Keywords

1 Introduction

Tropical forests hold the highest diversity of animals on Earth, but human-driven disturbance has led many populations to decline and species at imminent risk of extinction (Dirzo et al. 2014; Ceballos et al. 2015). Hunting for subsistence has been considered a major driver for the extinction of vertebrates since the Pleistocene (Ripple et al. 2014; Young et al. 2016) and even today it is a major cause of decline in the populations of large-bodied vertebrates (Ripple et al. 2014). Humans have now affected most of the Earth’s land area and just 15% of the natural areas are protected for nature (Sloan et al. 2014). Today we have lost more than a third of all the forest world cover (Hansen et al. 2013). In addition to hunting and deforestation, other human-driven disturbances have been incorporated to threaten the persistence of biodiversity in the rainforests: new emerging diseases, the widespread of invasive species, expansion of roadway network, and traffic that increase wildlife-vehicle collisions and the alterations in climatic condition.

The Atlantic Forest of South America is vanishing rapidly. It used to stretch from latitudes 3° to 30° S and to occupy an area of approximately 150 million ha (Ribeiro et al. 2009). It is estimated that the Atlantic Forest holds 861 species of birds (213 endemic) (Hasui et al. 2018; Rodrigues et al. 2019), more than 300 reptiles (95 endemic), 512 species of amphibians (260 endemic) (Vancine et al. 2018), and 321 species of mammals (89 endemic) (Bovendorp et al. 2017; Lima et al. 2017; Muylaert et al. 2017; Gonçalves et al. 2018a, b; Culot et al. 2019; Souza et al. 2019, see also Figueiredo et al. 2021 Chap. 9). This biome has suffered intense deforestation, fragmentation, and defaunation in the twentieth century (Dean 1997) and currently it is restricted to less than 16 million ha, or 12% of its original distribution (Ribeiro et al. 2009). As a result, the Atlantic Forest has the highest number of endangered species compared to other Brazilian biomes, having 593 vertebrate species threatened by extinction, and six already extinct in the wild (www.iucnredlist.org; ICMBio 2018).

We defined defaunation as the global, local, or functional extinction of animal population or species (Dirzo et al. 2014). For instance, one endemic amphibian (Phrynomedusa fimbriata), four endemic birds (Cichlocolaptes mazarbarnetti, Philydor novaesi, Glaucidium mooreorum, and Mitu mitu) and one endemic mammal from Fernando de Noronha Island (Noronhomys vespucci) are now globally extinct. The Giant Otter (Pteronura brasiliensis) is locally extinct in the Atlantic Forest, but still occur in the Amazon and Pantanal, and the jaguar (Panthera onca) is functionally extinct in the Atlantic Forest, where the species is so rare that the top predator function is lost from most of the biome (Galetti et al. 2013a).

In this chapter, we identify and quantify the magnitude of the main anthropogenic drivers of defaunation in the Atlantic Forest of South America (Fig. 14.1). We discuss the ecological consequences of defaunation in the Atlantic Forest and stimulate a debate on how to revert the loss of fauna to restore biodiversity and ecosystem functions and services. How many individuals are lost per year due to human-driven causes? And what are the consequences of the defaunation to human-well-being and ecosystem processes? Here, we attempt to answer these important questions according to each of the human-driven disturbances identified as follows.

Fig. 14.1
figure 1

Conceptual model on the magnitude of the effects of defaunation in both richness and abundance of tetrapods in the Atlantic Forest. The size of the circles indicates the number of species and individuals affected by direct (red colors) or indirect (black colors) drivers. A direct driver influences ecosystem processes and can therefore be identified and measured to differing degrees of accuracy. An indirect driver operates more diffusely, often by altering one or more direct drivers, and its influence is established by understanding its effect on a direct driver. Both indirect and direct drivers often operate synergistically. The magnitude of the impact of defaunation in the Atlantic Forest is hard to estimate, but we can predict that large-scale effects will affect more species and individuals depending on the driver

Full size image

2 Infectious Diseases

Wildlife can be negatively impacted due to diseases (Doherty et al. 2017) which are commonly a silent threat for wild animals (Hudson 2002). Emerging diseases may become one of the major defaunation drivers in the Anthropocene. Patterns of epizootia are characterized as a disease that is only occasionally found in a population, but which can spread very rapidly. Emerging infectious diseases (EIDs) of free-living wild animals can be classified into three major groups on the basis of key epizootiological criteria: (1) associated with spill-over from domestic to wildlife populations; (2) related directly to human intervention via host or parasite translocations; and (3) with no overt human or domestic animal involvement (Daszak et al. 2000). These phenomena have biological implications as many wildlife species are reservoirs of pathogens that threaten domestic animal and human health, and wildlife EIDs pose a substantial threat to the conservation of global biodiversity (Daszak et al. 2000).

Infections not necessarily cause death but also symptoms that lead to the decline of the population, directly or indirectly due to reducing competition capacity for hunt or escape from the predator, infertility, altering secondary sex ratios and movement patterns, morbidity, and increasing susceptibility to other infectious diseases (Preece et al. 2017). This scenario can be reinforced by the stress due to the impact caused by environmental changes, which consequently reduce the immunity capacity contributing to new infections (Fig. 14.2).

