Decreasing dietary diversity following habitat loss: the case of the thin-spined porcupine in the Atlantic forest


The thin-spined porcupine (Chaetomys subspinosus Olfers, 1818) is an endemic and threatened rodent from the Atlantic Forest biological hotspot. Previous studies have demonstrated it follows a strictly leaf-based diet, limited to a few tree species, although such information is derived from few individuals (n = 7) resident in small forest fragments. We aimed to evaluate whether such dietary specialization persists when animals inhabit larger forest fragments. For this, we assessed the diet composition of 19 radiotracked individuals inhabiting small (< 50 ha; n = 10) and large (> 500 ha; n = 9) forest fragments in southern Bahia state. We compared the composition and diversity of the diet in terms of tree parts and species consumed in these contrasting fragment size environments. Secondly, we aimed to evaluate the influence of leaf chemical composition on the consumption of plant species. Our findings show that the thin-spined porcupine is a strict folivore, with individuals responding to the reduction in forest size by reducing the diversity of plant species consumed, but not by feeding on new plant parts. Although the diet was richer in larger forest fragments, certain tree species were the most consumed in both fragment-size categories and the unique item consumed were leaves. Fiber influenced positively the leaf consumption of the plant species, while there was no effect of other chemical characteristics. The influence of the habitat size reduction on food diversity may be a risk factor for the species.


An animal’s diet composition is based on a complex balance between metabolic requirements, food availability, and costs of acquisition, digestion and detoxification (Ganzhorn et al. 2017; Jensen et al. 2015). Information regarding the dietary composition of a threatened species under different wild conditions is key for supporting in situ and ex situ conservation actions, and a powerful means to identify behavioral changes in the face of anthropic disturbances. In particular, the massive loss and the increasing disturbance of remnant tropical forests may force change in the dietary composition of arboreal mammals, mainly as a result of change in the availability and quality of the food items (Cristóbal-Azkarate and Arroyo-Rodríguez 2007; Ganzhorn 1995; Mekonnen et al. 2018). For example, the decay of floristic diversity (and associated changes in the ratio of shade tolerant/intolerant tree species, plant chemistry and fruit quality) due to forest loss and isolation may “force” animals to consume lower diversity or quality of species or parts of plants (Rivera and Calmé 2006). Such conditions may impose nutritional or metabolic constraints that ultimately affect behavior and persistence of some sensitive species (Cristóbal-Azkarate and Arroyo-Rodríguez 2007; Morante-Filho et al. 2018).

The thin-spined porcupine (Chaetomys subspinosus, Olfers1818) is a browsing arboreal rodent (Erethizontidae; Vilela et al. 2009), endemic to the highly fragmented Atlantic Forest (Ribeiro et al. 2009). The genus is monospecific and the species is currently recognized as vulnerable to extinction (Bonvicino et al. 2018; Catzeflis et al. 2017). It is one of the smallest arboreal folivore mammals of the world (1.25–1.75 kg). Previous studies found a diet strictly based on tree leaves and concentrated on a few plant species (85–90% of the diet comprised of four species, de Souto Lima et al. 2010; Giné et al. 2010). Pera glabrata and Tapirira guianensis were identified among the four plant species most consumed by animals in two regions, 800 km distant from each other, located in the state of Bahia (Giné et al. 2010) and Espirito Santo (de Souto Lima et al. 2010). In addition, Giné et al. (2010) reported a diet characterized by high protein content, although nutritional traits alone were unable to explain the preference among the most-consumed species. It is possible, therefore, that other features not previously addressed, such as plant secondary metabolite levels may drive food selection of this species.

However, only a few animals were included in studies focused on diet composition aspects published to date (n = 7), and all of them were monitored in small and disturbed forest fragments (0.7–18 ha; Giné et al. 2010; ~ 2.1–59 ha; de Souto Lima et al. 2010). Since other porcupine species commonly consume a broader diversity of plant parts and species (Charles-Dominique et al. 1981; de Abreu et al. 2016; Passamani 2010; Roze 19842012; Tenneson and Oring 1985) we first ask ourselves: is Chaetomys really a strict folivore and dietary specialist (i.e., with a diet concentrated on a narrow set of plant species) or is the conclusion an artifact created by restricted food availability in small and disturbed forest fragments? In addition, we asked whether nutritional or anti-nutritional characteristics (primary and secondary compounds, respectively) could help to explain the selection and the intensity of consumption of plant species.

To answer such questions we monitored a larger number of animals (n = 19) inhabiting small and large forest fragments (< 50 and > 500 ha, respectively) from southern Bahia. We compared the diet composition in terms of plant parts and species consumed in the large and small forest fragments. We evaluated the influence of plant primary and secondary compounds on the consumption of plant species. Previous studies in the region have shown that larger habitat fragments offer a higher tree diversity and higher availability of high-quality native fruits (Pessoa et al. 2017ab; Rocha-Santos et al. 2017). Therefore, we predicted that individuals using larger fragments would choose a more diversified diet, in terms of plant parts and species, than those using smaller fragments. Finally, we predicted that protein content would have a positive effect on the food selection, while fiber, dry matter, tannin and phenolics would have negative effects, as expected of an arboreal mammal (Ganzhorn et al. 2017).

Materials and methods

Study area

We conducted the study in the Atlantic Forest of the cacao growing region from southern Bahia (38°90′–39°32′W and 14°29′–15°23′S), eastern Brazil. The local annual average temperature is 24–25 °C, and the annual rainfall averages 2000 mm/year (Mori et al. 1983). The landscape is variegated, composed of a mosaic that includes mainly mature and secondary forest, as well as shaded cocoa plantations (Giné et al. 2015). Forests are preferred by Chaetomys, while the shaded cocoa plantations and open areas are avoided (Giné et al. 2015). The bulk of the remaining forest grows on poor sandy coastal soils and is classified as a Tropical Lowland Rainforest (Thomas et al. 1998).

