Anthropogenic habitat loss and fragmentation represent major threats to biodiversity and ecosystems (Fahrig, 2003; Maxwell et al., 2016). Tropical rainforests – the home of the majority of non-human primates – are amongst the most strongly affected ecosystems (Estrada et al., 2017). Habitat loss, i.e., deforestation, results in local, regional or even global extinction of primates and other inhabitants of tropical forests, unless a species is capable of switching to a completely different habitat. In contrast, fragmentation can have variable effects and outcomes, from negative, through neutral to even positive (Ewers and Didham, 2006; Benítez-Malvido and Arroyo-Rodríguez, 2008; Fahrig, 2017). Some primate taxa may be more resilient to habitat fragmentation than others, but this resilience may reach its limits when other disturbances (e.g., the outbreak of diseases) affect their populations (Bicca-Marques et al., 2020; Strier, 2021).

The specific outcome of habitat fragmentation depends on a number of factors including the size and shape of fragments, the distance between fragments, the structure and permeability of the landscape (matrix) surrounding fragments, and the life history and ecological traits of the primate species (Benchimol and Peres, 2014; Galán-Acedo et al., 2019). Larger fragments are likely to provide more and higher quality habitat, which can sustain larger groups or populations. However, while some studies found a relationship between fragment size and presence or abundance of a primate species, others did not (Cristóbal-Azkarate et al., 2005; Arroyo-Rodríguez and Dias, 2010; Marsh et al., 2013). The shape of fragments is likely to influence habitat quality through edge effects: the proportion of a fragment that is affected by edge effects increased with the irregularity of the shape (which results in a larger ratio between the perimeter and the area; Benítez-Malvido and Arroyo-Rodríguez, 2008; see Murcia, 1995, for edge effects in fragments in general). On the other hand, irregularly shaped fragments, particularly when large, may increase the probability that a fragment will be encountered by traveling groups or individuals (Benítez-Malvido and Arroyo-Rodríguez, 2008). The distance between fragments, and the structure and permeability of the landscape will also affect this probability, and the probability that a single group can occupy more than one fragment (Prugh et al., 2008; Pozo-Montuy et al., 2011; Benchimol and Peres, 2014). Roads are another factor that influences the permeability of the landscape; they can limit the movement of primates, and crossing roads may be lethal (Asensio et al., 2017; Hetman et al., 2019). Finally, tree density and canopy cover are obvious characteristics of fragments that directly influence tree-living organisms like primates, as they obtain most food from (plant-based diet) or within (invertebrate and vertebrate prey) trees.

The San Martín titi monkey, Plecturocebus oenanthe (Pitheciidae; Platyrrhini; Primates; previously Callicebus oenanthe; see Byrne et al., 2016 for taxonomic change) is endemic to the San Martín region in Peru (Vermeer and Shanee, 2020). It is categorized as Critically Endangered (IUCN category: CR), due to an estimated population reduction of ≥80% since around 2000 caused by habitat loss (Vermeer and Shanee, 2020). Throughout this region, agricultural activities have led to heavy deforestation and fragmentation. The species’ original geographic range covered 14,686 km², but by 2013, 55.6% of the original range was lost, and of the remaining 6,500 km² only about 1,900 km² (13% of the original range) were considered as “good” quality habitat (Shanee et al., 2013). These authors recommended studying the presence of P. oenanthe in marginal habitat, which would include highly fragmented areas. Therefore, in this study we originally aimed at examining fragment occupation by P. oenanthe and characterizing fragments in which the species is present with regard to size, tree density and other variables in comparison to fragments in which the species is not detected. Unexpectedly, titi monkeys occupied all except one of the surveyed forest fragments. Therefore, this paper quantitatively describes the characteristics of occupied fragments and discusses the implications for the conservation of San Martín titi monkeys, in order to identify the range of characteristics under which this species exists in the area, and to identify the minimum values (e.g., fragment size) and maximum values (e.g., distance to nearest fragment) that can be tolerated. As most groups were not habituated to humans, we could not collect data on group size and composition.

Quite unexpectedly, our study found P. oenanthe to be present in all but one of the surveyed fragments. Also, titi monkeys were present in 2019 in all fragments surveyed in 2015, despite further fragmentation and reduction in forest cover between 2015 to 2019 ( Table 1 ). Forest fragments varied strongly in all characteristics that we considered, except for shape. This indicates that the critically endangered P. oenanthe seems to be able to adapt to and to persist in a fragmented habitat embedded in an agricultural matrix, at least in the short term.

In our study, P. oenanthe occupied fragments as small as 0.3 ha. This is similar to dusky titi monkeys, Plecturocebus moloch, which occupy fragments as small as 0.5 ha (Michalski and Peres, 2005). Presence in such small fragments does not necessarily mean that a group occupies this fragment only. Rather, it is conceivable that titi monkeys occupy two or more neighboring fragments. However, we detected titi monkeys in the majority of cases upon the first visit to a fragment, which makes it plausible that they were indeed residing in the respective fragment only. In any case, residing in and travelling between different fragments remains an option. Such travelling would be facilitated if distances between fragments are short (Fahrig, 2017), which is the case in our study. This would allow titi monkeys to employ landscape complementation (i.e., fragments provide different resources, e.g., food in one, sleeping sites in another) and landscape supplementation (i.e., fragments provide supplementary resources, e.g., different food species) (Dunning et al., 1992).