Fig. 14.2
figure 2

Two-way flow of diseases between domestic dogs and wild carnivores, and dynamic of yellow fever disease in monkeys, associated with an impact due to the human concept regarding the risk to become infected. Some alterations in wild animals can be caused by these infectious diseases, which result in their population decline and even local extinction

Full size image

Yellow fever virus (YFV) causes a devastating disease that is resulting in widespread defaunation in the Atlantic Forest; it is a highly pathogenic virus for New World primates (Monath and Vasconcelos 2015). This disease is currently considered an EID, once a gradual expansion of that has been observed toward the southeast of Brazil, an area that was considered YF-free for almost 80 years without vaccine recommendations (Romano et al. 2014). In later 2016, the virus re-emerged in Brazil, initiating the largest epidemic of sylvatic YFV ever recorded in the Country (Cardoso et al. 2019). It was recorded more than 5500 deaths of monkeys during this outbreak (Brasil et al. 2019) (Fig. 14.2), affecting mostly howler monkeys (Alouatta spp.), marmosets (Callithrix spp.), titi monkey (Callicebus spp.), and the critically threatened muriqui (Brachyteles spp.) (Abreu et al. 2019).

Domestic dogs are the most important reservoirs and maintainers of virulent pathogens that affect wild animals (Doherty et al. 2017), specially carnivore species (Courtenay et al. 2001), favoring pathogens transmission across wild, domestic animals and humans (Whiteman et al. 2007). Important diseases transmitted from dogs to wild carnivores are named as “The Big Three” (rabies, distemper, and parvovirosis) due to the strongly negative impact over their populations (Gompper 2014). In Brazil, the risk of transmission of rabies from domestic dogs to wild animals is reduced once they are frequently immunized by Brazilian public policy programs, but the disease is still common in wild animals of Atlantic Forest, such as bats, foxes, and marmosets (Megid et al. 2015; Gonçalves et al. 2020). In case of parvovirosis and distemper virus disease, they are frequent in stray and semi-domiciled domestic dogs (Curi et al. 2016), which are normally not immunized, so their contact with the wild environment can be a threat for wild animals (Martinez et al. 2013). Parvovirosis is a disease with high lethality for very young puppies, and as an aggravating factor the virus can remain for a long time in favorable environments (Megid et al. 2015).

Canine distemper virus (CDV) is considered one of the most important diseases for domestic dogs, with high lethality (Negrão et al. 2006), and is endemic in Brazil (Megid et al. 2015). Dogs are the most important reservoirs; however, some species are susceptible to the virus, such as Canidae (foxes), Mustelidae (otter, ferret, and badger), Procyonidae (coati), Myrmecophagidae (anteaters), and Felidae (wild cats) (Megid et al. 2015). Although outbreaks of CDV disease in wild animals in the Atlantic Forest have not yet occurred, it has been classified as a threat to biodiversity with elements associated with spill-over by domestic dogs (Whiteman et al. 2007). In a large fragment of Atlantic Forest, CDV in wild felids seems to be related with home range use and the close association with domestic dogs living in nearby areas, as 31.5%, 11.3% and 34.6% of the sampled jaguars, pumas, and dogs were seropositive for the disease, respectively (Nava et al. 2008).

3 Hunting

Hunting for food or retaliation against predators of domestic species continues to be widespread in the Atlantic Forest even within protected areas (Cullen Jr. et al. 2001; Galetti et al. 2017; Sousa and Srbek-Araujo 2017), causing local extinctions or decline of many mammal populations (Canale et al. 2012; Galetti et al. 2009). In the Atlantic Forest, the abundance of medium and large vertebrates is on average 37% to 90% lower in intensely hunted areas compared to low hunted areas (Galetti et al. 2017). Through cascading effects, the abrupt decline or even extinction of populations probably have had pervasive impacts on ecosystems structure and dynamics (Dirzo et al. 2014).

At least one bird (Alagoas curassow, Pauxi mitu) and one mammal species (Giant otter, Pteronura brasiliensis) had gone extinct in the wild due primarily to hunting, although both species were threatened by habitat loss and fragmentation of the Atlantic Forest. Other game species (e.g., Aburria jacutinga and Crax blumenbachii) had their distribution reduced in more than 95% (BirdLife International 2019) and species such as the lowland tapir (Tapirus terrestris) and the white-lipped peccary (Tayassu pecari) were locally extinct in most of the Atlantic Forest (Jorge et al. 2013; Ferreguetti et al. 2018).

Hunting is ubiquitous in virtually all fragments in the Atlantic Forest and although it is considered an unregulated activity since 1967, hunting for sport or subsistence never stopped. The recent invasion of wild boar, including feralized domestic pig, and consequently legalization of pig hunting may increase wildlife poaching, due to weak law enforcement. On the other hand, wild pigs may function as a shield to wildlife since it is a preferable target among hunters (Desbiez et al. 2011).

At least 43 bird, 40 mammal, and 14 reptile species are known to be consumed as food in the Atlantic Forest. The impact of hunting on the vertebrate community in the Atlantic Forest differs from other Neotropical sites because while hunting in the Amazon is more evenly distributed among primates, large rodents, and ungulates (Bodmer et al. 1997; Alvard 1998; Peres 2000), in the Atlantic Forest, hunting concentrates heavily on large birds (guans, tinamous), ungulates, large caviomorph rodents, and armadillos (Cullen Jr et al. 2001; Galetti et al. 2017).

