Pollination and seed dispersal are two fundamental aspects of plant biology in which animal involvement can both increase genetic diversity, and reduce mortality (Vranckx et al., 2012). Although seed dispersal is key to maintaining plant diversity, seed dispersal by reptiles (saurochory) is sometimes regarded as a rare phenomenon (Valido and Olesen, 2007). This notion is challenged by growing evidence of reptiles as seed dispersers (González-Castro et al., 2015), especially because of the tendency for most reptiles to gulp fruits whole, which provides little opportunity for seed damage.

South American river turtles consume fruits, and these can then be dispersed along water margins (Moll and Jansen, 1995) or carried further during long-distance migrations (Kubitzki and Ziburski, 1994). Amazon giant river turtles (Podocnemis expansa, up to 55 kg; Smith, 1979) are known seed dispersers and migrate great distances between feeding and nesting grounds. This species was considered overwhelmingly abundant in the past (Von Humboldt, 1877), but has declined due to hunting and egg collection. All three continental tortoises in South America (Chelonoidis carbonarius, 5.9 kg; C. chilensis 1.7 kg; and C. denticulatus, 6.4 kg) are known seed dispersers (Varela and Bucher, 2002; Strong and Fragoso, 2006). Moreover, they can reach high densities; up to 125 animals/km² for C. carbonarius and C. denticulatus combined (Moskovits, 1988), and go up to 54 kg in the largest species, C. denticulatus (Pritchard and Trebbau, 1984). The two large and widely distributed lizard genera in the Neotropics, tegu, and iguana lizards (Tupinambis and Iguana), also disperse seeds (Benítez-Malvido et al., 2003; de Castro and Galetti, 2004), and occur at high densities, reaching 83 and 122 individuals/km² for tegus and iguanas, respectively (Munoz et al., 2003; Bovendorp et al., 2008).

Asiatic species like the forest tortoise (Manouria emys, up to 37 kg) and northern river terrapin (Batagur baska, up to 30 kg) are potential seed dispersers (Standford et al., 2015) but this remains uninvestigated. Although they are the largest hard-shelled chelonian species in Asia, both lack reliable population estimates, but are under extinction risk (IUCN, 2000; ATTWG, 2016). Asiatic frugivorous lizards like the Philippine varanids (Varanus bitatawa, V. mabitang, and V. olivaceus) reach up to 10 kg and act as seed dispersers (Bennett, 2014). All Philippine varanids are considered under risk by the IUCN (Gaulke and Demegillo, 2008).

In Africa, the huge and widely-distributed leopard tortoise (Stigmochelys pardalis; up to 48 kg), and spurred tortoise (Centrochelys sulcata; up to 93 kg) are both potential seed dispersers (Milton, 1992). While the role of leopard tortoise as a seed disperser has been investigated, any effort has been made to characterize the seed dispersal potential in spurred tortoises. Reported densities of leopard tortoises reach 85 individuals/km² (Mason et al., 2000). There are no population records for spurred tortoise density, though wild populations in its native Sahel are now scarce, and it is considered Vulnerable (IUCN, 1996). Some smaller African reptiles, such as spiny tailed lizards (Uromastyx aegyptia, up to 3 kg), are seed dispersers in desert environments (Wilms and Wagner, 2009), and can reach densities of 400 individuals/km² (Kordges, 1998) but suffer from extensive persecution (Figure 2C) and are considered Vulnerable by IUCN (Wilms et al., 2012).

On tropical islands, giant tortoises are frequently the main seed dispersers (Blake et al., 2012). Many of these species have already been exterminated (Rhodin et al., 2015): five became extinct after human colonization of the Pacific Ocean's Mascarene Islands (Cylindraspis triserrata, C. vosmaeri, C. peltastes, C. inepta, and C. indica), while two were lost on Madagascar (Aldabrachelys grandidieri and A. abrupta), and seven from the Caribbean archipelago (including Chelonoidis alburyorum, C. cubensis, and five other undescribed species), and four from the Galápagos (Chelonoidis abingdoni, C. nigra, C. phantastica, and Chelonoidis sp.). These past losses are currently impacting seed dispersal on the islands of Madagascar and Mauritius; respectively, baobab and ebony species are at risk due to the absence of their co-evolved chelonian dispersers (Moolna, 2008; Andriantsaralaza et al., 2014). Since 57% of Earth's tortoises, including several large-sized continental species, were extirpated or made extinct by early or modern Homo species during the Pleistocene (Rhodin et al., 2015), similar effects in other ecosystems may have occurred or may still be occurring.