We studied porcupines living in forest fragments that had previously been part of a common continuous forest to reduce the influence of different original floristic composition over our results. The region maintains large fragments with high forest cover in and around protected areas such as the Una Biological Reserve (REBIO-Una) and Serra do Conduru State Park (PESC), but the fragments become smaller and the landscape more deforested to the south and north of these protected areas (Fig. 1). The forest remnants located in more deforested landscapes of this region contain more early successional vegetation, characterized by tree assemblages that are less diverse, dense, tall and thick-stemmed, as well as a vegetation with lower overall basal area and higher canopy openness than those in more forested landscapes (Rocha-Santos et al. 20162017). In addition to the reduction of shade-tolerant species, the reduction of the forest cover in this region directly affected the biomass, nutritional quality (protein and lipid contents) and functional diversity of tree fruits. (Pessoa et al. 2017ab).

Fig. 1

Map of the study region in the Bahia state of Northeastern Brazil, highlighting the protected areas Serra do Conduru State Park (northern black polygon), the Una Biological Reserve (southern black polygon), the forest fragments and sites where 19 thin-spined porcupines were radio-tracked

Capture, radio-tracking and feeding observations

We captured and monitored 19 free-living thin-spined porcupines (five males and 14 females) in ten different forest fragments (Fig. 1) following the method described by Giné et al. (2010). Among these, nine animals were captured in large forest fragments (> 500 ha) within or near to protected areas (REBIO-Una and PESC) and ten animals in small forest fragments (< 50 ha) from cocoa farms, the latter located in more deforested landscapes (see sampling effort in the Table 1 of the Appendix A). After capture, we sedated the animal using an intramuscular injection of ketamine (5 mg/kg) and xylazine hydrochloride (2 mg/kg). We fitted each individual with a 40-g VHF radio-collar (Model LB81/MS6A; 3.9 × 2.0 × 1.9 cm; Telonics Inc., Mesa, Arizona), to which we previously attached reflector tape for better nocturnal detection using halogen flashlights. All individuals were released at the site of capture. The procedures were conducted under the legal approval and consent of the Brazilian Federal Authority and followed the guidelines of the American Society of Mammalogists (Sikes et al. 2011).

Table 1 Diet composition of thin-spined porcupine in Atlantic Forest from southern Bahia

Beginning 1 week after capture, we recorded the behavior of the animals by direct visualization during a total of 1247 h distributed in 110 time series of observations performed in the crepuscular-nocturnal period (17h00–6h00) between 2005 and 2013 (Table 1 in Appendix A). In every sampled time series, we recorded the individual behavior during each 10-min interval using the instantaneous focal-sampling method (Altmann et al. 1974). When the animal was feeding, we recorded the type of plant foraged (tree, shrub, liana, and epiphyte), the part of the plant consumed (leaves, fruit, flowers or other), the maturity of consumed leaves (young or undefined) and identified plant species in the UESC/Herbarium. Because leaves appear monochromatic at night, we adopted the criteria used by Giné et al. (2010), defining young leaves as those located up to the 3rd node from the outer part of the tree branch, or “undefined regarding maturity state” when they feed in other tree branch parts. Later, we classified plant species consumed into four distinct ecological groups: liana species, shade-tolerant, shade-intolerant, and exotic tree species. This classification was based on previous studies (Giné et al. 2010) and expert opinion (L. Rocha-Santos, pers. comm.).

Diet composition and main food species

We evaluated the relative importance of each plant part and plant species in the animal diet by estimating the consumption frequency (CF), which corresponds to the proportion of records that each tree parts or species was consumed in relation to total feeding records obtained. We also calculated the consumption frequency of each plant family and ecological group.

Leaf chemical content assessment

To investigate the influence of the leaf nutritional quality on the food chosen by porcupines, we sampled 200 g of young leaves from 5 to 10 adult trees (diameter at breast height > 10 cm) from 17 plant species consumed by porcupines (six most heavily consumed species and 11 species less consumed). Each sample was composed of subsamples randomly collected from different branches and heights of each tree. We preferred to collect leaves from trees consumed recently, but sampled other trees of the same species when necessary. For comparison we collected young leaves from trees of 13 plant species that were not-consumed during our sampling, which were randomly selected among those present in the animal’s home range.

Each sample of the fresh leaves was weighed and dried at 40° C in an air forced circulation stove until weight stabilization. The dry matter (DM) was calculated as the percentage of the original weight left of each sample after drying. Then, the dry material was milled to 0.25 mm and chemical analyses were performed in the Animal Nutrition Laboratory (LANA/CENA-USP) to determine the content of crude protein (CP), acid detergent fiber (ADF), total phenolics (TP) and condensed tannins (CT). To determine the CP content (g/kg dry weight), we first measured the nitrogen (N) content using the micro-Kjeldahl method (AOAC 2005), then multiplying by 6.25, assuming that the N content of the crude protein is 16% (1/0.16 = 6.25) as recommended when this value is not known for a food (FAO, 2003). We measured the ADF (acid detergent fiber) content (g/kg dry weight), which represents mainly the fraction of cellulose and lignin, using the assay of Van Soest et al. (1991). We determined the TP and CT content using Folin-Ciocalteu (according to Makkar et al. 1993) and Butanol–HCl (Porter et al. 1986) assays, respectively. The TP and CT content were estimated in tannic acid equivalents (g/kg− 1 dry weight) and leucocyanidin equivalents (g/kg− 1 dry weight), respectively. The values of CP, ADF, TP, and CT were transformed to percentages (%).