The lack of quantitative information on the magnitude of hunting is partly explained by the fact that hunting activities are illegal in Brazil. Therefore, most of the studies on the effects of hunting are based on comparative densities of game species (Canale et al. 2012; Galetti et al. 2017), but not effective measures of harvest from local populations. So far, one study estimated that each hunter harvest 100 kg/year of game meat in a study area in Southeastern Brazil (Nobre 2007).

4 Animal Trade

The illegal capture and trade of wild animals are certainly main defaunation drivers in the Atlantic Forest since the European invasion to South America. The highly priced red-and-green macaw (Ara chloropterus) was known to occur from the Northeastern Brazilian coast to Rio de Janeiro state and today is extirpated in most of the Atlantic Forest, with a small relict population in Morro do Diabo State Park in São Paulo state and a reintroduced population toward the north in Espírito Santo State.

The illegal capture and trade of wild animals are widespread throughout the Atlantic Forest (Alves et al. 2013) where a total of 31 species are known to be illegally traded. Birds represent the most traded group among all animals in Brazil, and the same pattern is observed in the Atlantic Forest (Renctas 2007). Species belonging to the families Emberizidae (finches) and Psittacidae (parrots and parakeets) stand out among the taxa of wild birds most sold as pets in Brazil, corroborating a trend reported in previous studies (e.g., Gastañaga et al. 2011). There are many reasons for the observed preference for species in these families. The Emberizidae are traded as cage birds because of the popular appeal of their colorful plumage. Moreover, they are extremely resistant, and their small size allows that large numbers of them being kept together in small cages, which facilitates smuggling (Frisch and Frisch 1995; Sick 1997).

The animal trade has recently acquired a new ally – the internet. A survey carried out by Renctas (2007) found 4892 advertisements in Brazilian and international websites promoting the illegal sale and exchange of wild animals from the Brazilian fauna. Most were advertisements for birds and reptiles, but mammals, amphibians, and ornamental fish were also offered (Renctas 2007). The internet illegal market must be viewed with considerable alarm because of the efficiency with which virtual markets allow buyers and sellers to connect with ease and speed that was never possible. These vast potential markets pose new challenges to legislators and enforcement agencies.

5 Invasive Species

Invasive species can impair native species through competition for space or resources, through direct predation and through spreading diseases that were not previously present in the environment. Humans also transport new diseases from one area of the globe to another. In the Atlantic Forest, three invasive species are known to have widespread distribution and impact on native wildlife: domestic or feral cats, dogs, and the wild pigs Sus scrofa (Fig. 14.3, see also Vitule et al. 2021 Chap. 13).

Fig. 14.3
figure 3

Potential distribution for the domestic dog, cat, and the wild pig, estimated as a weighted (accuracy-based) ensemble of projections from three modeling methods (generalized linear models – GLM, random forest – RF, bioclimate envelope – BioClim). The domestic cat and dog are spread out most of the Atlantic Forest, while the wild pig is concentrated in the southernmost regions (from L. Sales, unpublished)

Full size image

Although the occurrence of domestic cats and dogs is usually associated with human habitations (Srbek-Araujo and Chiarello 2008; Paschoal et al. 2018), they can reach high densities in forest remnants (Paschoal et al. 2018). This happens mostly because domestic cats and dogs are allowed by their owners to forage freely (Torres and Prado 2010). Forest fragmentation also contributes much to this scenario (Cullen Jr. et al. 2001; Manor and Saltz 2004, see also Vitule et al. 2021 Chap. 13). As a consequence of forest fragmentation , domestic animals have more access to wild areas. Cats have already been pointed out by different studies as the major predator of many birds and mammalian species, causing a decline in many of their populations around the world (Bonnaud et al. 2010). Cats can kill 700 reptiles, 150 birds, and 50 native mammals per square kilometer per year in the USA (Read and Bowen 2001). Dogs are also strong predators, and mammals tend to be their most common victims (Galetti and Sazima 2006).

The third most destructive species for wildlife along with dogs and cats is probably the wild boar, which threatens native fauna by competing for resources, destroying micro-habitats, or depleting prey-base (Ilse and Hellgren 1995; Lowe et al. 2000, see also Vitule et al. 2021 Chap. 13). They also impact natural environments by shifting water quality of streams and accelerating invasion of exotic plants over natural or disturbed areas when dispersing their seeds (Lynes and Campbell 2000; Dovrat et al. 2012; Rosa et al. 2019, see also Vitule et al. 2021 Chap. 13). Wild pigs are hosts of several zoonoses and inflict losses for agriculture by directly destroying crops or transmitting diseases to livestock (Ruiz-Fons 2017; Maciel et al. 2018). In the Atlantic Forest, wild pigs are widespread, likely representing the largest biomass of terrestrial wildlife in many Atlantic Forest fragments (Pedrosa et al. 2015; Rosa et al. 2017; Beca et al. 2017; Rosa et al. 2020, see also Vitule et al. 2021 Chap. 13).

