

Field work was carried out in the holm oak Quercus ilex dehesa woodlands of the Cabañeros National Park (Central Spain, Ciudad Real province, 39°24’ N, 38°35 W). Dehesas are savanna-like man-made habitats resulting from shrub removal and tree thinning and pruning to enhance herb growth for livestock [38]. The studied dehesas were opened through tree thinning from the original Mediterranean forests in the late 1950s. Currently they have no livestock but wild ungulate populations of red deer Cervus elaphus and wild boars Sus scrofa. Deer densities were around 0.14 ind./ha [39] and boars are abundant but at unknown densities [40]. Acorns fall from trees from mid-October to late November [9].

The study area covers around 780 ha, with two sites separated by 1500 m. Average tree density in the area is 20.4 trees ha-1, although it is higher at site 1 (30.05 and 7.4 trees per ha at site 1 and 2, respectively). Tussocks and grasses are the main vegetation cover around trees (94.1%, on average), whereas shrub cover is low (<1%) [23]. At each site there is an ungulate exclosure (150 and 4.65 ha, site 1 and 2 respectively) made with wire fences 2 m tall and 32 cm x 16 cm mesh. The exclosures prevent the entrance of ungulates but not of mesocarnivores (mainly common genets Genetta genetta and red foxes Vulpes vulpes; pers. obs. based on scat searches) and raptors. In addition, ungulate exclosures have modified the structure of the vegetation. Lack of ungulate browsing has resulted in taller vegetation around trees (21.3±1.3 vs 8.3±0.7 cm inside and outside exclosures, respectively) and higher covers of taller resprouts under canopies (30.1±3.4 vs 19.3±2.5% and 79.9±13.4 vs 23.2±3.1 cm, respectively [41]). To evaluate the effects of the presence of ungulates on mouse foraging decisions at each site we established half of focal trees outside the exclosure and half of them inside it (see below). At site 1 we worked in the southernmost 5.72 ha of the site 1 exclosure and in the whole 4.65 ha site 2 exclosure, both paired with a close-by similar area outside the exclosure.

Experimental design

Tree occupancy by mice was established by means of live trapping using Sherman traps (23 × 7.5 × 9 cm; Sherman Co., Tallahassee, USA) baited with canned tuna in olive oil mixed with flour and a piece of apple. Water-repellent cotton was provided to prevent the cooling of the individual captured overnight. Traps were set during two consecutive days during the new moon of January 2012. High capture probability of M. spretus (detectability: 0.88±0.03 SE; [42]) pointed out that false negatives in occupancy was unlikely. Among trees known to be occupied by Algerian mice, we randomly selected ten trees inside and ten outside in each of the two exclosures (40 focal trees in total).

We paired focal trees according to their proximity and we randomly assigned a predator scent treatment to one of them. Predator scent treatment consisted of placing fresh genet feces (10 g) mixed with distilled water close to a corner of the cages where acorns were placed [23]. Genets are generalist predators whose presence and scats are known to influence rodent behavior [20, 22, 43]. Fresh feces were collected from two captive common genets housed in the Cañada Real Open Center (Madrid, Spain).

Fresh acorns were collected from holm oaks growing near the study area in October 2011 and stored dry in a cooler (4°C) until use. Sound acorns, with no marks of insect damage [44], were weighed with a digital balance to the nearest 0.01 g. To offer a full range of acorn sizes, in each cafeteria trial (combination of tree, month and moon light, see below) we randomly selected 5 large (>10 g), 5 medium-sized (5–10 g) and 5 small (1–5 g) acorns. Acorns were placed under the canopy of each focal tree inside a 50 cm × 50 cm x 15 cm galvanized-steel cage to prevent acorn consumption by birds or ungulates [44]. Cages were located 1.2 m on average (range 0.3–2.7 m) from focal tree trunks. A metal wire (ø 0.6 mm, 0.5 m length) with a numbered plastic tag was attached to each acorn [45]. After removing any naturally-present acorns within the cages, we randomly placed acorns in the intersection of a 3 row x 5 column grid. To track mouse choices, acorn size for each position was noted. Acorns were left exposed to mice for three consecutive nights, then removed. Acorns carried away from cages were searched by looking for plastic tags in circles around focal trees (up to 30 m away, where most acorns are initially deposited [27, 46]). Searches were performed during the following days of acorn offering (24 and 72 hours). We tracked the status of acorns that were transported and not predated throughout the experiment. We considered an acorn to be predated when it was either completely consumed (only wire and tag was found, sometimes with husk remains attached) or partially consumed in its apical side thus removing the embryo. To account for changes in night brightness and acorn availability [21, 47], the cafeteria experiment was repeated four times during the full-moon and new-moon periods of November 2011 and February 2012. No official permits or protocol approvals were legally necessary since we did not manipulate individual mice except for checking whether trees were occupied or not by means of live traps. We followed Guidelines of the American Society of Mammalogists for the use of wild mammals in research [48]. We performed all manipulations with disposable latex gloves, to avoid effects of human odor on rodent behavior [49].

Mouse foraging behavior

A video-camera OmniVision CMOS 380 LTV (OmniVision, Santa Clara, USA) (3.6 mm lens) monitored mouse foraging activity within each cage. Cameras were set on 1.5 m tall tripods located 2.5 m from each cage, powered by car batteries (70 Ah, lead acid) connected to a solar panel (ono-silicon erial P_20; 20 w). Video-cameras were connected to ELRO recorders with dvr32cards (ELRO, Amsterdam, Netherlands) and took continuous recording for three consecutive days autonomously (recorded in quality at 5 frames s-1). Events with rodent activity, from the entry of the individual into the cage up to the exit from it, were located and separated using Boilsoft Video Splitter software (https://www.boilsoft.com/videosplitter/) [43]. Within each foraing event, we noted which acorn was handled (selection) and if it was removed outside the cage or not. For removed acorns we measured transportation distances (cm) and noted its status (predated or not after transportation).

Data analysis

To assess acorn selection by rodents, we fitted a hierarchical multinomial model. For each foraging event, we modeled which acorn was handled (out of those available in the cage) as a function of acorn size (g), moon phase (new/full), month (February, November), ungulate presence (yes/no), predator scent (yes/no), acorn availability in the cage (g) and the two-way interactions between size and environmental effects. Local acorn availability was measured as total acorn mass in the cage during the event. Both acorn size and availability were scaled previous to the analyses (mean = 0, SD = 1) so that we could compare the magnitude of covariate effects. Focal tree was introduced as a random factor in the intercept term to account for repeated sampling during the experiment. Subsequently, we evaluated the effects of acorn size, competition and risk on the foraging decision of carrying acorns outside the cage or not (acorn removal, hereafter). To this end we used a hierarchical logistic model. Our response variable was acorn removal (yes/no). Our explanatory variables and random effects were the same as in the multinomial model.

Finally, we analyzed the effect of acorn size and environmental covariates (and their two-way interaction) on seed removal distances and initial fates. Our response variables were transportation distances of acorns (cm, log-transformed) and deposition status (viable or predated). We used a hierarchical Gaussian model in the former case, and a hierarchical logistic model in the latter. Our explanatory variables and random effects were the same as in the previous models. In all four models (acorn selection, removal, transportation distance and deposition fate) we used uninformative priors (S1 File). All analyses were performed employing a Bayesian approach with JAGS 3.4.0 [50]. We checked for convergence for all model parameters (Rhat < 1.1) and that the effective sample size of posterior distributions was high (>800). We estimated the mean and credible interval of posterior distributions, calculated the proportion of the posterior distribution with the same sign of the mean (f) and evaluated the predictive power of our models by means of posterior predictive checks (S1 and S2 Files).

Simulating scatter-hoarding decisions

To estimate the joint effect of seed size, competition and risk on acorn dispersal we designed a probability transition model in which simulated mice adapted their foraging behavior to the environmental context (S3 File). Before model run, we parameterized mouse scatter-hoarding decisions (from acorn selection to initial fate of transported acorns) following the same scheme of regressions explained in the previous section. Here, we only used data from November, the period of peak acorn falling in our study system. Thus, we did not include month as a covariate. For each behavioral submodel (selection, removal and initial fate), we obtained posterior distributions of parameters by running 50000 iterations in three chains (in all cases Rhat< 1.1, and Neff> 1000).