Forest fragment size, diet diversity, richness, and similarity

Based on mapping provided by Giné et al. (2015), we classified sampling sites into small (< 50 ha) and large forest fragments (> 500 ha), following the fragment size categories defined by Ribeiro et al. (2009) for the Brazilian Atlantic Forest. Then, using the data of the presence-absence of plant species in the individual diet of each nocturnal time series (our sampling unit), we performed a rarefaction curve for the diet of each fragment-size category. For this, we used the function ‘specaccum’ and method ‘rarefaction’ from the R ‘vegan’ package (Oksanen et al. 2019), generating the mean observed richness and its confidence intervals from 1000 random permutations of the data. Since the sampling effort differed among forest fragment size categories, we compared the richness (S) in each fragment size category based on the varying effort performed among the categories. In addition, we compared dietary diversity between these two fragment size categories using Shannon's Diversity Index (H'; Shannon 1948) and the method of Hutcheson (1970) based on the t-test (Magurran 2004). Finally, we assessed the similarity of the diet among the two categories by the Quantitative Sørensen Index (Magurran 2004) using the R ‘divo’ package (Sadee et al. 2019).

Chemical factors influencing the leaf consumption of plant species

We evaluated which chemical features of the leaves significantly influence the plant species consumption using two different approaches: those influencing the occurrence of consumption (binary variable) of the plant species among those available and those influencing the consumption frequency (CF) of the plant species in the diet. For both response variables, we initially performed a General Linear Model (GLM) with all the candidate explicative chemical variables (DM, ADF, CP, TP, CT). The ratios among the variables (i.e., CP:ADF, CP:TP, and CP:CT ratio) were not used due to high correlation (r > 0.65) with the single variables. The GLMs built to examine the factors influencing the occurrence of consumption of plant species were run with ‘glm’ function and ‘binomial distribution’ (link ‘logit’) of the R statistics package (R Core Team 2018). To evaluate the factors influencing the consumption frequency, we fitted ‘negative binomial generalized linear models’ using the ‘glm.nb’ function of the R ‘MASS’ package (Ripley et al. 2019) and using the total number of feeding records in each plant species as response variable instead of the relative consumption frequency. We used the ‘dredge’ function from the R ‘MuMIn’ package to test models defined by all possible variable combinations and the null models. We ranked the candidate models based on Akaike’s Information Criterion corrected for small sample size (AICc). Models were considered equally plausible to explain the observed pattern when AICc (Δi) differences were lower than two (Burnham and Anderson 2002). Then, we determined the best model choosing the most parsimonious model among those with lower AICc and corresponding higher Akaike weights (ωi—Burnham and Anderson 2002). All analyses were performed in the R environment (R Core Team 2018).


Diet composition

We obtained a total of 950 feeding records (Table 1 in Appendix A), identifying the plant part and species consumed in 92.0% (n = 874) and 90.3% (n = 858) of these records, respectively. The thin-spined porcupines fed exclusively on leaves (100%), 98.1% obtained from trees and 1.9% from lianas. At least 50.5% (n = 441) of the total feeding records included young leaves, whereas the maturation stage could not be identified for the remaining records. There was no evidence of consumption of any other item, including no evidence of drinking water, geophagy or coprophagy.

The thin-spined porcupines fed on 190 individual plants belonging to 60 species from 28 families (Table 2 in Appendix A). Despite the high number of species consumed, two species comprised ~ 50% of the thin-spined porcupines’ diet (Albizia pedicellaris, Inga thibaudiana), both of the Ingeae tribe, and seven species comprised 75%, which included three species of genus Inga (Table 1). Only 16 plant species had more than 1 individual plant used during the sampling. Most prominent of these were Inga thibaudiana, Pera glabrata, Albizia pedicellaris and Humiria balsamifera with more than ten trees consumed of each species (Table 1). Of 19 animals monitored, at least seven fed on these four tree species, confirming the high importance of these species in the diet of the thin-spined porcupine. On the other hand, 25 of the 60 plant species identified were rarely consumed (< 3 instantaneous records), with 39 plant species represented only by one tree used and 42 plant species exclusively composing the diet of one studied animal (Table 2 in Appendix A). The most consumed family was Fabaceae (58.6%), followed by Euphorbiaceae (14.1%), and Anacardiaceae (6.3%). The Fabaceae family also comprised most of the species consumed (n = 13).

Table 2 Diet composition of the thin-spined porcupine inhabiting the small and large forest fragments from Atlantic forest in southern Bahia

Forest fragment size and the diet composition

We found that the animals inhabiting small forest fragments consumed 30 plant species from 17 families, while those in large fragments consumed 34 plant species from 19 families (Table 2). Although the number of species and families identified in the diet on these conditions was close, the diet diversity was significantly lower in small (Shannon's Diversity Index; H' = 2.04) than larger fragments (H' = 2.98; t465 = 10.78; P < 0.001). The rarefaction curve plot also showed that in the small forest fragments the diet was poorer in plant species than in the larger fragments (Fig. 2). The consumption in the small forest fragments was concentrated on fewer families and plant species than those recorded in large fragments (Table 2). For example, 4 species comprised ~ 80% of the animal diet from small forest fragments while 13 species composed this amount in large forest fragments. The Fabaceae was the most consumed family in both conditions but was 3.8 times more dominant in the diet of animals located in small (71.1% of the diet) than large fragments (19.3%).

Fig. 2

Rarefaction curves of diet richness (S) of the thin-spined porcupines in relation to the number of hours (sampling efforts) that these were observed in small (light grey) and large (dark grey) forest fragments from the Atlantic Forest, Bahia, Brazil. Bar describes the 95% confidence intervals

In general, shade-intolerant tree species contributed more than shade-tolerant species to the porcupine’s diet (81.4 and 13.8%, respectively), although the number of shade-intolerant species consumed was slightly lower than shade-tolerant (22 and 25 species, respectively). The shade-intolerant species contributed more to the animal diet in the small than the large forest fragments (88.5 and 48.8%, respectively; Table 2), while the inverse occurred to shade-tolerant species (6.7 and 42.7%, respectively). The remainder consisted of lianas (1.4 and 2.1%), exotic trees (0.0 and 0.9%) and unidentified species with respect to the ecological group (7.2 and 1.7%) (Tables 1 and 2).