6 Wildlife Collision (Infrastructure)

Brazil has the fourth biggest road network in the world (DNIT 2019) and the road density is higher in the Southeast region. The effects of roads and traffic on wildlife vary and range from habitat loss (Forman et al. 2003), a reduction in habitat quality in a zone adjacent to the road (e.g., noise, lights, pollution, visual disturbance) (Parris et al. 2009), barrier effect, including interruption of migration and dispersion (Lesbarrères and Fahrig 2012) and direct mortality through collisions with vehicles (Forman and Alexander 1998; Fahrig and Rytwinski 2009). Direct road mortality has the potential to alter the demographic structure of wildlife populations (Steen and Gibbs 2004) and create local population sinks (Nielsen et al. 2006). Such changes may alter the structure and function of communities and ecosystems adjacent to the road (Trombulak and Frissell 2000).

At least 179 species of birds, 92 species of mammals, 71 species of reptiles, and 25 species of amphibians have been recorded roadkilled in different projects on the roads that cross the Atlantic Forest (Grilo et al. 2018). Fifteen species are listed as vulnerable, including the jaguar (Panthera onca), puma (Puma concolor), maned wolf (Chrysocyon brachyurus), hoary fox (Lycalopex vetulus), giant anteater (Myrmecophaga tridactyla), pygmy brocket deer (Mazama nana), maned sloth (Bradypus torquatus), brown howler monkey (Alouatta guariba), and small wild cats (Herpailurus yagouaroundi, Leopardus colocolo, Leopardus geoffroyi, Leopardus guttulus, and Leopardus wiedii). The northern tiger cat (Leopardus tigrinus), listed as Endangered in Brazil, is another victim of roadkills, as well as other four near-threatened species, the black howler monkey (Alouatta caraya), greater guinea pig (Cavia magna), black-bellied slider (Trachemys dorbigni), and shrike-like cotinga (Laniisoma elegans).

In São Paulo State alone, a survey in 6.500 km of paved roads (18% of total length of paved roads in the State) recorded 37,744 roadkilled individuals, from 32 medium to large-sized mammals (0.6 animals roadkilled/km/year) in 10 years (Abra et al. unpublished). The most roadkilled species are common and generalist mammals such as Capybara (Hydrochoerus hydrochaeris, n = 12,614; 33.42%), European hare (Lepus europaeus, n = 5406; 14.32%), crab-eating fox (Cerdocyon thous, n = 4957; 13.13%), nine-banded armadillo (Dasypus novemcinctus, n = 2375; 6.29%), porcupine (Coendou sp., n = 2299; 6.09%), six-banded armadillo (Euphractus sexcinctus, n = 1537; 4.07%), southern tamandua (Tamandua tetradactyla, n = 1193; 3.16%), and raccoon (Procyon cancrivorus, n = 906; 2.40%). These species together account for more than 80% of all roadkills in the state of São Paulo.

The frequencies of roadkills vary both temporally and spatially and the extension of these impacts depends on the characteristics of the roads, such as road density, traffic volume, landscape features, proximity with protected areas, the wildlife population and their natural history (Fahrig et al. 1995; Frair et al. 2008; Freitas et al. 2015; Rytwinski and Fahrig 2013). Beyond vehicles, animals also collide with electric power lines and wind turbines, although these collisions are restricted to birds and bats. The information about animal collisions with transmission lines is scarce for the Atlantic Forest, but the available reports show that migratory birds are the main victims of this type of collisions (Marques et al. 2020).

7 Habitat Loss, Fragmentation, and Degradation

The conversion of natural landscapes to agriculture and cattle fields represents 80% of habitat loss and fragmentation worldwide being the main cause of global biodiversity loss (Laurance 2007). Changes in the landscape matrix can affect the persistence of several species, such as birds, mammals, and others in the forest fragments (Mazerolle and Villard 1999). Population sizes of large and specialist species, for example, tend to decline, recruitment rates and genetic diversity decrease, reduced due to higher inbreeding and extinction rates (see Lira et al. 2021). Meanwhile, generalist and small species tend to be benefited by the habitat loss due to the reduction and/or extinction of predators or competitors, or by the exploitation of new resources from the matrix around (Beca et al. 2017). There is evidence that habitat fragmentation severely affects the composition of local communities in the Atlantic Forest (Beca et al. 2017, see also Lira et al. 2021), creating landscapes with impoverished communities and simplified network interactions (Fahrig 2003).

The Brazilian territory has 1.2 million km2 of pastures and one of the largest cattle herds with about 218 million heads, being considered the largest beef exporter in the world (Vale et al. 2019). Although human-made pasture areas could be considered heterogeneous, associated with deforestation age and type/intensity of management (Dias-Filho and Ferreira 2013), the replacement of rainforest by open farmlands has potentially severe consequences for animal biodiversity and forestry. Only a few forest species can maintain viable populations in tropical livestock systems (Esquivel et al. 2008). Moreover, livestock impacts on wildlife can be direct, through interference competition, or indirect with changes in vegetation structure that influence the availability of natural resources and nesting sites (Gonçalves et al. 2017).

The relationship between habitat and animal loss is complex to estimate, particularly because we do not have enough information on how many animals there were in the area before deforestation. However, based on a few sites where wildlife densities have been estimated, we can calculate that for every km2 of forest destroyed thousands of vertebrates (particularly mammals and birds) and millions of invertebrates are extinct locally if they cannot move to adjacent forests.