Model setup mimics our experimental design, 20 trees outside and 20 inside exclosures paired according to a predator scent treatment (presence vs. absence). Simulations begin under new moon conditions with focal trees offering 15 acorns of large, medium and small sizes (5 each). Acorn size is sampled from empirical distributions of these size categories. In each focal tree, the number of foraging events is drawn from a Poison distribution with mean equal to the average number of events observed in the corresponding moon phase. During each foraging event, simulated mice decide which acorn to handle and whether to remove it or not. If removed, mice decide to predate it or not after mobilization and acorn availability in the cage is updated. Once all foraging events (of all trees) are simulated, acorn dispersal is modelled under full moon conditions (S1 Fig in S3 File).

For each model run we sampled parameter of behavioral submodels (selection, removal and deposition) from posterior distributions fitted to data. Thus, in our simulations, mice adapted their decisions to acorn size and availability (in the experimental cage), characteristics of the focal tree (i.e. ungulate and predator scent presence), and the moon phase in which the foraging event occurs (new or full moon). After each model run (simulated mice foraging under new and full moon conditions), the program tracked the size and status of handled acorns and the environmental covariates in which the foraging event occurred. We run the model 1000 times and plotted deposition rates of viable acorns and their size with respect to the moon phase and tree characteristics (predator scent and ungulate presence). See S3 File for detailed model specifications and S1 Fig in S3 File for a summary of the process overview.

Results

Before setting the cafeteria experiments in November, we removed from cages 53.3 acorns/m2 on average (range: 0–104). No acorns were found in February. We detected Mus spretus activity in 18 and 26 trees in the new and full moon of November; and in 26 and 24 trees in February. Therefore, we finally monitored 1410 acorns instead of 2400 (40 initial focal trees x 15 acorns per trial x 2 months x 2 moonlight conditions). Mice (M. spretus) handled 986 acorns (69.5% of those offered). Out of them, 288 (29.2%) were carried outside cages and 211 (73.2%) were relocated, 67 of which (31.8%) were not predated after transportation, and 8 out of these (11.9%) were found buried.

Foraging decisions in the focal tree: Selection and removal

In general, mice preferentially handled larger acorns, but the positive effect of size was modulated by environmental conditions. Size-driven selection occurred in the absence of competition with ungulates (Fig 1A) and predator scent (Fig 1B). In addition, mouse selectivity was enhanced under low local acorn availability (Table 1, selection). Among handled acorns, mice preferentially removed smaller ones outside the cages. Such behavior occurred when risks were low due to reduced night brightness (new moon, Fig 1C) or lack of predator scent (Fig 1D), as well as when ungulates were absent (Table 2). Acorn availability at local and landscape scales did not modify size effects, although they changed removal rates. They were lower during the acorn fall peak (13% in November vs 24% in February), whereas local acorn availability (i.e. in cages) enhanced removal (Table 2, removal).

Fig 1

Mouse foraging decisions during acorn selection and removal (upper and lower panels, respectively).

Size of handled acorns in the presence or absence of (A) ungulates and (B) predator scent. Size of acorns (removed away from the cage or not) (C) under new or full moon conditions and (D) in the presence or absence of predator scent. Point colors depict whether the acorn was selected or removed (yes, blue) or not (no, orange). In all cases acorn size is expressed in grams. Points represent mean values, bars standard errors (N = 1677 foraging events).

Table 1

Summary table of the effects of size, moonlight, month, ungulate presence, predator scent and local acorn availability (and their interactions with size) on the probability of acorn selection and removal.

A total of 1677 foraging events were analyzed.

Process	Fixed effect	Mean	HPD1	f
Acorn selection	Size	0.19	[0.09, 0.29]	1.00	**
	Moon (Full)	0.03	[-6.22, 6.35]	0.50
	Month (February)	-0.03	[-6.12, 6.06]	0.50
	Ungulate (Yes)	-0.05	[-6.27, 6.16]	0.51
	Scent (Yes)	0.01	[-6.16, 6.28]	0.50
	Availability	-0.02	[-6.32, 6.19]	0.50
	Size*Moon	0.07	[-0.03, 0.17]	0.93
	Size*Month	-0.06	[-0.16, 0.04]	0.88
	Size*Ungulates	-0.13	[-0.23, -0.03]	0.99	**
	Size*Scent	-0.08	[-0.18, 0.01]	0.96	*
	Size*Availability	-0.04	[-0.09, 0.01]	0.95	*
Acorn removal	Size	-0.50	[-0.94, -0.07]	0.99	**
	Moon (Full)	0.07	[-0.27, 0.39]	0.65
	Month (February)	0.77	[0.43, 1.11]	1.00	**
	Ungulate (Yes)	-0.22	[-0.96, 0.47]	0.73
	Scent (Yes)	0.20	[-0.53, 0.93]	0.72
	Availability	0.29	[0.12, 0.46]	1.00	**
	Size*Moon	0.29	[-0.02, 0.60]	0.96	*
	Size*Month	-0.09	[-0.41, 0.22]	0.71
	Size*Ungulates	0.24	[-0.07, 0.55]	0.94	∙
	Size*Scent	0.30	[0.00, 0.59]	0.98	*
	Size*Availability	0.10	[-0.07, 0.26]	0.88

Mean of posterior distribution, highest posterior density interval (HPD) and percentage of the posterior distribution with the same sign as the mean (f) are shown. Effects with f ≥ 0.95 are in bold. • depicts f ∈ [0.90, 0.95)

Table 2

Summary table of the effects of size, moonlight, month, ungulate presence, predator scent and local acorn availability (and their interactions with size) on acorn mobilization distances and the probability that it is deposited in a viable status (vs predated).

A total of 211 acorns that were mobilized outside cages and retrieved were analyzed.

Process	Fixed effect	Mean	HPD	F
Mobilization distance	Size	0.16	[-0.51, 0.83]	0.68
	Moon (Full)	-0.67	[-1.27, -0.06]	0.98	*
	Month (February)	0.54	[-0.14, 1.2]	0.94	∙
	Ungulate (Yes)	-0.75	[-1.59, 0.14]	0.95	*
	Scent (Yes)	0.09	[-0.73, 0.98]	0.57
	Availability	-0.01	[-0.31, 0.29]	0.52
	Size*Moon	-0.07	[-0.66, 0.49]	0.60
	Size*Month	-0.33	[-0.94, 0.28]	0.86
	Size*Ungulates	0.18	[-0.44, 0.81]	0.71
	Size*Scent	0.22	[-0.36, 0.79]	0.78
	Size*Availability	-0.16	[-0.49, 0.17]	0.83
Viability after deposition	Size	-1.20	[-2.15, -0.33]	1	*
	Moon (Full)	0.42	[-0.38. 1.22]	0.85
	Month (February)	-1.58	[-2.46, -0.75]	1	*
	Ungulate (Yes)	0.67	[-0.32, 1.69]	0.91	∙
	Scent (Yes)	-0.16	[-1.14, 0.80]	0.63
	Availability	0.52	[0.10, 0.97]	0.99	*
	Size*Moon	0.66	[-0.12, 1.46]	0.95	*
	Size*Month	0.49	[-0.33, 1.34]	0.88
	Size*Ungulates	0.59	[-0.22, 1.40]	0.92	∙
	Size*Scent	0.21	[-0.52, 0.94]	0.72
	Size*Availability	-0.24	[-0.73, 0.25]	0.84

Mean of posterior distribution, highest posterior density interval (HPD) and percentage of the posterior distribution with the same sign as the mean (f) are shown. Effects with f ≥ 0.95 are in bold. • depicts f ∈ [0.90, 0.95).