Evolutionary unique groups of giant tortoises were also eliminated during human colonization of Sahul (present New Guinea, Australia, and Tasmania, before sea levels rose). The 200 kg-plus Meiolanidae, a family of horned tortoises, were giant herbivores. Anthropogenic causes meant they had disappeared long before we could understand their ecological role in ecosystems. Insular seed disperser extinctions were not restricted to tortoises: the giant iguanas (Brachylophus gibbonsi, Lapitiguana impensa) of Tonga and Fiji archipelagos in Pacific also vanished upon human colonization, causing disruption to the island's seed dispersal ecology (Meehan et al., 2002).

Pollination by reptiles (Figure 1C), although mostly restricted to insular lizards, is a widespread phenomenon occurring on islands in all tropical seas and includes at least 50 gecko species (Godínez-Alvarez, 2004). Low abundances of insect, both as prey and pollinators, probably drove insular lizards to consume fruit pulp and nectar. This, combined with density compensation, niche expansion, and low predation pressure, has resulted in lizards being important pollinators on many island ecosystems. Given their restricted distributions and vulnerability to introduced species, insular lizards are a highly threatened taxa (Olesen and Valido, 2003). The largest skink on Earth, Chioninia coctei, went extinct after European colonization of Cabo Verde (Vasconcelos et al., 2013). Several other extinctions have occurred between pollinating lizards, including gecko pollinators as Phelsuma edwardnewtoni and P. gigas in the Mascarene archipelago (Hansen et al., 2006).

When the area for colonization is limited, the encrustation of other animals may become a viable strategy for sessile and sedentary marine animals. Reptiles generally have a hard tegument and can consequently act as habitat for epibionts—organisms that live on the surface of others. Barnacles (infraclass Cirripedia) are found on all species of marine turtles (Hayashi, 2012). As a result, sea turtles are known to play key roles in barnacle phylogeography (Pinou et al., 2013).

Crocodilians that inhabit brackish waters can also play host to barnacles and epibionts, and these have been found on Crocodylus porosus (Monroe and Garrett, 1979), Crocodylus acutus (Cupul-Magaña, 2011), and subtropical species as Alligator mississippiensis (Nifong and Frick, 2011). Barnacles have also been found on unexpected hosts such as brackish water turtles (Arndt, 1975; Seigel, 1983; Cedillo-Leal et al., 2013), and marine snakes, which are associated with 48 barnacle species (Pfaller et al., 2012).

In sea turtles, barnacles are only one of the many associated epibionts, and more than 200 species have been found on a single turtle species (Frick et al., 2004). A known epibiont of sea turtles are remoras (Echeneidae), which have been recorded associating with all sea turtles (Lazo-Wasem et al., 2011). Sessile forms include algae, bryozoans, cnidarians, foraminiferans, mollusks, poriferans, and tunicates. Sedentary species include amphipods, polychaete worms, and tanaids. The minitature reef produced by sessile and sedentary epibiota creates conditions for colonization by motile organisms such as annelid worms, copepods, echinoderms, decapods, dipterans, ostracods, platyhelminth worms, protozoans, sipunculid worms, and peracarids, which sometimes results in the formation of an ecosystem on the turtle hard integument (Frick et al., 2004).

Organism movement by reptiles is likely to have several benefits for the associating species: (a) seed survival is greater away from the parent plant (Fricke et al., 2013); (b) pollination enhances the volume, speed, and synchrony of fertilization (Kormann et al., 2016); (c) epibiosis allows species that attach to reach specific sites that are favorable for survival, such as nutrient- or prey-rich microhabitats (Hayashi and Tsuji, 2007). Finally, at larger scale, these movements may allow associated organisms to colonize vacant habitats and, or, new regions. However, many of the reptiles performing gene transport roles in tropical regions are threatened with extinction (Figure 2A) or are already extinct—with severe knowledge gaps on role of gene dispersal at Table 1.