For instance, at Cocha Cashu Biological Station, Peru, scientists have estimated a density of 263 understory birds, 288 primates, 127 marsupials, and 33 carnivores in 100 ha (1 km2) lowland Amazon forest (Terborgh et al. 1984). In the Brazilian Atlantic Forest, the density of mammals varied with altitude, rainfall, forest productivity, but hunting pressure and forest size explained most of its diversity and abundance (Galetti et al. 2009). The mean density of terrestrial medium and large-bodied mammals is about 28 individuals per km2 and the mean density of primates is about 90 individuals per km2 (Galetti et al. 2017; Chiarello 2000) where highland forests have lower densities. For small mammals (rodent and marsupials), forest landscapes with a high proportion of forest cover (80–100%) are expected to hold about 1123 individuals per km2 of forest specialist species (76%) that are unable to persist in smaller fragments (Bovendorp et al. 2018). A decline in forest cover favored rodent abundance of a few disturbance-adapted species (Bovendorp et al. 2018) represented by 73% of individuals in landscapes with ≤10% forest cover. Birds and mammals only represent a small fraction of the number of individuals and species that 1 km2 of rainforest can sustain. For example, in the Brazilian Amazon, each hectare can contain one billion of invertebrate individuals (Wilson 1987).

8 Climate Change

Climate change will redistribute the biodiversity as we know it, with negative effects to ecosystem services and human well-being (Pecl et al. 2017, see also Vale et al. 2021). In the Atlantic Forest, where most forest remnants are small (<50 ha) and embedded within human-dominated matrices (Ribeiro et al. 2009), climate-driven migrations will be hampered by the lack of habitat for most terrestrial species, especially those that are canopy-dependent (Gouveia et al. 2016, see also Vale et al. 2021). Projections of climate change effects on the biodiversity of the Atlantic Forest usually indicate range shrinks, upslope and poleward movements (Ferro et al. 2014, see also Vale et al. 2021). Climate-driven faunal movements may force native species outside protected areas (Ferro et al. 2014) and/or drive invasive species inside protected areas (Loyola et al. 2014, see also Vale et al. 2021).

Amphibians are probably the taxa most readily threatened by climate change (Lemes et al. 2014; Loyola et al. 2014). Ecological characteristics such as limited dispersal abilities, water dependency on at least one life stage, and special physiological requirements make amphibians particularly sensitive to future climatic changes (Lawler et al. 2010). At the Atlantic Forest, up to 10% of amphibian species will lose all climatically suitable area by the year 2070. If species are not able to adapt to changing conditions, or migrate to newly suitable habitat, this will translate into physiological stress, with effects on fitness, reproduction, and survival, ultimately leading to species extinctions (Lawler et al. 2010). Unexplained population declines in pristine regions have already been reported on more than 29 Brazilian amphibian species, including the families Leptodactylidae (19), Hylidae (7), Centrolenidae (2), Dendrobatidae (2), and Bufonidae (Eterovick et al. 2005). Species inhabiting higher elevations seem to be at most risk and the frog species Colostethus carioca, C. olfersioides, Crossodactylus dispar, C. gaudichaudii, Cycloramphus boraceiensis, C. duseni, C eleutherodactylus, C. granulosus, C. semipalmatus, Hylodes babax, Paratelmatobius lutzii, Thoropa lutzi, and T. petropolitana seem to have undergone population declines in pristine areas due to climate change (Eterovick et al. 2005, see also Vale et al. 2021).

As species relocate in response to climate change, colonization by invasive species may also be enhanced. For example, alterations in climatic patterns may drive the invasive American bullfrog Lithobates catesbeianus into reserves currently established in the Atlantic Forest (Loyola et al. 2014, see also Vale et al. 2021), posing another threat to this already imperiled fauna. Climate change may accelerate disease spread among amphibians in the Atlantic Forest. The fungus Batrachochytrium dendrobatidis has been associated with population declines and local extinctions of several amphibian species and is now broadly distributed in the Atlantic Forest (Carnaval et al. 2006; Carvalho et al. 2017). The interaction between climate change per se and the spread of such lethal pathogens may be the cause of widespread amphibian extinctions in South American highland forests.

The spread of vector-borne diseases will likely follow the climate-driven redistribution of parasites. In addition to range expansion, the higher temperatures expected for the twenty-first century are likely to enhance transmission rates of emerging diseases to the wildlife. Climate change may, therefore, interact with other defaunation agents to reduce vertebrate biomass in the Atlantic Forests. Yellow fever has recently decimated hundreds of howler monkeys and was concomitant with extraordinarily high temperatures in South America in the last decade. However, some studies suggest that climate is rarely the main driver of epidemic bursts (Reiter 2001). Human activities and impacts on local habitats may interact with higher temperatures to affect pathogen cycles in complex ways. Habitat degradation, for example, may enhance the vulnerability of howler monkeys to yellow fever (Chapman et al. 2005).

Most projections of climate change effects on the Atlantic Forest biodiversity, however, do not account for landscape permeability across migratory routes. If landscape mosaics are impermeable to the movement of habitat-specialist species, populations will be confined to forest remnants with unsuitable climate conditions (Gouveia et al. 2016). Within such patches, novel climates are likely to exceed the amplitude, extremes, and seasonality characteristics to which such species are adapted (Ribeiro et al. 2016, see also Vale et al. 2021). Primates and other canopy-dependent groups, in addition to other forest-specialist species, may not be able to move as the climate changes (Pecl et al. 2017). The interaction between climate and land use changes is likely to prevent climate-driven migrations, by degrading the landscape configuration for Atlantic Forest vertebrates, such as titi monkeys (Gouveia et al. 2016), lion tamarins (Meyer et al. 2014), and mountain birds. In addition, climate change will also redistribute key food resources (Raghunathan et al. 2015), thus potentially affecting the fauna dependent on them.