The similarity of the diet composition among the small and larger fragments was 23.25% according to the Quantitative Sorensen Index. Only four plant species were present in the diet of the animals living in both fragment size categories (Albizia pedicellaris, Pera glabrata, Humiria balsamifera and Inga affinis), with the two former species among the three most frequently consumed in both situations.

Chemical factors influencing the leaf consumption of plant species

The GLM analysis indicated that the model containing only the ADF variable was the better model to explain plant species consumption from those in the individual home ranges (Table 3), (Chi-square test; d.f. = 1; P = 0.032). No other chemical variable explained the occurrence of consumption or the consumption frequency of plant species by the porcupines (Table 3). The plant species consumed in small and large fragments were not significantly different in their chemical characteristics, including the levels of foliar CP (12.6 ± 4.6 and 12.4 ± 4.6%, respectively; Student's t-test; t = − 0.10053, d.f. = 19, P = 0.921), ADF (61.8 ± 8.6 and 54.8 ± 10.4%, t = − 0.80693, d.f. = 19, P = 0.4297), DM (42.8 ± 7.3 and 39.4 ± 8.7%, t = -0.96356, d.f. = 19, P = 0.3474), TP (13.3 ± 4.9 and 11.7 ± 7.6%, t = − 0.57246, d.f. = 19, P = 0.5737), and CT (10.6 ± 5.2% and 11.6 ± 9.4%, t = 0.29067, d.f. = 19, P = 0.7745). The plant species not-consumed randomly sampled in home ranges presented 11.7 ± 5.6% of CP (mean ± SD), 50.1 ± 12.2% of ADF, 40.4 ± 10.4% of DM, 14.6 ± 9.6% of TP, and 6.6 ± 7.1% of CT. Chemical values of each plant species consumed are available in Table 3 in Appendix A.

Table 3 AICc-based model selection performed to evaluate which chemical characteristics have influenced the consumption occurrence (a) and consumption frequency (b) of plant species by the radiotagged thin-spined porcupines (Chaetomys subspinosus)


Based on the largest animal sample so far assessed from this secretive species, our study unveiled key aspects of the dietary composition of animals monitored under contrasting conditions of forest size which are representative of the conservation status of the Atlantic forest realm. The thin-spined porcupines reduce their dietary diversity when facing forest loss, while maintaining a strict folivore diet regardless of fragment-size conditions. Although some shade-intolerant plant species (Albizia pedicellaris and Pera glabrata) represented the main food items in both fragment-size conditions, we observed that the porcupines are able to consume a wider range of plant species than previously known, including a higher diversity of shade-tolerant species when foraging within large forest fragments. Fabaceae plants were crucial in the diet of the porcupines in small forest fragments, but not in large forests. We reported a positive influence of fiber content (i.e., ADF) on feeding choices of plant species, which was contrary to what we predicted, and the thin-spined porcupines did not discriminate other chemical variables, such as crude protein, condensed tannins, and total phenol levels. We here discuss our results in the light of the ecology of the porcupines and other arboreal folivorous mammals.

Dietary aspects of a strict folivore

Our study reconfirmed that the thin-spined porcupine follows a strict folivore diet, as shown in the few available systematic studies of free-ranging animals (Chiarello et al. 1997; de Souto Lima et al. 2010; Giné et al. 2010) and in disagreement with anecdotal observations on this species (Kuniy 2005; Moojen 1952). The observed high degree of folivory is unlikely to be a result of environmental or food restrictions and here we offer evidence that it is an intrinsic characteristic of this porcupine species. Considering that all other New World porcupine species rely heavily on a wide diversity of plant parts in the wild, including fruits, seeds, flowers, bark and leaves as food items (Charles-Dominique et al. 1981; de Abreu et al. 2016; Passamani 2010; Roze 1984; Tenneson and Oring 1985), our results are in line with the notion that the thin-spined porcupine is the most folivorous species of the family Erethizontidae (de Souto Lima et al. 2010; Giné et al. 2010).

The dietary characteristics here described are of paramount importance when conservation actions include a range of management purposes (maintenance, rehabilitation, and reintroduction). For instance, the supply of fruit for captive animals could cause irreversible damage such as severe gastrointestinal disorders (e.g., uncontrollable diarrhea, gastritis), as reported for other folivorous mammals (Collins and Roberts 1978) and observed in previous experiences with this species (G.A.F. Giné; person. obs.). So it is important first to treat the target species as a strict folivorous arboreal mammal in future conservation management. The second concrete contribution to management is a list of the plant species consumed, with the relative importance of each plant for their diet. The species Albizia pedicellaris, Pera glabrata, Inga thibaudiana and Humiria balsamifera appear to be important components of the porcupine diet whenever present in a forest fragment. Therefore, we recommend the use of these plant species in management. Several other species also may be considered, although the species consumed more rarely may represent feeding of an exploratory nature.

The thin-spined porcupines show a high degree of specialization at the species level and not only at the individual level as previously suggested by de Souto Lima et al. (2010) and Coltrane (2012). Individuals from different sites and under different situations feed heavily on some plant species with hard physical and chemical defenses (ADF > 50%, TP > 10% and CT > 10%, Table 3 in Appendix A). For example, Pera glabrata was heavily consumed in the Espirito Santo and Bahia (de Souto Lima et al. 2010; Giné et al. 2010), as well as in both fragment-size conditions. On the other hand, Chaetomys dwelling in large forest fragments can consume a more diverse spectrum of plant species than previously known. Therefore, despite the high degree of dietary specialization, thin-spined porcupines showed some flexibility in adapting their diet when feeding in forest fragments of different size. This is a typical characteristic of a facultative specialist, which is the category ascribed by and Shipley et al. (2009) and Coltrane (2012) to animals capable of consuming a narrow set of highly defended foods when limited by their environment, but also have a broad enough fundamental niche allowing expansion of their diet under changed environmental conditions.