Foraging decisions after removal: Distances and predation after deposition

Mice transported acorns shorter distances under new moon conditions (Fig 2A) and when ungulates were present (Fig 2B). During lean periods (February) transportation distances and post-dispersal predation increased (Table 2, Month). In addition, larger acorns were preferentially predated (Fig 2C), though the presence of ungulates and full moon conditions attenuated this negative effect (Fig 2D, Table 2). Regarding the microhabitat of deposition, viable acorns were frequently found under tree canopies or close to oak resprouts (96.8% and 97.9% of acorns, inside and outside exclosures, respectively). Only 2.4% of transported acorns were deposited in open areas.

Fig 2

Mouse foraging decisions during transportation and after deposition (upper and lower panels, respectively).

Removal distances under (A) new and full moon conditions and (B) in the absence and presence of ungulates. (C) Seed size effects on the probability of predation after deposition (black line represents mean effects and shaded area 0.95 credible intervals). (D) Proportion of acorns escaping predation after deposition in the absence and presence of ungulates. Sample size was 267 for mobilization distances and 211 for analyses of initial seed fate.

Transition probability model for acorn dispersal

Under optimal conditions (new moon, no predator scent or ungulates), post-dispersal predation rates were higher (Fig 3A) and simulated mice preferentially consumed large acorns (i.e. viable acorns -blue bars- were smaller, Fig 3B–3D). In contrast, predation risks and ungulate presence precluded acorn predation after mobilization and attenuated selection. As a result, the proportion of viable acorns increased and they were larger (Fig 3A–3D).

Fig 3

Results from simulations of the probability transition model for acorn dispersal.

(A) Proportion of acorns escaping predation in the presence or absence of environmental stressors (i.e. full moon, ungulates present or predator scent, black bar) in comparison to more optimal conditions (i.e. new moon, ungulates absent, no predator scent, grey bar). Size of predated (yellow) and viable (blue) acorns under (B) new vs full moon conditions and in the presence or absence of (C) ungulates and (D) predator scent. Bars represent mean values (±s.e.) across 1000 simulations.

Discussion

Overall, our work shows that mice are able to adapt their foraging decisions to the presence of ungulates and perceived predation risk, and that such behavioral adjustments affect the fate of acorns at initial stages of the dispersal process. When not exposed to stressful factors, mice preferentially consumed in situ large acorns and carried away small ones. Furthermore, seeds were more likely to be predated after deposition. In contrast, under stressful conditions (increased predation risk and ungulate presence) mice foraged opportunistically and reduced their activity outside tree canopies. As a result, predation rates of seeds decreased, and larger acorns had a higher probability of survival, at least in the short term. This bolsters the idea that interactions with third-party players can modify the qualitative component of dispersal effectiveness of scatter-hoarding rodents [12, 15, 51].

As expected, larger and more valuable acorns were preferentially handled by mice, which adapted this behavior to the environmental context [12]. In line with previous work, mice foraged opportunistically in trees with predator scent, probably because they devoted more time to vigilant behaviors [15, 43] at the expenses of acorn discrimination [21]. In contrast, acorn availability effects did not follow the expectations of increased selectivity in scenarios of food depletion or competition [27, 51, 52]. Seed size effects were similar between acorn fall peaks and lean periods. In addition, mice foraged randomly in the presence of ungulates, while selected larger seeds in their absence. These unexpected results may respond to some particularities of our study system. On one hand, dehesas are characterized by a high acorn production [53, 54], and hence, the effects of competition for seeds may have been attenuated [13]. On the other, ungulates not only compete with rodents for acorns, but also are important modulators of vegetation structure in dehesas. Outside exclosures, grazing and trampling by ungulates has led to scarcer and shorter vegetation around focal trees, whereas inside exclosures tall resprouts, grasses and tussocks can be found (S1 Fig in S5 File). Such changes in vegetation structure allow mice to forage under shelter, devoting less time to vigilant behaviors [43], and hence, selecting the most profitable food items. In line with previous work, our results suggest that predation risks rather than competition modulate mouse foraging decisions in dehesas [43]. Also, that the effects of ungulate presence on vegetation structure can strongly affect the foraging behavior of scatter-hoarding rodents [17].

Larger acorns tend to be carried away, transported farther and preferentially cached in forest habitats [7, 26, 55, 56]. However, in our study larger acorns had a higher probability of being predated (in situ and after transportation) and seed size did not affect transportation distances. Again, these results highlight that in dehesas environmental conditions are particularly harsh for rodents. In general, rodents preferentially carry away small seeds when the costs of transporting large ones are unaffordable [12]. In the presence of ungulates, low antipredatory cover and high trampling risks may have triggered transportation costs [13, 57], deterring mice from carrying large seeds away. Seed size effects were not fixed, but depended on direct and indirect cues of risk. Preferential removal of small seeds only occurred in trees with no predator scent or under new moon conditions, reflecting that only when risks are reduced mice can take the time to select among the seeds available [15, 21].

Regarding initial seed fate, we expected higher predation rates when acorns were deposited close to tree canopies [13, 16]. Nonetheless, this relationship blurred in our system. In the presence of ungulates, larger acorns had a higher probability of escaping predation in spite of being mobilized nearby source trees. In dehesas, outside ungulate exclosures the pervasiveness of open land cover forces mice to concentrate their activities beneath canopies [13, 23, 41], and decreases the likelihood that mobilized acorns are encountered and consumed [58]. Accordingly, in our simulations, suboptimal conditions (due to increased risks or ungulate presence) discouraged mice from selecting which acorns to handle and carry away, and from consuming seeds after transportation. Consequently, predation rates were reduced and larger acorns had a higher probability of survival. In principle, these results suggest that intermediate levels of stress can enhance the probability of acorn dispersal by rodents (as suggested by [59]). Nonetheless, a high proportion of acorns were deposited within resprouts growing under tree canopies, where the establishment of a new seedlings is highly unlikely [60]. Therefore, it remains an open question whether mice can act as local acorn dispersers in dehesas by outweighing a very low dispersal quality with high removal rates (as found in forest habitats [29]).

This work builds on previous research analyzing the effects of competition and risk on mouse foraging behavior in dehesas [43]. In the present study, by accounting for many stages of the scatter-hoarding process (from initial acorn selection to predation after transportation [57]) and including the entire acorn fall season [26] as well as contrasting moon light conditions [21], we obtained a more in-depth understanding of the role of mice as acorn dispersers (or predators) in dehesas. Overall, our results show that suboptimal conditions for mice can reduce predation rates and increase the probability that larger ones survive acorn handling. Also, they suggest that in dehesas the effects of ungulates on mouse foraging decisions are mediated by their impacts on vegetation cover rather than by competition. Nonetheless, low caching rates (<1%) prevented us from analyzing scatter-hoarding (in spite of tracking 1410 acorns). Such difficulties are commonplace in dehesas, where caching rates by mice are low (reported values lie between 1.83% [13] to 7.52% [58]). In this context, mechanistic models like ours result particularly useful [61]. They allow to simulate a high number of foraging events, and hence, monitor the fate of those that are rare but important from a demographic perspective (e.g. survival of cached acorns). However, to quantify seed dispersal effectiveness by mice (sensu [30]), our model needs information about caching rates and long-term survival. To obtain robust estimates of caching rates, a higher number of acorns could be tracked at the expense of simplifying the number of environmental factors being evaluated (e.g. only ungulate presence) and of not videorecording foraging events. In the case of cache survival rates, we believe that sowing acorns and monitoring artificial caches seems the only way to achieve adequate sample sizes. Although such approach does not allow to evaluate the foraging decisions made by cache owners, it informs about survival rates of cached acorns from pilferers and ungulates as well as the probability of emergence and one-year survival [60, 62]. Once this information is available, it could be easily included in our transition probability model. This version of our model will be able to inform if changes in their short-term mouse foraging decisions modulated by ungulate presence and predation risks have an imprint on seedling recruitment.