Models based on future expectations of greenhouse gas emissions predict large-scale changes, high rates of loss of climatically suitable areas, and reorganized communities of small mammals (rodents and marsupials) in the Atlantic Forest (Bovendorp et al. unpublished).

9 The Spatial Distribution of Defaunation in the Atlantic Forest

Defaunation does not affect all the groups of animals evenly. Thirty-four percent of the amphibians, 26% of the mammals, and 16% of the birds reported to the Atlantic Forest are under some category of threat (CR, EN, NT, VU, DD; www.iucnredlist.org). Seven mammal, 10 bird, and one frog species are at critical risk of extinction in the Atlantic Forest (Bello et al. 2017; Bovendorp et al. 2017; Culot et al. 2019; Gonçalves et al. 2018a; Hasui et al. 2018; IUCN 2019; Lima et al. 2017; Muylaert et al. 2017; Vancine et al. 2018). In general, apex predators, other carnivores, large-bodied mammals, and large herbivores were among the most defaunated functional groups (Bogoni et al. 2018; Nagy-Reis et al. 2020). Similarly, the defaunation process is not evenly distributed along the Atlantic Forest. According to Bogoni et al. (2018), most Atlantic Forest remnants are classified as medium to high levels of defaunation of mammals (i.e., half of the species were lost). Regions dominated by forest converted into cropland and cattle pastures, timber extraction, and forest edges comprise the most defaunated areas in the Northeast and Southwest parts of the Atlantic Forest (ombrophilous mixed forests and semi-deciduous forests; Fig. 14.4). Historically, these areas have presented the highest human population expansion since the seventeenth century with the highest concentration of sugarcane, coffee, and cacao plantations of colonial and modern Brazil (Ribeiro et al. 2009). In addition to bioclimatic processes related to forest fragmentation, hunting, and species invasion play important roles in the maintenance of these forest fragments. The Northeast and Southwestern Atlantic Forest concentrate the remaining indigenous lands of the Atlantic Forest, where hunting is a common practice (FUNAI 2019). Besides, the inner part of the Atlantic forest has suffered the strongest invasion of feral pigs (Pedrosa et al. 2015). On the other hand, areas that are not under high defaunation process also face threats. Particularly, the concentration of roads and the recent blooms of yellow fever can decimate the remaining animal populations of the center of the Atlantic Forest (Fig. 14.4).

Fig. 14.4
figure 4

Defaunation of mammals in the Atlantic Forest and its drivers (according to Bogoni et al. 2018). The index range between 0 (not defaunated) to 1 (highly defaunated). Red colors represent areas with high concentration of each driver while blue colors represent low concentration of each driver

Full size image

10 Functional Defaunation, Trophic Cascades and the Loss of the Atlantic Forest Top Predators

Species extinctions and reductions in animal populations promote pervasive trophic cascades that affect ecological processes and functions, disease dissemination, fire dynamics, biogeochemical cycles, and more (Dirzo et al. 2014). An iconic example is the case of Yellowstone National Park (USA), where the extinction of wolves (Canis lupus) triggered a series of trophic cascades through increasing the population of its prey, the elk (Cervus elaphus), which in turn caused significant changes in the habitat through limiting the recruitment of certain plant species and ultimately led to profound changes in riparian ecosystems (Beschta and Ripple 2018). Top-down control by herbivores is widely determinant for plant communities (Jia et al. 2018). Therefore, annihilation of such animals is likely to significantly modify the shape of ecological communities as we see now. Although defaunation is a local process, its effects could scale up globally, for example, through shifts in carbon stocks (Bello et al. 2015). Upscaling effects are likely to be more pronounced in tropical forests.

Most of the Atlantic Forest fragments are under high levels of defaunation, having lost animal populations from every functional group (Bogoni et al. 2018), with likely consequences for ecosystem structure and functioning. Estimates suggest that the largest apex predator (jaguar – Panthera onca), the largest herbivore (tapir – Tapirus terrestris), the largest seed predator (white-lipped peccary – Tayassu pecari), and the largest arboreal seed disperser (muriqui – Brachyteles spp.) are missing from 88% of the Atlantic Forest (Jorge et al. 2013). Jaguars are functionally extinct in most of its original distribution (Galetti et al. 2013a; Nagy-Reis et al. 2020) and less than 300 individuals are left in the whole biome, divided into sub-populations of extremely low densities, of which only three are considered viable (Paviolo et al. 2016). Historically, the main driver of jaguar population declines was land conversion and fragmentation, which not only reduced their habitat, but further resulted in population declines of their prey. Currently, retaliation to livestock predation and roadkill are the main threat to the few remaining jaguar individuals that still roam through the Atlantic Forest. According to the trophic cascade model such declines should contribute toward increasing prey population densities, higher consumption of edible plants, and a series of cascading effects that should alter entire ecological communities throughout the whole biome. Critically, empirical evidence for top-down control of herbivorous prey by jaguar is still missing, yet evidence of trophic cascades is well supported by defaunation studies on herbivory, seed dispersal, and indirect interactions at lower trophic levels.