Dietary selectivity

Small herbivores are often more selective in their dietary requirements, as smaller body sizes tend to increase the metabolic energetic demand per mass unit while limiting an increase of the relative capacity of the digestive tract (Milton 1979; Parra 1978). Chaetomys likely maximizes its nutritional gain by selecting young leaves, as here shown and previously reported (de Souto Lima et al. 2010; Giné et al. 2010). Young leaves commonly present a higher content of protein, water, and soluble sugar, and lower levels of fiber and plant secondary metabolites than mature leaves (Kursar and Coley 2003).

On the other hand, the decision criteria for the consumption of a plant species remain incompletely understood. We found that the porcupines failed to discriminate against primary or secondary compounds, such as CP, DM, TP, and CT, except that the plant species consumed were more fibrous (ADF) than those available. Based on assumptions about constraints imposed by body size, we expected that the thin-spined porcupines to select plant species with lower fiber levels maximizing energy and nutrient intake per unit of food, while minimizing the total amount of indigestible or slowly digestible material (fiber content) that they have to process. Thus, we can think of no plausible reason for the positive selection of ADF, but some mechanisms may help to understand how this species deals with leaves of plant species with high-fiber content.

First, it is possible that the choice of young leaves minimizes the fiber content to acceptable levels, making the choice of plant species with lower fiber levels less important. Second, an exceptionally enlarged cecum of this species may enable these animals to survive regardless of the nutritional limitations imposed by high ADF. For example, it is known that the North American porcupine (Erethizon dorsatum) with its enlarged caecum is more efficient in digesting highly lignified fiber than many ruminants and large hindgut fermenters (Coltrane and Barboza 2010; Felicetti et al. 2000). Chaetomys is at least 60% lighter than Erethizon (1.6 and 4.1 kg, respectively) and, remarkably, their fermenting cecum is ~ 53% longer (41.5 cm and 27 cm, respectively), according to measurements provided in previous studies (Giné et al. 2010; Vispo and Hume 1995). Therefore, the cecum of C. subspinosus has characteristics that suggest a high digestive ability to ferment and process a highly fibrous diet, likely producing supplemental energy. Finally, the mastication of the food into fine particles and the colonic separation previously observed in Erethizon, which allow retaining mainly fine particles, solutes and microorganisms in the cecum (Felicetti et al. 2000; Vispo and Hume 1995), may maximize the exposure of fiber material to fermentation and the concentration of digestible nutrients, while minimizing the gut-residence time of large particles, including most of the indigestible fraction (Foley and Cork 1992). Such a characteristic can be shared among porcupine species.

Previous studies reported that Erethizon dorsatum can reach N balance with diets comprising 10% of crude protein (Fournier and Thomas 1997) or even lower levels (Felicetti et al. 2000). We lack such detailed information for Chaetomys, but the average concentration of crude protein on leaves available in the environment and in those plant species reported in its diet were similar and even higher that 10% (11.7 and 12.6%, respectively). This would explain the lack of a positive selection for plant species with high-protein content, as the thin-spined porcupine can meet such nutritional requirement feeding according to the average availability of crude protein in the local forests. Nevertheless, it is important to highlight that some of the most consumed plant species (e.g., genus Inga, Tapirira and Mimosa) present higher protein content (> 16%, Table 3 in Appendix A), suggesting an important role of this species in supplying proteins to animals.

We did not detect other chemical factors influencing the consumption of plant species. Although these animals apparently do not drink water (Giné et al. 2010), the water content of the leaves (inversely proportional to DM) was definitely not a factor that significantly influenced the choice of plant species to be consumed. In this case, choosing young leaves may be sufficient. Likewise, the selection of species was not influenced by the levels of TP and CT, which are linked to chemical defenses and anti-nutritional effects (Adams et al. 2009; Ganzhorn et al. 2017). According to Felicetti et al. (2000), although the mechanisms have not yet been elucidated, porcupines appear to have water reabsorption mechanisms as well as mechanisms to neutralize the effect of tannin on protein digestion. Such mechanisms need study.

Impoverished diet and conservation implications

The reduction of the dietary diversity of the thin-spined porcupines experiencing forest loss, mainly by decreasing the number of shade-tolerant trees consumed, and heavily relying on shade-intolerant species (Table 2), was similar to the pattern of floristic alteration commonly observed in this region and other parts of the Atlantic Forest. In this biome, forest loss often leads to a loss of local tree diversity, particularly shade-tolerant species (Benchimol et al. 2017; Rocha-Santos et al. 2017; Tabarelli et al. 2010). Therefore, our results support the observed change in dietary composition as a consequence of forest loss.

Very few extant mammals are strictly arboreal and folivorous (Eisenberg 1978), making the Chaetomys subspinosus even more interesting from the point of view of ecology, evolution, and conservation. Because the species is highly dependent on forest tracts, and 80.9% of such forest within its geographic distribution encompasses remnants smaller than 50 ha (Giné and Faria 2018), a large part of the current population is likely to rely on a more impoverished plant assemblages to forage. However, it is unclear whether and to what extent the observed dietary impoverishment ffects the species. On one hand, some dietary characteristics indicate that the species may be tolerant to forest loss or disturbance, as (1) they consume leaves, which tend be an abundant and dense resource even after habitat reduction, (2) they consume leaves of shade-intolerant tree species (mainly Fabaceae), which tend be abundant in secondary vegetation, and (3) they feed from many different plant species and have the ability to adapt their diet to the plant species available in the habitat. In addition, at least two observations support the notion that plant species consumed in small forest fragments are not necessarily second choice food sources: (1) some of the plant species most consumed in small fragments were also those most consumed in large fragments, (2) these food species were highly selected in small forest fragments (consumed in a much higher proportion than their local availability, Giné et al. 2010); and (3) the nutritional quality of the most consumed plant species in the small forest fragments was apparently similar to those of trees consumed in the large fragments.