Concluding remarks

Our mechanistic approach provides new insights about the joint effect of habitat structure, competition and risk on the foraging behavior by scatter hoarders and its potential consequences on acorn dispersal. In the presence of ungulates and when predation risks were high, mice acted as opportunistic foragers and concentrated their activities beneath tree canopies. As a result, predation rates decreased and larger acorns had a higher probability of survival (at least in the short term). These results suggest that inefficient foraging by mice can reduce acorn predation and may promote dispersal. Also, they highlight the importance of competition and risk as modulators of the spatial and temporal dynamism of oak-rodent interactions [2]. Finally, though future work is needed to estimate long-term cache survival and seedling establishment, our results support the view that the presence of the full set of acorn consumers and top predators can facilitate seed dispersal effectiveness in conditional mutualisms [2, 15, 51]. This may be particularly important in habitats like dehesas, which depend on scatter-hoarders to ensure their long-term sustainability [53, 63]

(XLSX)

Click here for additional data file.

Structure of models and priors.

(DOCX)

Click here for additional data file.

Posterior predictive checks.

(DOCX)

Click here for additional data file.

Specifications of transition probability model for acorn dispersal.

(DOCX)

Click here for additional data file.

Code for the transition probability model.

(DOCX)

Click here for additional data file.

Changes in vegetation structure in dehesas.

(DOCX)

Click here for additional data file.

16 Dec 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Pedro G. Blendinger, Dr

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Methods section, please provide additional information regarding the permits you obtained for the work. Please ensure you have included the full name of the authority that approved the field site access and, if no permits were required, a brief statement explaining why.

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

D. López, M. Fernández and C. L. Alonso helped during fieldwork. D. López, B. Ramos and M. de Pablo pre-processed the video recordings, and D. Gallego, D. Valero, A. Velasco, C. J. González and E. Sánchez visualized the recordings noting seed choices. Authorities of the Cabañeros National Park provided the official permissions to carry out field experiments. J. España provided common genet scats. This study is a contribution to the projects RISKDISP (CGL2009-08430) and VULGLO (CGL2010-22180-C03-03), funded by the Spanish Ministry of Economy, and REMEDINAL3-CM (S2013/MAE-2719), funded by the Autonomous Community of Madrid. We declare no conflict of interest.

We note that you have provided funding information. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

MD. Grant number S2013/MAE-2719. REMEDINAL3-CM project, funded by the Autonomous Community of Madrid. NO

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This study uses an experimental approach to investigate how the exclusion of ungulates and the presence of predator cues affects the removal, transportation, and short-term fate of Quercus ilex acorns handled by Algerian mice. The results show that the mice preferred large acorns for immediate consumption, but that this selectivity was changed by the presence of ungulates or predator cues, as well as the availability of more acorns. Acorns that were removed for caching were generally smaller, but this preference was changed in experiments with predator scents or ungulates present. Caching distances were generally very short and further reduced by the presence of full moon and outside of ungulate exclosures. Finally, the immediate post-dispersal predation of acorns was determined by acorn size, timing, and the presence of ungulates.

Based on their findings, the authors discuss how “environmental stress”, exerted by potential competition with ungulates and predation pressure, simulated by scents, moves the mouse-oak interaction from the predation end of the seed fate spectrum towards mutualism. This interpretation is hard to follow, based on the presented data, as the differences in “dispersal” distance (maybe a mean difference of 10cm) are biologically irrelevant and the time span over which post-dispersal predation is monitored does not suffice to make inferences about its effects on seed fate that leads to plant establishment. Therefore, I’d suggest to focus the discussion more on the decisions the mice face and how these are influenced in the experiment, rather than suggesting that this affects the whole ecosystem. Likely, the most important drivers are the presence of shrub cover, water availability/drought, and herbivory by large ungulates. Nonetheless, the experiment is interesting in teaching us about decisions that rodents make when handling seeds under different conditions. Therefore, I suggest the authors shift the discussion towards the behavioral ecology of the rodents, and away from effects of competition and predation on dispersal effectiveness.

L48: “seed dispersers” – otherwise the readers may think of natal or breeding dispersal of the hoarders themselves. Generally, please make sure to refer to “seed dispersal”, rather than simply “dispersal” (eg L65) for clarity

L48: To avoid the awkward term “acorn-bearing trees”, you could rephrase to “Scatter-hoarders are key seed dispersers in temperate and Mediterranean forests dominated by oaks [1-5]”

L56: “...space and time, making it difficult to predict...” (add comma and “it”)

L58: here and throughout: unless referring specifically to multiple individuals, please consider using the singular “mouse” rather than “mice” (e.g. “oak-mouse” in abstract etc).

L61: Please explain to the reader why competition encourages seed mobilization (just saw it on L66; consider moving that part forward a bit)

L58-72: While I understand the need to keep it short, it seems that the dynamics that turn a potential seed predation event into a seed dispersal event are oversimplified. Transportation distance alone does not make the interaction more antagonistic or mutualistic. While I know that not all aspects of seed fate can easily be quantified, it should at least be noted here that not only seed transport, but also consumption after caching can result in a seed predation event.

L123: please provide latin species name for the Algerian mouse

L127: do the mice live in or below the trees? (ie are they arboreal?) Consider rephrasing to “Mouse occupancy below target trees was established...”

L136: “consisted of...”

L151: For how many days did you search? This is really important for your definition of viability. Consider the fact that cached acorns may be retrieved and consumed weeks later.

L179: “acorn availability (g)” is this the natural crop of the tree or the overall mass of acorns provided to the mice during the experiment?

L182: scaling: please describe how and why

L183: “..intercept term to account for repeated sampling.”

Figures: Consider using “predator scent” to label the plots. Also, please add sample sizes to captions.

Figures 1 & 3: Does the color add any information? Otherwise consider making all figures black and white and differentiating yes/no with symbols or grayscale colors (eg Fig 3A)

L252: The last sentence reads as if it contradicts itself (more mobilization when no food, but more when lots of food). Consider rephrasing

L274: “environmental stress” is a very broad and loaded term, which could mean temperatures, etc. Please be concise

L280: “....probability to survive the first days after caching.” Since the whole point of caching is subsequent consumption, which may occur long after caching, I would be careful with this interpretation. However, by providing the time period over which you monitored seed fate, you can make this statement more accurate.

L282: Please expand on this notion of intermediate stress. The reference you provide in the introduction, Lichti et al. Biol Rev, only discusses the effect of intermediate seed availability on seed dispersal by rodents, but not overall stress. In the context of your study, you only compared the presence and absence of putative stressors, thus not providing quantitative support for the role of “intermediate stress”. Also, the argument that the recovery of cached seeds is affected slightly misses the point. Isn’t it about the moving of seeds in the first place? Why would the mouse take the risk to cache under high predation pressure, but then avoid that risk during recovery? In general, I think the “viability” argumentation is very limited by the time period over which the transported acorns were monitored (a few days, I assume).

L300: “Risk rather than competition modulates the effect of ungulate presence on acorn selection”. It seems to me that table 1 shows the opposite. The effect size of ungulate presence (size * ungulate) is nearly twice that of size*scent. Why wasn’t the interaction scent * ungulate included? It seems the experiment would be optimal to test the interaction between the two putative drivers of acorn selection.

L307: rephrase, hard to follow

Reviewer #2: General comment not to the Authors: "PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here."

I have not proof-read the paper.

The study has a very interesting aim which is to explore whether biological integrity of the system can have positive effects on scatter hoarding and thus regeneration of the keystone oak. Results are generally nicely written-up, but figures need revisions. I have some comments regarding presentation, with the hope that these revisions can make the reading of the study easier.

Well explained experimental design – 40 trees, 10 assigned to a combination of herbivore exclusion crossed with genet scent addition.

Data analysis: sample size for the analysis needs to be provided, as it appears like model can be overfitted: 6 fixed effects + many interactions.

I am not familiar with Bayesian framework, so pardon the question. How the effects can be considered meaningful (or significant, but it is not about semantics) if their 95% CI overlap zero?