Large animals disproportionately drive lateral nutrient movement, and their widespread demise is disrupting the processes that result in the efficient distribution of nutrients from the oceans on to dry land (Doughty et al., 2015), as in aquatic turtles that lay on land. While biogeochemical cycles determine the limits of nutrient flows from terrestrial to aquatic ecosystems, the movements of these organisms can partially reverse this trend (Helfield and Naiman, 2001). Large reptiles can act as nutrient transporters in two main ways: (a) through reproduction, when aquatic reptiles lay eggs in terrestrial environments (Bouchard and Bjorndal, 2000), and (b) through predation, when aquatic reptiles are killed and dragged onto land by terrestrial predators (Veríssimo et al., 2012). Due to reproduction being a risky phase of aquatic reptile life history, during which predation is common, those processes are not necessarily separate.

Reproduction-based nutrient transport can be demonstrated in the worldwide phenomena of turtles nesting on sandy beaches (Bouchard and Bjorndal, 2000). These annual events result in the terrestrial transfer of enormous volumes of nutrients originally gathered by turtles on aquatic feeding grounds. For example, the Mexican Pacific population of the olive ridley turtle was estimated to consist of ~10 million individuals (Spotila, 2004), of which some 5 million were likely to be adult females averaging 35 kg (Miller, 1997). Olive ridley turtles breed annually, laying 1–3 clutches averaging 2.8 kg (Elgar and Heaphy, 1989). This would result in the annual transfer of 14,000–42,000 tons of egg-based material on to land, a considerable entry of nutrients for an oligotrophic sand beach. Similarly, in freshwater beaches of Amazonia (Figure 1D), more than 6,000 giant river turtles, each weighing around 35 kg, offer an input of more than 16 tons of eggs on a single beach each year (Alho and Pádua, 1982). Considering just the Brazilian population monitored in a subsample of 120 nesting beaches (Cantarelli et al., 2014), would give a total of nearly 2,000 tons of eggs. However, these figures are for only one species and one set of beaches. When the other six Podocnemididae species of river turtle in Amazonia are also considered, and the other countries and beaches in which the species also nest, the annual volume of nutrient transfer represented by deposited eggs may be relevant to nutrient cycling in the amazon basin. A ~50% emergence rate (Hernández and Espinosa-Blanco, 2010; Brost et al., 2015), indicates that half of these nutrients will remain to enter terrestrial food chains in both fresh and saltwater ecotones. Nutrients can also directly enter into the terrestrial food chains through plant root absorption in both fresh and saltwater (Stegmann et al., 1988; Hannan et al., 2007); not to mention the egg shells that are slowly absorbed. If these figures are amplified to include multiple sea and freshwater turtle species nesting in multiple beaches around the globe, and we have a nutrient transport phenomenon of global scale and impact that may parallel or surpass the much-celebrated salmon-derived nitrogen-transport system (Helfield and Naiman, 2001).

Since aquatic turtles, most lizards and crocodilians need to lay their eggs on land, this is far from being a uniquely South American phenomenon. In Africa, several large aquatic reptile species, including Nile giant softshell turtles (Trionyx triunguis, up to 100 kg; Keller, 2005) and Nile crocodiles (Crocodylus niloticus, averaging 44 kg; Hutton, 1987) are abundant, widespread and nest on sandy beaches. Meanwhile, in tropical Asia, sandy beach-nesting species include giant softshell turtles of the genera Chitra and Pelochelys (which can reach up to 120 kg; Bonin et al., 2006), batagur river turtles, gharials (Gavialis gangeticus, averaging 31 kg; Whitaker and Basu, 1982) and mugger crocodiles (Crocodylus palustris, averaging 65 kg; Bayani et al., 2011). Although the nests of these species have the potential to transport the embodied nutrients into terrestrial ecosystems, this widespread phenomenon has been little-addressed. With the exception of Nile and mugger crocodiles, all of these sand-nesters are now extirpated throughout most of their ranges, and are all considered under risk (Stevenson and Whitaker, 2010; Rhodin et al., 2011).