11 The Loss of Critical Seed Dispersal Functions

Seed dispersal is a critical ecosystem process heavily affected by defaunation . About 89% of all woody plant species in the Atlantic Forest are animal dispersed (Almeida-Neto et al. 2008), with more than 331 vertebrate frugivore species (including birds, mammals, fishes, amphibians, and reptiles) in the biome (Bello et al. 2017). Vertebrate species inhabiting the Atlantic Forest vary in their functional role as seed dispersal agents, and therefore the differential defaunation might affect ecosystem processes in different ways. For example, muriqui monkeys (Brachyteles spp.) and tapirs (Tapirus terrestris) are efficient seed dispersers thought to play a complementary role in seed dispersal through their different size-selective and spatial seed deposition patterns. This is also true even within functional guilds, as different primate species also have different seed deposition patterns (Culot et al. 2017). Large terrestrial species contribute disproportionately to long-distance dispersal (Pires et al. 2018) and thus might play a critical role in genetic flow among populations of fruiting trees. Scatter hoarders are effective seed dispersal agents that reduce predation of seeds by insects, and whose dispersal function collapses in fragments below <1000 ha (Galetti et al. 2006).

By taking a closer look at the quality of seed dispersal mediated by Atlantic Forest mammals, evidence shows species may differentially affect seed fate because their feces attract different abundances and richness of dung beetles (Scarabaeidae: Scarabaeinae), which secondarily disperse those seeds (Lugon et al. 2017). Extirpation of large- and medium-sized herbivores in the Atlantic Forest has modified the structure of dung beetle communities and has likely caused co-extinctions (Fig. 14.5; Culot et al. 2013; Genes et al. 2019). Dung beetle abundance and diversity respond to mammal declines because they rely on mammalian droppings for feeding and nesting (Andresen 1999). While manipulating the feces, dung beetles secondarily disperse seeds to microsites that are generally favorable for germination and recruitment (Nichols et al. 2008). As a result, defaunation is likely to additionally affect plant communities through declines in secondary seed dispersal mediated by an impoverished community of dung beetles. Generally, smaller-sized dung beetles dominate in defaunated areas (Fig. 14.5; Culot et al. 2013). Since their size restricts the size of the seeds they can secondari0ly disperse, the dispersal of larger-seeded plant species could be negatively affected.

Fig. 14.5
figure 5

Evidence of the ecological consequences of defaunation in the Atlantic Forest. Non-defaunated areas (ND) in green and defaunated areas (D) in orange. (a) Mean abundance decline of large frugivorous birds (Galetti et al. 2013b) and arboreal and terrestrial mammal mean biomass loss in defaunated fragments (Galetti et al. 2017). (b) Wildlife extirpation leads to increases on small-mammal (SMa) and dung beetle abundance (DBa), while small-mammal (SMr) and dung beetle richness (DBr) decreases (Galetti et al. 2015a; Culot et al. 2013). (c) Defaunation also affects plant-animal interactions through increasing the number of seed disperser interactions that are missing (MI, Marjakangas et al. 2018), increasing seed predation (SP, Galetti et al. 2015b), increasing germination time (GT), and increasing the number of seedling (SR, Villar et al. 2020). Defaunation decreases seed dispersal (SD, Galetti et al. 2006), maximum seed dispersal distance (SDd) (Culot et al. 2017), and trampling (TR) (Villar et al. 2020). (d) Defaunation and changes in plant-animal interaction patterns alter ecological functions and services, increasing mean seedling species richness (SSr), total seedling biomass (Sb), and total seedling productivity (SPp, Villar et al. 2020), decline in seedling beta diversity (SBd), seedling species evenness (SSe) (Villar et al. 2020), available nitrogen in soil (NS), total phosphorus in soil (PS), and organic matter in soil (OS) (Villar et al. 2020), and reduction in the number of arbuscular mycorrhizal fungal (AMF) spores (MFa) and AMF morphotype richness (MFr) (Paz et al. unpub.) Loss of large bird seed dispersers caused the reduction of palm seed size (ESS, Galetti et al. 2013b). Defaunation of large herbivores also erodes carbon storage (Bello et al. 2015)

Full size image

Fruits from large and dense wood tree species in the Atlantic Forest are generally dispersed by large frugivores, and therefore the extinction of the latter may cause a reduction in up to 5.8% carbon stocked on forests over time (Fig. 14.5; Bello et al. 2015). Furthermore, the functional extinction of large seed dispersers and large seed and seedling predators can have synergic negative effects on the recruitment of large-seeded plants, reducing recruitment success from 30% in an intact community to 7.5% in a strongly defaunated one (Fig. 14.5; Culot et al. 2017), with subsequent effects on carbon storage.

The loss of seed dispersal function carried out by large frugivores has also a strong impact on the phenotypic and genotypic architecture of tree populations. In forest fragments smaller than 10.000 ha, bird dispersal of large seeds has been lost and only interactions between small generalist birds and small-seeded plant species remain (Emer et al. 2018). Using a key resource tree species as a model system (the juçara palm Euterpe edulis), recent studies have shown that the functional extinction of birds drives rapid evolutionary changes in seed size (Fig. 14.5; Galetti et al. 2013b), that also has a genotypic basis (Carvalho et al. 2016). Critically, such studies indicate that selection for smaller seeds on defaunated landscapes occurs on ecological time scales and at a very fast pace, suggesting that the functional extinction of large seed dispersers might have profound and perhaps irreversible consequences on seed traits over evolutionary timescales.