Figures needs major revisions. First, All figure captions should explain what is being showed, not provide the interpretation. The current version of Fig 1 is difficult to interpret, and the figure caption does not help as instead explaining what is shown, it presents the interpretation. Acorn size was categorial here with three levels, yet only two are shown. Then, if the acorn size was category, what is the point of showing it at y-axis which should be the place of response variable – I guess we are here mostily interested in probabilities? I suggest that all panels at Figure 1 show Probability at y-axis, acorns size at x-axis, and how the probability of each process changes with size depending on the treatment (ungulates, scent, moon etc).

Boxplots at Fig 2 should include data points in the background.

L50: The net outcome of the interaction does not depend on whether seeds are consumed or cached, as seeds are usually both consumed and cached in each interaction. The key is the balance between predation and dispersal, and the balance of the benefit (improved recruitment) vs cost (predation and thus reduced recruitment).

L58: This sentence suggests that corvids are inefficient dispersers in oaks savannas, yet we have papers showing the contrary both in dehesas (Baroja et al: https://besjournals.onlinelibrary.wiley.com/doi/abs/10.1111/1365-2745.13642) as well as in oak savannas in other regions (e.g. Pesendorfer et al, papers from California).

L70: “In the absence of stress… “ this sentence is oversimplification and sounds like rodents never store seeds in the absence of stress, which is not true. Simply put, if there is lots of food (no stress) we do expect that rodents will start to store.

L112: what does it mean that they were opened?

L121: averages of both areas are needed to support a statement that they have similar tree abundance

Reviewer #3: This is a comprehensive study as to the joint effect of habitat structure, competition and predation risk on dispersal effectiveness in an oak-mice system. They found that intermediate stress (presence of predator or grazer) could increase dispersal effectiveness and then facilitated interaction towards the mutualistic side, providing new evidences on conditional mutualism. I think this is a good contribution to the study of the field. I have only a few revision suggestions, mainly by including a few previous similar studies:

Line: 147-148. The wire-linked plastic seed tagging method was proposed by Xiao et al. (2006). You need to include the original reference: Xiao et al, 2006, Forest Ecology and Management, 223:18–23；

Line 296-299. You found ungulate presence would benefit dispersal of oak acorns. A previous study had similar results. You should include the reference in the discussion: Zhang et al. (2009), Wildlife Research, 36: 610–616;

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

14 Feb 2022

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Done

2. In your Methods section, please provide additional information regarding the permits you obtained for the work. Please ensure you have included the full name of the authority that approved the field site access and, if no permits were required, a brief statement explaining why.

Done. No offcila permits were necessary, as we explained in the former version, now developed in lines 169-173

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

D. López, M. Fernández and C. L. Alonso helped during fieldwork. D. López, B. Ramos and M. de Pablo pre-processed the video recordings, and D. Gallego, D. Valero, A. Velasco, C. J. González and E. Sánchez visualized the recordings noting seed choices. Authorities of the Cabañeros National Park provided the official permissions to carry out field experiments. J. España provided common genet scats. This study is a contribution to the projects RISKDISP (CGL2009-08430) and VULGLO (CGL2010-22180-C03-03), funded by the Spanish Ministry of Economy, and REMEDINAL3-CM (S2013/MAE-2719), funded by the Autonomous Community of Madrid. We declare no conflict of interest.

We note that you have provided funding information. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

Done

MD. Grant number S2013/MAE-2719. REMEDINAL3-CM project, funded by the Autonomous Community of Madrid. NO

4. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Done

Submitted filename: MoranLopez_etal_BiolIntegrity_Response_to_Reviewers.docx

Click here for additional data file.

25 Apr 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Pedro G. Blendinger, PhD

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #4: (No Response)

Reviewer #5: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #4: Partly

Reviewer #5: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #4: N/A

Reviewer #5: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #4: Yes

Reviewer #5: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #4: Yes

Reviewer #5: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #4: The manuscript titled ‘Biological integrity of dehesa ecosystems favors acorn dispersal over predation in the mouse-oak mutualism’ explores the effects of predation risk cues and competition on seed size selection by rodents. The manuscript generally reads well but does not present clear novel idea. However, it combines few earlier ideas into more comprehensive study which presents complex interactions in dehesa system. Authors applied simple, commonly used both field and statistical methods presented in the Manuscript. The manuscript has been noticeably improved after previous review rounds. However, I still suggest a major revision before publication and provide some comments.

Title: I do not think that a term “biological integrity” reflects what has been presented in the study. Biological integrity represents the capability of supporting and maintaining a balanced, integrated, adaptive community of organisms having a species composition, diversity, and functional organization comparable to that of the natural habitat of the region before human alterations were imposed. I do not find such comparisons in the manuscript. If Authors still consider this term as somehow appropriate in the manuscript, please describe it more thoroughly in both introduction and discussion. If not, please change the title into more informative. Short title/Running headline:, i.e. “Predation and competition favor dispersal in mouse-oak interaction”, fits much better.

Authors use rarely used terms in seed dispersal studies: ‘select’, ‘mobilize’, ‘mobilization’, ‘transportation’ etc. What is the differences between “seeds selected” vs “seeds removed” (e.g. Fig 1)? Seeds selected include both those consumed in situ and those removed while seeds removed – only those taken away from the seed stations? I highly suggest to use commonly used terms, constistently throughout the text what makes interpretation much easier (not only less confusing but also increasing the probability of finding the article in searching results): ‘handle’/’handling’, ‘remove’/’disperse’, and ‘removal’/’dispersal’, etc. Otherwise, it creates a great confusion.

Line 122: It is not clear how many sites there are in total. Two with two subplots (one open and one exclosure) within each? Four separate plots?

Lines 129-130: Authors have mentioned that plant community structure is similar on both open and excluded areas. However, in Lines 315-316, they have assumed that obtained differences in rodent foraging can be caused by changes in vegetation cover and shrub layer. So, were there differences observed or not? If so, add some information about such changes, thus, I think ungulates may act functionally as vegetation changers rather than competitors in this system.

In the results, there is no information of:

- rodent diversity and abundances based on live-trapping and video recordings. Authors do not present any data confirming that seeds were handled by Algerian mice only while merely suggest that “The Algerian mouse (Mus spretus) is the most abundant scatter-hoarding rodent in the area [44] (…)”. This information seems to be crucial since seed:rodent ratio strongly affects rodent decisions. However, I assume Authors did not mark individuals (lines 169-171) so they can only provide information regarding rodent activity – while this not really reflects abundances and is a bit tricky: few bold individuals can be very active and they can provide a picture of higher abundances while shy individuals can be recorded or captured only once.

- numbers/percentages of acorns occurred in each category of seed removal process (seeds removed, consumed in situ, consumed after removal, cached etc.),

- three-way interactions, e.g. Size*Ungulates*Month. I am pretty sure that foraging activity (thus, the effect) of ungulates may depend on season.

I suggest to add these data to the results. Especially that, I suppose, these may differ between two seasons.

Lines 153-154: I wonder whether the results would be similar if Authors treated “seed size” as a numerical instead of categorical factor. Nature usually does not use such clear categories. Additionally, a term “seed size” is not suitable here as “size” refers to other measurements, such as length, width or volume, while Authors are focused on seed mass [g] instead. I suggest to change this nomenclature.

Line 193: What actually does ‘seed availability’ mean in this study? It is not explained in the text. Is it about naturally occurring acorns? Experimental acorns left in the seed stations? How was this measured?

Lines 203-204: Were distances measured for all seeds removed from seed stations (both consumed and cached)? “Seed dispersal distances” should be used and analyzed for cached seeds only, while “seed removal distances” can be analyzed for all seeds. If all seed distances were analyzed altogether ,“seed fate” should be added as a fixed factor to reveal differences in removal distances between cached and consumed seeds. Otherwise it is a bit confusing. Usually, consumed seeds are removed on shorter distances than cached so adding consumed seeds into analysis can alter the results.

Lines 252-253: The information regarding number of seeds as well as trees used should be provided in methods.

Lines 253-254: What about exclusures vs. controls? Was mouse activity altered by predator scent?