In addition, it is expected that feeding on small amounts of a wide diversity of species can bring advantages due to nutritional complementarity (Westoby 1978), as well as helping to metabolize the secondary compounds by different physiological pathways without overloading the organism with a particular toxin (Freeland and Janzen 1974; Iason and Villalba 2006). The detoxification process may require greater energy loss and consequently affect the metabolism, body condition and behavior of the porcupines (Coltrane and Barboza 2010). Thus, there is an urgent need for studies analyzing how the porcupines respond physiologically and behaviorally to habitat reduction and the associated change in the diet composition to evaluate if the dietary diversity loss can be an additional risk factor to this species.


  1. Adams JM, Rehill B, Zhang Y, Gower J (2009) A test of the latitudinal defense hypothesis: herbivory, tannins and total phenolics in four North American tree species. Ecol Res 24:697–704.

    CAS  Article  Google Scholar 

  2. Altmann J, Loy J, Wagner S (1974) Observational study of behavior: sampling methods. Behaviour 49:227–266.

    CAS  Article  Google Scholar 

  3. AOAC (2005) Official methods of analysis of AOAC. AOAC International, Gaithersburg

    Google Scholar 

  4. Benchimol M, Mariano-Neto E, Faria D, Rocha-Santos L, Pessoa MS, Gomes FS, Talora DC, Cazetta E (2017) Translating plant community responses to habitat loss into conservation practices: forest cover matters. Biol Conserv 209:499–507.

    Article  Google Scholar 

  5. Bonvicino, C.R., D’Andrea, P.S., Bezerra, A.M.R., Percequillo, A., Portella, A., Christoff, A.U., Almeida, A.M., Carmignotto, A.P., Silva, C.R., Raices D.S.L., Medeiros, L., Hingst-Zaher, E., Fernandes, F.A., Ximenes, G.E., Lessa, G., Moreira, J., de Oliveira, J.A., Cherem, J., Tiepolo, L.M., Reis, M.L., Weksler, M., Alvarez, M.R., Faria, M.B., Gonçalves, P.R., Peres, P.H.A.L., Paresque, R., Vilela, R.V., de Freitas, T.O., Leite, Y.R., 2018. Chaetomys subspinosus, in: Instituto Chico Mendes de Conservação da Biodiversidade (Org.), Livro Vermelho da Fauna Brasileira Ameaçada de Extinção: Volume II. Mamíferos, ICMBio, Brasília, pp. 459–462.

  6. Burnham KP, Anderson DR (2002) Model selection and inference: a practical information-theoretic approach, 2nd edn. Springer-Verlag, New York

    Google Scholar 

  7. Catzeflis F, Patton J, Percequillo A, Bonvicino C, Weksler M, 2017 Chaetomys subspinosus. (Accessed on 20 June 2018) [WWW Document]. IUCN Red List Threat. Species.

  8. Charles-Dominique P, Atramentowicz M, Charles-Dominique M, Gérard H, Hladik A, Hladik CM, Prévost MF (1981) Les mammiferes frugivores arboricoles nocturnes d´une forest guyanaise: inter-relations plantes-animaux. Rev. d’écologie. La terre la vie Paris 35:341–435

    Google Scholar 

  9. Chiarello AG, Passamani M, Zortéa M (1997) Field observations on the thin-spined porcupine, Chaetomys subspinosus (Rodentia; Echimyidae). Mammalia 61:29–36.

    Article  Google Scholar 

  10. Collins L, Roberts M (1978) Arboreal folivores in captivity—Maintenance of a delicate minority. In: Montgomery GG (ed) The ecology of arboreal folivores. Smithsonian Institution Press, Washington, D.C., pp 5–12

    Google Scholar 

  11. Coltrane JA (2012) Redefining the north American porcupine (Erethizon dorsatum) as a facultative specialist herbivore. Northwest Nat 93:187–193.

    Article  Google Scholar 

  12. Coltrane JA, Barboza PS (2010) Winter as a nutritional bottleneck for North American porcupines (Erethizon dorsatum). J. Comp Physiol B Biochem Syst Environ Physiol 180:905–918.

    Article  Google Scholar 

  13. Cristóbal-Azkarate J, Arroyo-Rodríguez V (2007) Diet and activity pattern of howler monkeys (Alouatta palliata) in Los Tuxtlas, Mexico: effects of habitat fragmentation and implications for conservation. Am J Primatol 69:1013–1029.

    PubMed  Article  Google Scholar 

  14. de Abreu TCK, da Rosa CA, Aximoff I (2016) New record of feeding behavior by the porcupine Coendou spinosus (F. Cuvier, 1823) in high-altitude grassland of the Brazilian Atlantic Forest. Mammalia.

    Article  Google Scholar 

  15. de Souto Lima RB, Oliveira PA, Chiarello AG (2010) Diet of the thin-spined porcupine (Chaetomys subspinosus), an Atlantic forest endemic threatened with extinction in southeastern Brazil. Mamm Biol 75:538–546.

    Article  Google Scholar 

  16. Eisenberg JF (1978) The evolution of arboreal herbivores in the class Mammalia. In: Montgomery GG (ed) The ecology of arboreal folivores. Smithsonian Institution Press, Washington, D.C., pp 135–152

    Google Scholar 

  17. Felicetti LA, Shipley LA, Witmer GW, Robbins CT (2000) Digestibility, nitrogen excretion, and mean retention time by North American porcupines (Erethizon dorsatum) consuming natural forages. Physiol Biochem Zool 73:772–780.