The type and quality of vegetation around fragments will also be critical, as it determines the degree of isolation of forest fragments, matrix permeability and connectivity (Prugh et al., 2008; Pozo-Montuy et al., 2011; Benchimol and Peres, 2014). In our study areas, shrubland, young secondary forest, agroforestry coffee plantations, and banana plantations or cocoa plantations would allow for above ground or arboreal travelling, while unshaded coffee plantations, pineapple plantations and pastures could only be crossed via terrestrial travelling. So far, terrestrial foraging but not terrestrial travelling has been observed in P. oenanthe (DeLuycker, 2007). However, other Plecturocebus species regularly forage and travel on the ground, including travelling between forest fragments (Martínez and Wallace, 2011; Souza-Alves et al., 2019). Travelling between forest fragments, both above and on the ground, has been reported in different species of howler monkeys, genus Alouatta (e.g (Asensio et al., 2009; Arroyo-Rodríguez and Dias, 2010; Serio-Silva et al., 2019). On the way to surveying fragments, we observed titi monkeys feeding on bananas, mandarins and “taperiba”, Spondias mombin (Anacardiaceae) in farmland; moving through agroforestry areas, shrubland, timber plantations; and resting in a large tree with vines in an agroforestry area. Farmers reported having seen the titi monkeys feeding on farm fruit and they even left fruit out for them. They also reported titi monkeys moving through agroforestry areas and feeding on “guava”, Psidium guajava (Fabaceae), trees that provide shade for the coffee.

Apart from finding sufficient resources in fragments or complementing/supplementing resources by occupying two or more fragments, travelling between fragments is also extremely important for the social dynamics of these pair-living primates. Upon maturity, titi monkeys of both sexes usually disperse from their family group (Bossuyt, 2002; Van Belle et al., 2016). This may become increasingly more difficult if travelling between fragments is impeded, or if no unoccupied fragments are available. This could increase group size and – in the long term – change the social and mating system. Guyana sakis, Pithecia pithecia, and golden-faced sakis, Pithecia chrysocephala, usually live in small family groups with a single breeding pair (Norconk, 2007), but on an island and in an isolated forest fragment, respectively, groups contain multiple adult males and females (Setz and Gaspar, 1997; Norconk, 2006).

All species of the genus Plecturocebus, including P. oenanthe, occupy small home ranges, mostly <10 ha, and even in continuous forest, home ranges can be as small as 1.7 ha ( Supplementary Table 1 ). This may dispose members of this genus to persist in small fragments, provided these harbor sufficient resources, both food and safe places for sleeping and resting. Whether the shape of fragments is an important factor, too, must remain open. In our study, most fragments were round (SI 1.2-1.4), which minimizes the amount of edge (Benítez-Malvido and Arroyo-Rodríguez, 2008).

Tree density varied strongly between fragments. However, we must point out that the number of sampling points for the point-centered quadrat method is lower than theoretically requested (Ganzhorn et al., 2011). Nevertheless, since the potential error applies to practically all fragments, the results still suggest that P. oenanthe can persist in forest fragments with quite divergent tree densities. The same holds true for canopy height.

Distance to roads also varied strongly; some fragments were immediately next to a road while others were a several hundred meters away. Thus, nearby roads do not appear to deter titi monkeys from living in forest fragments. While roads can constrain travel between fragments and crossing roads may be lethal (Asensio et al., 2017; Hetman et al., 2019), Galán-Acedo et al. (2019) did not find any significant effect of road abundance in fragmented landscape on primate populations.

In summary, our study demonstrates that P. oenanthe can persist in a highly fragmented habitat, seemingly tolerating a broad range of fragment characteristics, giving hope for the future of this Critically Endangered species. Thus, the major threat may not consist in forest fragmentation per se, but in forest loss, as suggested by Galán-Acedo et al. (2019) for primates in general. Nevertheless, there is currently no point for a re-evaluation of this categorization. First, detailed data are needed to see whether population densities remain stable or continue to decrease. Second, data are needed on social dynamics (group size and composition). Unfortunately, we have data neither on population density nor on group size and composition, to see whether these variables have changed since 2015. Also, we do not know whether all groups or only some are reproductive, which would be crucial for the long-term persistence in fragments and for the stability of the population. Despite continued occupation of all fragments surveyed in 2015 and 2019 and occupation of all fragments added in 2019, we cannot rule out that reproduction is impaired and population size is decreasing. However, observations on the presence of infants and juveniles suggest that reproduction can take place in groups living in fragments (DeLuycker, 2007: two infants during study period in one group in a 3 ha fragment; Hodges, 2020: two groups with one infant each in fragments of 2.15 ha and 4.15 ha, respectively; Huashuayo-Llamocca, 2017: one juvenile in group living in a 4 ha fragment). Only the collection of demographic data during more detailed and long-term monitoring of fragment occupation can resolve this issue. Third, data are needed on the productivity of forest fragments, to evaluate whether they can provide enough food resources for P. oenanthe in the long run. In any case, the reduction in forest cover in the two study areas from 2015 to 2019 is an alarming signal. Prevention of further habitat loss, along with maintaining (or increasing) connectivity between fragments, remains a conservation priority. Modelling connectivity under scenarios with increasing fragmentation or scenarios with reforestation will help to guide conservation efforts (Schaffer-Smith et al., 2016), but ground-based evaluation of populations, fragment occupation and permeability of the matrix in different parts of the distributional area of P. oenanthe remain crucial.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

Ethical approval was not required for the study involving animals in accordance with the local legislation and institutional requirements because the study was conducted as part of the research and conservation activities of the Proyecto Mono Tocón.

CLRV: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Writing – original draft. ZLCC: Investigation, Writing – original draft. EWH: Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Supervision, Validation, Writing – original draft, Writing – review & editing.