Lines 255-256: The diagram showing the whole seed fate process is clearly needed – this should include all numbers of seeds in all stages. Otherwise it is hard to follow throughout the results section. Please also add percentages to all numbers. What about missing seeds? Using seed-tagging method always provides some seeds that cannot be found.

Lines 271-272: Could removal distances vary between exclosures vs. controls (the effect of ungulates) due to differences in vegetation structure? For example, if ungulate activity leads to vegetation surrounding the focal trees (because it is normally grazed in open spaces) then removal distances will be shorter (because vegetation is closer to focal trees in open spaces/controls while more scattered in exclosures). Moreover, did Authors check for microhabitats chosen for caching in exclosures vs. controls?

Line 288: Please change ‘relaxed’ into more appropriate terminology, such as ‘not exposed to stressful factors’. I assume mice can never be truly relaxed.

Reviewer #5: The combination of including predation risk with ungulate exclosures in a disturbed landscape makes for an interesting study with novel results. This research provides important insights regarding rodent foraging decisions in anthropogenic habitats which can inform oak woodland restoration.

Overall:

The use of the words “mobilized” and “mobilization” instead of “dispersed” and “dispersal” (especially when “dispersal” is used in the title) is confusing. Maybe the authors don’t consider movement away from the parent plant “dispersal” unless seedling recruitment is the final fate? Whatever the reason, there is a whole body of dispersal literature that uses the term “dispersed” or “dispersal” to describe seed movement away from the focal plant. Also, due to the definition of mobilization, the term can carry with it a military connotation and doesn’t seem to be the best term to replace “dispersal” if that was the authors’ intent (mobilization: 1. the action of a country or its government preparing and organizing troops for active service; 2. the action of making something movable or capable of movement). Trying to make the manuscript less repetitive by using “mobilization” or “movement” every so often is understandable, but it is overused in the manuscript and shouldn’t be used along with “distance” since “dispersal distance” is a pretty well-established term. The seed dispersal literature is a quite large body of work at this point with its own established terminology that creates continuity, so it’s confusing and unnecessary to use new terminology without any justification.

The Methods and Results are lacking some important details for a study with “dispersal” in the title. How far from the cages did the search area for dispersed acorns extend? What was considered “predated”? Were acorn fragments found or tags left behind? Since dispersed acorns were tracked “throughout the experiment,” was any secondary dispersal of cached acorns documented? If tagged acorns were removed and never found, could they not have been dispersed outside of the search area? 21% of removed acorns were not relocated but nothing is mentioned in the Results or Discussion about their potential fate. How many of those that were dispersed and recovered were cached/buried? “Scatter-hoarding” is the first word in the abstract, yet there is very little information about the “scatter-hoards” that were recovered in the Results. Cache depth and microsite placement are important components of effective seed dispersal but are not touched on at all. In terms of microsite deposition, were all 211 relocated acorns found under focal tree canopies? Were any moved to the canopies of neighboring trees or into the open?

In the Discussion, lines 344-348, the authors overstated the completeness of their study in terms of seed dispersal effectiveness. The study did not examine cache microsites and seedling recruitment where seeds are deposited (qualitative component), nor did they report details like number of visits or number if seeds per visit (quantitative) (Schupp et al. 2010).

Suggestions for minor edits are highlighted and written in red text in the attached PDF.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #4: No

Reviewer #5: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-21-34252_R1_LD review.pdf

Click here for additional data file.

21 Jun 2022

ACADEMIC EDITOR:

Thank you for submitting your revised manuscript. I apologize that this recommendation has taken longer than normal. It took us a long time to get reviewers for your revised manuscript, unfortunately the two reviewers of the first version were not available. All we agree that you have made many of the changes recommended by the former reviewers and the manuscript is much improved. However, both reviewers point out important details that

should be taken into account. I direct you to the (new) reviews, where they pointed some important methodological and conceptual aspects that still remain unclear, as well as an over interpretation of some results that should be softened. Because I think there is potential in this manuscript, I am returning it to you for additional major revision. Please

note that the concerns of the reviewers do need to be fully addressed.

Dear editor, thank you for providing us the opportunity to revise this work. Concerns raised by the new reviewers have been very helpful at clarifying some aspects of our methods and results. For instance, that the main effect of ungulate presence is mediated by changes in vegetation cover rather than on competition per se. In addition, comments have been very constructive so that our discussion better reflects our results. We believe that the current version has improved the first revision. Please see specific comments below.

COMMENTS FROM REVIEWERS

Reviewer #4: The manuscript titled ‘Biological integrity of dehesa ecosystems favors acorn dispersal over predation in the mouse-oak mutualism’ explores the effects of predation risk cues and competition on seed size selection by rodents. The manuscript generally reads well but does not present clear novel idea. However, it combines few earlier ideas

into more comprehensive study which presents complex interactions in dehesa system. Authors applied simple, commonly used both field and statistical methods presented in the Manuscript. The manuscript has been noticeably improved after previous review rounds. However, I still suggest a major revision before publication and provide some comments.

We appreciate the support to our work and the constructive comments that have helped us to be more precise and clearer about our results.

Title: I do not think that a term “biological integrity” reflects what has been presented in the study. Biological integrity represents the capability of supporting and maintaining a balanced, integrated, adaptive community of organisms having a species composition, diversity, and functional organization comparable to that of the natural habitat of the

region before human alterations were imposed. I do not find such comparisons in the manuscript. If Authors still consider this term as somehow appropriate in the manuscript, please describe it more thoroughly in both introduction and discussion. If not, please change the title into more informative. Short title/Running headline:, i.e. “Predation and

competition favor dispersal in mouse-oak interaction”, fits much better.

We appreciate this comment since the previous title could lead to confusions for readers. We have changed it to “Ungulate presence and predation risks reduce acorn predation by mice in dehesas”. We have slightly changed the title because in the case of ungulates our interpretation is that their main effect is mediated by changes in the vegetation cover rather than competition per se. Also, since we have decided to focus on acorn predation in the title since it is more precise with respect to our findings (we did not quantify seed dispersal effectiveness).

Authors use rarely used terms in seed dispersal studies: ‘select’, ‘mobilize’, ‘mobilization’, ‘transportation’ etc. What is the differences between “seeds selected” vs “seeds removed” (e.g. Fig 1)? Seeds selected include both those consumed in situ and those removed while seeds removed – only those taken away from the seed stations? I highly suggest to use commonly used terms, constistently throughout the text what makes interpretation much easier (not only less confusing but also increasing the probability of finding the article in searching results): ‘handle’/’handling’, ‘remove’/’disperse’, and ‘removal’/’dispersal’, etc. Otherwise, it creates a great confusion.

To define the stages of the foraging decisions processes we followed that of Wang et al. 2013 (Oikos). That is (1) when encountering acorns which one to handle (acorn selection); (2) whether to remove the selected acorn away or not, (3) how far to carry it and (4) whether to predate it once deposited or not (L185-188). Selection is the first foraging decision made by rodents when encountering acorn. It is a term frequently used in literature and better reflects the ecological process we are evaluating (i.e. which acorn to handle among those available). In fact, it is modeled as a multinomial regression. Therefore, we have decided to continue calling the first decision as “acorn selection” but refer to “handled acorns” throughout the text (e.g. Fig 1, L186, L192). In the case of removal, we were referring to the decision of removing the acorn outside the cage or not (L184 and 201, in the former version of the manuscript). However, it may have not been clear enough. We have highlighted this throughout the text and in figure 1 (e.g. Fig. 1, L201). Regarding the use of term dispersal, we decided to refer to transportation rather than dispersal because a high proportion of seeds that are taken away are finally predated.

Line 122: It is not clear how many sites there are in total. Two with two subplots (one open and one exclosure) within each? Four separate plots?

In the study area there are two sites and each one has an exclosure and a “control” plot (with ungulates). To evaluate the effects of ungulate presence we established at each site half of focal trees inside the exclosure and half outside it. We have rewritten this section to make it clearer (L118-132).