Predators can consume aquatic reptiles on a large scale. Jaguars (Panthera onca) are the largest felids in the Neotropics, and crocodilians can represent more than 70% of its prey (Da Silveira et al., 2010). Caimans are consistently killed by jaguars in Neotropical floodplain habitats as the Pantanal (Azevedo and Verdade, 2012), Amazon (Emmons, 1987), and Llanos (Scognamillo et al., 2003). In Central America jaguar prey on nesting sea turtles. Such nesting events can attract jaguars to the beach in their dozens, and modify their ranging behavior (Carrillo et al., 2009; Guilder et al., 2015). So extensive is the predation that, at some sites, sea turtles can comprise up to a third of the total annual jaguar diet (Chinchilla, 1997). This was probably a widespread behavior before human colonization eliminated the majority of both coastal jaguar populations, and sea turtle nesting areas.

Once killed by predators, aquatic reptiles are often dragged for hundreds of meters inland (Carrillo et al., 2009; Azevedo and Verdade, 2012) where the uneaten remnants of their carcasses will decay, releasing nutrients in terrestrial ecosystems. As in other cases, this phenomenon is not restricted to South America. In the Sundarbans mangrove of India, tigers prey on water monitors (Varanus salvadorii) that represent as much as 12% of their diet (Mukherjee and Sarkar, 2013). In Africa, fish eagles (Haliaeetus vocifer) and crowned eagles (Stephanoaetus coronatus) are known to prey on the semiaquatic Nile monitor (Boshoff and Palmer, 1994; Stewart et al., 1997), which in turn is a predator of crocodile nests and juveniles (Kofron, 1989). African fish eagles also prey directly on young Nile crocodiles (Cott, 1961). In Australia, water goannas (Varanus mertensi) are prey for dingoes (Canis lupus dingo; Sutherland et al., 2011), while both varanids and dingoes are frequent predators of crocodile eggs (Somaweera et al., 2011).

With high standing crop and large body size, reptiles may have a strong positive impact on nutrient flow from aquatic to terrestrial tropical ecosystems. As consumers of aquatic organisms and as prey for terrestrial predators, they can act as important nutrient vectors. While some species may still be highly abundant in some areas (Campos and Magnusson, 2016), numerous populations are in severe decline (Thorbjarnarson, 1999; Rhodin et al., 2011). In fact population sizes of both turtles and crocodilians are now well below the historical levels witnessed by the first naturalists on scientific expeditions (e.g., Von Humboldt, 1877), suggesting that for many species (Figure 2B), current numbers are a mere relic of their former abundance. In Table 2, further examples of threatened reptiles providing nutrient movement—as well as current gaps on knowledge—can be seen.

Reptiles can act as top predators in aquatic and terrestrial ecosystems. Crocodilians are conspicuously present as predators in aquatic-terrestrial interfaces in the tropical regions of all continents. Like other predators, crocodilians can exert top-down pressure on prey species populations (Nifong and Silliman, 2013). Such interactions have the potential to regulate prey populations, as well as modulate their access to key resources, such as water (Doody et al., 2007). Opposed to the opinions of those who study mammalian carnivores, crocodilians are the largest predators in tropical regions on all continents. Indeed, within a given area, pythons and crocodilians can surpass mammalian carnivores both in numbers and biomass by several orders of magnitude (Campos and Magnusson, 2016; Natusch et al., 2016).

Squamate reptiles, like snakes, are capable of overpowering bird and mammalian prey (Figure 1B). They are also capable of strongly impacting avian prey populations (Patten and Bolger, 2003), and intense declines have been recorded where exotic snakes species have been introduced, as, for instance, on the island of Guam (Richmond et al., 2014). One could argue that prey suppression occurred because Guam is an insular ecosystem with prey communities in delicate equilibrium. However, the same pattern of decline is seen in species preyed upon by introduced pythons in Florida's Everglades. Unlike Guam, the Everglades is a functionally diverse continental mammal community composed of large and medium-sized species where prey declines have been shown by correlational and experimental methods (Dorcas et al., 2012; McCleery et al., 2015).