    CAS  PubMed  Article  Google Scholar 

  18. Foley WJ, Cork SJ (1992) Use of fibrous diets by small herbivores: how far can the rules be ‘bent’? Tree 7:159–162

    CAS  PubMed  Google Scholar 

  19. Fournier F, Thomas DW (1997) Nitrogen and energy requirements of the North American porcupine (Erethizon dorsatum). Physiol Zool 70:615–620.

    CAS  PubMed  Article  Google Scholar 

  20. Freeland WJ, Janzen DH (1974) Strategies in herbivory by mammals: the role of plant secondary compounds. Am Nat 108:269–289.

    CAS  Article  Google Scholar 

  21. Ganzhorn JU (1995) Low-level forest disturbance effects on primary production, leaf chemistry, and lemur populations. Ecology 76:2084–2096.

    Article  Google Scholar 

  22. Ganzhorn JU, Arrigo- Nelson SJ, Carrai V, Chalise MK, Donati G, Droescher I, Eppley TM, Irwin MT, Koch F, Koenig A, Kowalewski MM, Mowry CB, Patel ER, Pichon C, Ralison J, Reisdorff C, Simmen B, Stalenberg E, Starrs D, Terboven J, Wright PC, Foley WJ (2017) The importance of protein in leaf selection of folivorous primates. Am J Primatol 79:1–13.

    CAS  PubMed  Article  Google Scholar 

  23. Giné GAF, Faria D (2018) Combining species distribution modeling and field surveys to reappraise the geographic distribution and conservation status of the threatened thin-spined porcupine (chaetomys subspinosus). PLoS ONE 13:1–22.

    CAS  Article  Google Scholar 

  24. Giné GAF, Duarte JMB, Faria D (2010) Feeding ecology of a selective folivore, the thin-spined porcupine (chaetomys subspinosus) in the Atlantic forest. J Mammal 91:931–941.

    Article  Google Scholar 

  25. Giné GAF, De Barros EH, Duarte JMB, Faria D (2015) Home range and multiscale habitat selection of threatened thin-spined porcupine in the Brazilian Atlantic forest. J Mammal 96:1095–1105.

    Article  Google Scholar 

  26. Hutcheson K (1970) A test for comparing diversities based on the Shannon formula. J Theor Biol 29:151–154.

    CAS  PubMed  Article  Google Scholar 

  27. Iason GR, Villalba JJ (2006) Behavioral strategies of mammal herbivores against plant secondary metabolites: the avoidance-tolerance continuum. J Chem Ecol 32:1115–1132.

    CAS  PubMed  Article  Google Scholar 

  28. Jensen LM, Wallis IR, Foley WJ (2015) The relative concentrations of nutrients and toxins dictate feeding by a vertebrate browser, the greater glider Petauroides volans. PLoS ONE 10:1–12.

    CAS  Article  Google Scholar 

  29. Kuniy AA (2005) Análise do conteúdo estomacal e intestinal de dois espécimes de ouriço-preto (Chaetomys subspinosus, olfers, 1818). Acta Biol Lepondensia 27:187–189

    Google Scholar 

  30. Kursar TA, Coley PD (2003) Convergence in defense syndromes of young leaves in tropical rainforests. Biochem Syst Ecol 31:929–949.

    CAS  Article  Google Scholar 

  31. Magurran AE (2004) Measuring biological diversity, 1st edn. Blackwell Publishing Ltd, Oxford, UK

    Google Scholar 

  32. Makkar HPS, Blümmel M, Borowy NK, Becker K (1993) Gravimetric determination of tannins and their correlations with chemical and protein precipitation methods. J Sci Food Agric 61:161–165.

    CAS  Article  Google Scholar 

  33. Mekonnen A, Fashing PJ, Bekele A, Hernandez-Aguilar RA, Rueness EK, Stenseth NC (2018) Dietary flexibility of Bale monkeys (Chlorocebus djamdjamensis) in southern Ethiopia: effects of habitat degradation and life in fragments. BMC Ecol 18:1–20.

    Article  Google Scholar 

  34. Milton K (1979) Factors influencing leaf choice by howler monkeys: a test of some hypotheses of food selection by generalist herbivores. Am Nat 114:362–378.

    CAS  Article  Google Scholar 

  35. Moojen J (1952) Os roedores do Brasil. Instituto Nacional do Livro, Rio de Janeiro, Brasil

    Google Scholar 

  36. Morante-Filho JC, Arroyo-Rodríguez V, Pessoa MS, Cazetta E, Faria D (2018) Direct and cascading effects of landscape structure on tropical forest and non-forest frugivorous birds. Ecol Appl 28:2024–2032.

    PubMed  Article  Google Scholar 

  37. Mori SA, Boom BM, de Carvalho AM, dos Santos TS (1983) Southern Bahian moist forests. Bot Rev 49:155–232.

    Article  Google Scholar 

  38. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, Mcglinn D, Minchin PR, Hara RBO, Simpson GL, Solymos P, Stevens MHH, Szoecs E (2019) vegan: community ecology package. R package version 2.5–6. Available from

  39. Parra R (1978) Comparison of foregut and hindgut fermentation in herbivores. In: Montgomery GG (ed) The ecology of arboreal folivores. Smithsonian Institution Press, Washington, D.C., pp 205–230

    Google Scholar 

  40. Passamani M (2010) Use of space and activity pattern of Sphiggurus villosus (F. Cuvier, 1823) from Brazil (Rodentia: Erethizontidae). Mamm Biol 75:455–458.

    Article  Google Scholar 

  41. Pessoa MS, Hambuckers A, Benchimol M, Rocha-Santos L, Bomfim JA, Faria D, Cazetta E (2017a) Deforestation drives functional diversity and fruit quality changes in a tropical tree assemblage. Perspect Plant Ecol Evol Syst 28:78–86.