12 The Ecological Consequences of Losing Large Herbivores

If in one hand seed dispersal is crucial for the maintenance of high plant diversity in the Atlantic Forest, antagonist interactions also play an important role; however, empirical evidence is still scarce. Lower herbivory and seed predation rates due to loss of fauna are expected to have variable impacts on the trophic interactions, depending on the abundance of the animals and regional environmental condition. For example, the dominance of palatable plant species was found to increase with loss of herbivores, reducing the local diversity of plants in a tropical forest (Harrison et al. 2013). As a result of the loss of large herbivores and increase in dominance of certain plant species, both functional and life-form diversity among plant communities can be affected (Bulascoschi-Cagnoni et al. unpublished, Souza et al. unpublished), which might in turn lead to changes in ecosystem functions, such as C storage, decomposition of organic matter, water and nutrient retention (Lavorel and Garnier 2002).

Another effect of defaunation of large mammals in the Atlantic Forest is the competitive release of small mammals (a phenomenon also known as “rodentization”, Fig. 14.5; Galetti et al. 2015a), which can trigger subsequent compensatory seed predation by small mammals. For instance, it has been shown that this process leads to a twofold increase in seed predation of a threatened keystone palm species (Euterpe edulis, Fig. 14.5; Galetti et al. 2015b), and demographic simulations calibrated with empirical data suggest that this increase in seed predation might have further consequences for recruitment rates in other large-seeded species (Fig. 14.5; Culot et al. 2017). On the other hand, long-term experiments in the Atlantic Forest show that excluding large mammals increases the abundance of palm species and alters the dominance of plant species, and results in an increase in community-level seedling recruitment, primary productivity, and biomass (Villar et al. 2020). Clearly, more evidence is needed to discern in which ecological contexts defaunation leads to increases or decreases in plant recruitment. Nonetheless, the rodentization of the Atlantic Forest due to the loss of large mammals may also have potential consequences for the increase of emerging diseases, such as hantavirus (Galetti et al. 2015a).

Non-trophic effects also result from defaunation processes. Some large herbivore/omnivorous mammals such as the peccaries can be considered as “ecosystem engineers” because of their indirect effects on plants, vertebrates, and soil organisms (Fleming et al. 2014; Coggan et al. 2018). These indirect effects relate with their foraging activities as they disturb the soil surface in search for soil organisms and roots, creating spatial heterogeneity that benefits other vertebrates (e.g., creating optimum ponds for anurans; Beck et al. 2010), soil invertebrates (e.g., termites, Coggan et al. 2018), and soil microbial communities (Eldridge et al. 2016). In addition to decreasing seedling recruitment, primary productivity and biomass, large seed and seedling predators of the Atlantic Forest such as the white-lipped peccary (Tayassu pecari), for instance, can also have substantial effects on seedling communities through trampling and increased soil compaction (Villar et al. 2020). Furthermore, as a result of their top-down control of plant communities and foraging behavior in large herds peccaries can affect soil nutrient cycling (e.g., ammonium- and nitrate-N; Villar et al. 2020, Fig. 14.5) and soil carbon dynamics, which consequently influence plant productivity and alpha- and beta-diversity (Villar et al. 2020, Fig. 14.5) and the diversity of the root-symbionts (e.g., arbuscular mycorrhizal fungi, Paz et al. unpublished, Fig. 14.5). Yet, the evidence at hand still represents the tip of the iceberg, and the longer-term consequences of local extinctions or reductions in the abundance of large wildlife from the Atlantic Forest still remain largely unknown.

13 How to Revert Defaunation and Its Consequences: A Tool Kit

The Atlantic Forest lost 88% of its original distribution and its biodiversity is rapidly eroding, but paradoxically we have experienced a few global extinctions. Very rare populations have been monitored over large periods (e.g., Golden lion tamarin) and there is still a knowledge gap on most vertebrate species population size (Joly et al. 2014).

Given the extent of the consequences of defaunation in the Atlantic Forest, mitigation strategies are imperative. Habitat restoration would clearly be effective in building space for defaunation mitigation. Despite all the logistical challenges, ambitious programs such as The Atlantic Forest Restoration Pact aims to restore 15 million ha by 2050 (Crouzeilles et al. 2019). However, considering that most of the matrix between forest fragments are composed by agricultural lands or cities, restoration will not be suitable to every patch and vertebrates will not be able to re-colonize most forest fragments without being assisted. Reintroduction of the golden lion tamarin (Leontopithecus rosalia) has successfully rescued its population from the brink of extinction (Kierulff et al. 2012), a good example of species restoration project. Although some progress has been made toward understanding the extent to which ecological processes can be restored through rewilding (Genes et al. 2019), and which areas would best benefit from the reintroduction of seed dispersers (Marjakangas et al. 2018), it is still unclear how to scale-up such conservation strategy in practice. Reversing the pervasive defaunation that occurred in the Atlantic Forest is by no means a straightforward task. Nonetheless, it will be fundamental to assure the persistence of the biodiversity in the Atlantic Forest remnants.