Lines 129-130: Authors have mentioned that plant community structure is similar on both open and excluded areas. However, in Lines 315-316, they have assumed that obtained differences in rodent foraging can be caused by changes in vegetation cover and shrub layer. So, were there differences observed or not? If so, add some information about such changes, thus, I think ungulates may act functionally as vegetation changers rather than

competitors in this system.

This is indeed an important issue that was not clear enough in the previous version of the manuscript. In general, the study area shows a low tree density and shrub cover (1%). It is a savanna-like environment where understory cover is mainly composed by grasses and tussocks. Lack of browsing by ungulates inside exclosures has resulted in taller resprouts under canopies and vegetation around trees (mainly herbs and tussocks). Such changes mediate the availability of antipredatory cover for mice. Information about differences in vegetation cover outside and inside exclosures has been clarified in the material and methods section (L120-122, L127-132, FigS5_1).

In the results, there is no information of:

- rodent diversity and abundances based on live-trapping and video recordings. Authors do not present any data confirming that seeds were handled by Algerian mice only while merely suggest that “The Algerian mouse (Mus spretus) is the most abundant scatter-hoarding rodent in the area [44] (…)”. This information seems to be crucial since seed:rodent ratio strongly affects rodent decisions. However, I assume Authors did not

mark individuals (lines 169-171) so they can only provide information regarding rodent activity – while this not really reflects abundances and is a bit tricky: few bold individuals can be very active and they can provide a picture of higher abundances while shy individuals can be recorded or captured only once.

In L138-140 we specified that to select focal trees we live-trapped mice and captured Mus spretus with a high detection probability. Also, we could identify the species in the video recording of foraging events. These issues have been clarified in L254-256, 258. We agree that without marking individuals or adjusting occupancy-detection models we are actually inferring a “naïve occupancy”. Nonetheless, in our study area the probability of detection of rodents under tree canopies is high (0.88±0.03) and we live-trapped mice during two consecutive nights (L138-140). Thus, we believe that differences between “naïve” and actual occupancies should be low. We also agree that seed:rodent ratios are important modulators of mouse foraging decisions. However, in our study, main differences in such ratios in November and February are driven by acorn availability (during and outside the acorn fall peak). In fact, no acorns were found beneath tree canopies in February (L254).

In sum, some information was already available in the previous version of the manuscript but it may have not been sufficiently highlighted. We have tried to clarify these issues. In addition, given the high detection probabilities in our area and large differences in acorn availability between months, we believe assumptions made by our approach are reasonable.

- numbers/percentages of acorns occurred in each category of seed removal process (seeds removed, consumed in situ, consumed after removal, cached etc.),

We have completed this information in L255-258.

- three-way interactions, e.g. Size*Ungulates*Month. I am pretty sure that foraging activity (thus, the effect) of ungulates may depend on season. I suggest to add these data to the results. Especially that, I suppose, these may differ between two seasons.

We agree that evaluating changes in the effects of ungulates between months is interesting. Also, other three-way interactions like moonlight, risk and size. We have been suggested multiple three-way interactions in previous revisions of this manuscript. Nonetheless, we have decided to keep our analyses as simple as possible for two main reasons. First, the aim of our study is to assess the “average” effect of competition and risk on size-driven foraging decisions by mice. To account for changes on “baseline” brightness and acorn availability, we performed our experiment under new and full moon conditions as well as in February and November (L169-172). We did not intend to analyze all factors and their possible interactions but to better understand the effects of acorn size, competition and risk. In this sense, our current analyses reflect our specific aims. The second reason is related to sample sizes. Even though we tracked a large number of acorns >1000, at stages of acorn transportation and initial fates our sample size is much lower, making it very difficult to obtain reliable estimates of three-way interactions. Nonetheless, we have tried to fit the model for acorn selection and removal following your suggestions. When fitting the multinomial regression (i.e. acorn section) we had a lot of problems for model convergence in the variance of the random term. This warned us that adjusting such complex models may not be possible with information in our data and design, as suspected. Therefore, given that our current analyses better reflect our specific aims and that our data does not allow to obtain reliable estimates of three-way interactions for all foraging decisions, we have kept the analyses as it was in the former version.

Lines 153-154: I wonder whether the results would be similar if Authors treated “seed size” as a numerical instead of categorical factor. Nature usually does not use such clear categories. Additionally, a term “seed size” is not suitable here as “size” refers to other

measurements, such as length, width or volume, while Authors are focused on

seed mass [g] instead. I suggest to change this nomenclature.

As stated in the previous version, we treated size as numerical in our analyses (now L193). However, the way we explained the experimental set-up may have been confusing. To have a full range of acorn sizes that were balanced, we selected 5 small, 5 medium and 5 large acorns. This has been rewritten 150-153 to avoid any confusion. The term acorn size is frequently used in cafeteria experiments that evaluate rodent foraging decisions. We have decided to keep the same term but we have specified more clearly that we are referring to grams (L193).

Line 193: What actually does ‘seed availability’ mean in this study? It is not explained in the text. Is it about naturally occurring acorns? Experimental acorns left in the seed stations? How was this measured?

Local acorn availability was measured as total acorn mass in the cage during the event. This information was in the previous version (L192-193, now L195-196).

Lines 203-204: Were distances measured for all seeds removed from seed stations (both consumed and cached)? “Seed dispersal distances” should be used and analyzed for cached seeds only, while “seed removal distances” can be analyzed for all seeds. If all seed distances were analyzed altogether ,“seed fate” should be added as a fixed factor to

reveal differences in removal distances between cached and consumed seeds. Otherwise it is a bit confusing. Usually, consumed seeds are removed on shorter distances than cached so adding consumed seeds into analysis can alter the results.

In general, when evaluating the so-called “dispersal” kernels of acorn mobilization by rodents all removed acorns are taken into account. Even though this is not measuring dispersal it provides very important information about the ability of rodents to move throughout the landscape and about their decision of “how far to carry the seed”. Since predation rates by rodents tend to be high, there is usually not enough sample size to estimate dispersal kernels on solely cached acorns, especially in managed systems like dehesas. We have used the term acorn transportation throughout the text (instead of dispersal) because we are aware that removal and dispersal are not synonyms in the case of mouse-acorn interactions. We have rewritten this sentence to clarify it L205-207.

We agree that frequently predated seeds are dispersed shorter distances. However, including seed fate as a fixed factor in the analyses of transportation distances would mix two foraging decisions (how far to carry a seed and whether to consume it or not). This would be a little bit confusing with respect to the aims of our work, which is evaluating the effects of acorn size and environmental variables on each foraging decision (selection, removal, transportation and predation). In any case, we have performed the analyses as suggested and there was no difference in mobilization distances mediated by seed fate (mean effect -0.21 [-0.27, 0.43]). Therefore, we have decided to continue with the current analyses.

Lines 252-253: The information regarding number of seeds as well as trees used should be provided in methods.

Information about the number of acorns tracked was a result since not all focal trees had mice activity. This was not clear enough so we have rewritten this part of the text (L256-261).

Lines 253-254: What about exclusures vs. controls? Was mouse activity altered by predator scent?

We specified new and full moon conditions in both months because these were the two factors defining our cafeteria experiment set-up (L167-169). Therefore, the combination of moon and month was relevant to specify the level of activity in our focal trees, and hence, the number of acorns tracked, removed etc. Probably it was not clear enough in the previous version of the manuscript. This part has been rewritten.

We have calculated the number of trees with activity outside and inside exclosures and also in trees with and without predator scent. We provide it below but we believe it is not relevant for our work.

Table S1. Number of trees showing mouse activity (in different treatments.

Month Moon Ungulate presence Predator scent

November New (18) No (11) No (9)

Yes (7) Yes (9)

Full (26) No (13) No (11)

Yes (13) Yes (15)

February New (26) No (13) No (10)

Yes (13) Yes (16)

Full (24) No (11) No (11)

Yes (13) Yes (15)

Lines 255-256: The diagram showing the whole seed fate process is clearly

needed – this should include all numbers of seeds in all stages. Otherwise it is hard to follow throughout the results section. Please also add percentages to all numbers. What about missing seeds? Using seed-tagging method always provides some seeds that cannot be found.