    Article  Google Scholar 

  42. Pessoa MS, Rocha-Santos L, Talora DC, Faria D, Mariano-Neto E, Hambuckers A, Cazetta E (2017b) Fruit biomass availability along a forest cover gradient. Biotropica 49:45–55.

    Article  Google Scholar 

  43. Porter LJ, Hrstich LN, Chan BG (1986) The conversion of procyanidins and prodelphinidins to cyanidin and delphinidin. Phytochemistry 25:223–230.

    CAS  Article  Google Scholar 

  44. R Core Team, 2018. R: a language and environment for statistical computing [WWW Document]. Version 3.0.0. R Found. Stat. Comput. Vienna, Austria.

  45. Ribeiro MC, Metzger JP, Martensen AC, Ponzoni FJ, Hirota MM (2009) The Brazilian Atlantic Forest: how much is left, and how is the remaining forest distributed? Implications for conservation. Biol Conserv 142:1141–1153.

    Article  Google Scholar 

  46. Ripley, B., Venables, B., Bates, D.M., Hornik, K., Gebhardt, A., Firth, D., 2019. MASS: support functions and datasets for venables and Ripley’s MASS. R package version 7.3–51.5. Available from

  47. Rivera A, Calmé S (2006) Forest fragmentation and its effects on the feeding ecology of black howlers (Alouatta pigra) from the Calakmul area in Mexico. In: Estrada A, Garber PA, Pavelka MSM (eds) New perspectives in the study of mesoamerican primates. Springer, New York, Boston, pp 189–213

    Google Scholar 

  48. Rocha-Santos L, Pessoa MS, Cassano CR, Talora DC, Orihuela RLL, Mariano-Neto E, Morante-Filho JC, Faria D, Cazetta E (2016) The shrinkage of a forest: landscape-scale deforestation leading to overall changes in local forest structure. Biol Conserv 196:1–9.

    Article  Google Scholar 

  49. Rocha-Santos L, Benchimol M, Mayfield MM, Faria D, Pessoa MS, Talora DC, Mariano-Neto E, Cazetta E (2017) Functional decay in tree community within tropical fragmented landscapes: effects of landscape-scale forest cover. PLoS ONE 12:1–18.

    CAS  Article  Google Scholar 

  50. Roze U (1984) Winter foraging by individual porcupines. Can J Zool 62:2425–2428

    Article  Google Scholar 

  51. Roze U (2012) Porcupines: the animal answer guide. John Hopkins University Press, Baltimore, Maryland, USA

    Google Scholar 

  52. Sadee C, Pietrzak M, Seweryn M, Wang C, Rempala G (2019) divo: tools for analysis of diversity and similarity in biological systems. R package version 1.0.1. Available from

  53. Shannon CE (1948) The Bell system technical journal. Bell Syst Tech J 27:379–423.

    Article  Google Scholar 

  54. Shipley LA, Forbey JS, Moore BD (2009) Revisiting the dietary niche: when is a mammalian herbivore a specialist. Integr Comp Biol 49:274–290.

    PubMed  Article  Google Scholar 

  55. Sikes RS, Gannon WL (2011) Guidelines of the American Society of Mammalogists for the use of wild mammals in research. J Mammal 92:235–253.

    Article  Google Scholar 

  56. Tabarelli M, Aguiar AV, Girão LC, Peres CA, Lopes AV (2010) Effects of pioneer tree species hyperabundance on forest fragments in Northeastern Brazil. Conserv Biol 24:1654–1663.

    PubMed  Article  Google Scholar 

  57. Tenneson C, Oring LW (1985) Winter food preferences of porcupines. J Wildl Manage 49:28–33

    Article  Google Scholar 

  58. Thomas W, Carvalho A, Amorim A, Garrison J, Arbeláez A (1998) Plant endemism in two forests in southern Bahia. Brazil Biodivers Conserv 7:311–322.

    Article  Google Scholar 

  59. Van Soest PJ, Robertson JB, Lewis BA (1991) Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J Dairy Sci 74:3583–3597.

    PubMed  Article  Google Scholar 

  60. Vilela RV, Machado T, Ventura K, Fagundes V, De J Silva MJ, Yonenaga-Yassuda Y (2009) The taxonomic status of the endangered thin-spined porcupine, Chaetomys subspinosus (Olfers, 1818), based on molecular and karyologic data. BMC Evol Biol 9:1–17.

    Article  Google Scholar 

  61. Vispo C, Hume ID (1995) The digestive tract and digestive function in the North American porcupine and beaver. Can J Zool 73:967–974.

    Article  Google Scholar 

  62. Westoby M (1978) What are the biological bases of varied diets? Am Nat 112:627–631

    Article  Google Scholar 

Download references


We thank the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) [Project Rede SISBIOTA 563216/2010-7; Project INCT, IN-Tree 465767/2014-1] and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) [Project CASADINHO/PROCAD 552198/2011-0] for financial support and fellowships for K. F. M. da Silva. We are grateful to Rubens Vieira Lopes and Marcos Alves for field assistance as well as Universidade Estadual of Santa Cruz for logistical support. We are grateful to José Lima da Paixão and Larissa Rocha Santos for identifying and classifying plant species. This research received the legal permit (license number 27021-1) from the Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio).

Author information



Corresponding author

Correspondence to Gastón Andrés Fernandez Giné.

Ethics declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Handling editor: Sabine Begall.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary file1 (PDF 71 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Giné, G.A.F., da Silva, K.F.M. & Faria, D. Decreasing dietary diversity following habitat loss: the case of the thin-spined porcupine in the Atlantic forest. Mamm Biol 100, 473–484 (2020).

Download citation


  • Threatened species
  • Arboreal folivore mammal
  • Diet composition
  • Feeding ecology
  • Food selection
  • Plant secondary metabolites
  • Forest fragments