Some of the information required like missing seeds was in the previous version of the manuscript. However, we did not provide enough information about the percentage and number of acorns throughout all decisions made by rodents. It has been completed as suggested in L256-261. We do not include a diagram because we would have needed to split different months, areas with and without ungulates, moon-light conditions… We believe that with the information now completed in the main text and our analyses we provide a full picture of our results.

Lines 271-272: Could removal distances vary between exclosures vs. controls (the effect of ungulates) due to differences in vegetation structure? For example, if ungulate activity leads to vegetation surrounding the focal trees (because it is normally grazed in open spaces) then removal distances will be shorter (because vegetation is closer to focal trees in open spaces/controls while more scattered in exclosures). Moreover, did Authors

check for microhabitats chosen for caching in exclosures vs. controls?

Indeed, the effects of ungulate presence may be related to change in vegetation structure rather than to competition (as discussed in the previous draft). Inside exclosures, vegetation around focal trees was higher and also oak resprouts. Accordingly, mice were able to perform an acorn selection behaviour and mobilized seeds further. We have rewritten some parts of the discussion to stress these ideas see L316-320, 322-324, 361-363.

Regarding microhabitats, we recorded the type of microhabitat where acorns were left after transportation. Most of acorns were deposited under tree canopies or resprouts (96.8 and 97.9% of them, inside and outside exclosures). This information is now provided L280-283.

Line 288: Please change ‘relaxed’ into more appropriate terminology, such as ‘not exposed to stressful factors’. I assume mice can never be truly relaxed.

Changed accordingly.

Reviewer #5: The combination of including predation risk with ungulate exclosures in a disturbed landscape makes for an interesting study with novel results. This research provides important insights regarding rodent foraging decisions in anthropogenic habitats which can inform oak woodland restoration.

We are grateful for the support to our work and constructive comments. We have tried to be clearer with respect to some terms used throughout the text. Also, we have toned down some parts of the discussion according to the concerns raised throughout the revision.

Overall:

The use of the words “mobilized” and “mobilization” instead of “dispersed” and “dispersal” (especially when “dispersal” is used in the title) is confusing. Maybe the authors don’t consider movement away from the parent plant “dispersal” unless seedling

recruitment is the final fate? Whatever the reason, there is a whole body of dispersal literature that uses the term “dispersed” or “dispersal” to describe seed movement away from the focal plant. Also, due to the definition of mobilization, the term can carry with it a

military connotation and doesn’t seem to be the best term to replace “dispersal” if that was the authors’ intent (mobilization: 1. The action of a country or its government preparing and organizing troops for active service; 2. the action of making something movable or capable of movement). Trying to make the manuscript less repetitive by using

“mobilization” or “movement” every so often is understandable, but it is overused in the manuscript and shouldn’t be used along with “distance” since “dispersal distance” is a pretty well-established term. The seed dispersal literature is a quite large body of work at this point with its own established terminology that creates continuity, so it’s confusing and unnecessary to use new terminology without any justification.

We agree that the term dispersal has been widely used for scatter-hoarding rodents. Nonetheless, from our perspective using this term is confusing because frequently rodents (and specially mice) consume seeds that have been carried away. In such cases they act as seed predators rather than dispersers. Therefore, coining the term dispersal to transported seeds may be confusing with respect to the actual role of mice. This is the main reason why we used the term “mobilization” throughout the text. We were not aware that “mobilization” could be confounded with military terms (English is not our mother tongue). Therefore, we have removed mobilization and have used “transportation” instead.

The Methods and Results are lacking some important details for a study with “dispersal” in the title. How far from the cages did the search area for dispersed acorns extend?

We searched for acorns within a 30 m-radius circle around focal trees (and cages). This represents an area where most acorns are deposited by mice (at least in the short term) and, in principle, a high proportion of Mus spretus home ranges (see comments below). Information now available in L160-162.

What was considered “predated”? fragments found or tags left behind? Since dispersed acorns were tracked “throughout the experiment,” was any secondary dispersal of cached

acorns documented?

Following Perea et al.’s (2011) experiments, we considered an acorn to be predated when it was completely consumed or partially consumed with embryo damage. Now in L164-167.

No secondary acorn dispersal event was observed in our experiments, as well as during previous work with this species in the same area (Muñoz & Bonal 2007, 2011); this behavior seems rare even for wood mice, a much more proficient acorn disperser owing to its larger size (Perea et al. 2011).

If tagged acorns were removed and never found, could they not have been dispersed outside of the search area? 21% of removed acorns were not relocated but nothing is mentioned in the Results or Discussion about their potential fate.

Unfound acorns may have been mobilized further than 30 m or predated by other acorn consumers characterized by larger home ranges than mice (e.g. ungulates). We searched for acorns within a quite large area around focal trees, where most acorns are usually deposited and compromising an area adequate for Mus spretus home ranges (e.g. Gray et al. 1998). Therefore, we believe that our findings reflect general patterns of acorn removal distances and predation rates. We prefer to not speculate about unfound acorns because their fate is very uncertain.

Gray, S. J. et al. 1998. Microhabitat and spatial dispersion of the grassland mouse (Mus spretus Lataste). – J. Zool. 246: 299–308.

How many of those that were dispersed and recovered were cached/buried? “Scatter-Hoarding” is the first word in the abstract, yet there is very little information about the

“scatter-hoards” that were recovered in the Results. Cache depth and microsite placement are important components of effective seed dispersal but are not touched on at all.

We now explicitly state that only 8 acorns (11.9% of seeds dispersed and not predated) were found buried (i.e. cached) L260.

In terms of microsite deposition, were all 211 relocated acorns found under focal tree canopies? Were any moved to the canopies of neighboring trees or into the open?

Most acorns were found beneath tree canopies or resprouts (now L280-283). Since a low proportion of acorns were found in open areas (2.37%) the actual probability of recruitment of transported seeds was very low. We have toned down some parts of the discussion accordingly. L349-353.

In the Discussion, lines 344-348, the authors overstated the completeness of their study in terms of seed dispersal effectiveness. The study did not examine cache microsites and seedling recruitment where seeds are deposited (qualitative component), nor did they report details like number of visits or number if seeds per visit (quantitative) (Schupp et al. 2010).

We agree that in the previous version the completeness of our study was overstated. In fact, low caching rates (<1% of the seeds removed) did not allow us to evaluate the effects of environmental factors on actual scatter-hoarding by mice. We are confident that the present work provides new information about how different environmental factors affect mouse foraging decisions in dehesas, and hence, potential recruitment. This is quite important since corvids are usually lacking when dehesas do not have nearby forests. Nonetheless, we agree that some parts of the discussion were too optimistic about our approach. Therefore, we have rewritten them and hopefully now they better reflect the strengths and limitations of our work. L358-382.

Suggestions for minor edits are highlighted and written in red text in the

attached PDF.

Thank you for the throughout revision of the text. It has been very helpful to correct some grammar mistakes and clarify some sentences.

L85. In the sentence: “Nonetheless, the strength and even sign of acorn size effects on mouse foraging decisions are not unequivocal, but context-dependent”. Commented: “What’s the sign of acorn size?”.

In this sentence strength and sign are referring to the effects of acorn size. They can be positive or negative (sign) and have a different absolute value (strength).

11 Jul 2022

Ungulate presence and predation risks reduce acorn predation by mice in dehesas

PONE-D-21-34252R2

Dear Dr. Díaz,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you will receive an e-mail detailing the required amendments. When these have been addressed, you will receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Pedro G. Blendinger, PhD

Academic Editor

PLOS ONE

1 Aug 2022

PONE-D-21-34252R2

Ungulate presence and predation risks reduce acorn predation by mice in dehesas

Dear Dr. Díaz:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Pedro G. Blendinger

Academic Editor

PLOS ONE