Gigantism in reptiles is commonly considered an island phenomenon, a by-product of competition-release; the Komodo dragons (Varanus komodoensis) being a frequently cited example. Since body size is such an important factor in species interactiveness, this misconception must be addressed: (i) several living giant snake species such as anacondas and pythons (e.g., Eunectes notaeus, E. murinus, Python molurus, P. reticulatus, P. sebae) occur in mainland environments, where they successfully compete with mammalian carnivores; (ii) Komodo dragons were formerly widespread over Sahul (Hocknull et al., 2009), where they competed with Thylacoleo carnifex, an extinct 100 kg-plus marsupial carnivore (Wroe et al., 1999), so that their current distribution is a misleading relict; (iii) other species of large lizards occur on mainland Asia, such as the water monitor, a species that exceeds 20 kg, which is smaller than only two of southeast Asia's largest mammalian predators, the tiger (Panthera tigris, 227 kg) and the leopard (Panthera pardus, 42 kg; Corlett, 2011); (iv) in the early Pleistocene, southeast Asia had giant species of monitors like Varanus sivalensis that coexisted with carnivore faunas similar to current ones (Hocknull et al., 2009). Despite this, reptiles are often completely ignored as predators in both present (Jorge and Galetti, 2013), and past ecological communities (Malhia et al., 2016). As suggested from this brief list, the popular notion of food chain apices occupied solely by mammalian carnivores must be reassessed in tropical ecosystems.

Insular giant tortoises are another example of a popular misconception regarding reptile body size: rather than being a byproduct of competition release, they evolved from continental species that were already large (Braithwaite et al., 1973; Pritchard, 1996). For instance, the largest tortoise ever, Megalochelys atlas, was an early Pleistocene continental species, which disappeared from mainland Asia and its nearest islands after the arrival of Homo erectus (Hirayama et al., 2015). An entire genus of extinct continental tortoises (Hesperotestudo spp.) contained around a dozen species that all surpassed 100 cm in length. The genus was exterminated in Central and North America following human colonization in the Pleistocene (Rhodin et al., 2015). In addition, it is worth noting that current large-sized continental species like the african spurred tortoise and the asian forest tortoise are likely to have reduced in size, as a result of a long history of exploitation (Lomolino et al., 2001).

Chelonians, popularly regarded as slow moving animals, can also act as top predators in aquatic ecosystems. The alligator snapping Turtles (Macrochelys species, 21 kg), snapping turtle (Chelydra serpentina, 6.2 kg), and several Asiatic giant softshell turtles (Rafetus swinhoei, Chitra indica, C. chitra, Pelochelys cantorii, up to 200 kg), can be top consumers of fish, birds and mammals in aquatic environments, and themselves have no predators once they have reached adult size (Steyermark et al., 2008; Corlett, 2011).

As predators, reptiles can also provoke the same types of human-wildlife conflict as large mammalian carnivores. Small and large livestock species are common prey for reticulated pythons in Southeast Asia (Shine et al., 1998), and for anacondas in South America (Miranda et al., 2016), while cattle, buffalo and sheep fall prey to the Burmese python in Pakistan and India (Goursi et al., 2012; Bhupathy et al., 2014). Crocodiles are also known to take a toll on livestock in both Asia and Africa (Butler, 2000; Bhatnagar and Mahur, 2010). Since reptiles do not possess the same charisma as large carnivores (Bonnet et al., 2002), this research theme and conservation issue is substantially under-explored (Figure 2D).

Reptiles can act as scavengers. Varanids in Australia, Asia, and Africa are consumers of carrion (Shine and Harlow, 1998; Dalhuijsen et al., 2015), removing flesh from carcasses before they spread disease and increasing the flux of energy in trophic cascades. The Nile crocodile plays this role when scavenging the carcasses of the hundreds of ungulates that die crossing large rivers during annual migrations, through which they may reduce the spread of disease and increase water quality (Subalusky et al., 2017). While ornithologists have quantified the scale of carcass cleaning service by vultures (Pain et al., 2003), similar knowledge of carrion use by reptiles is scant. This is an interesting research gap as well as conservation issue.

As with other processes described here, the action of reptiles as trophic actors has been severely reduced by extinctions and extirpations. Several species of crocodilians, for example, are already lost. The human expansion during the late Pleistocene-Holocene wiped out the last clade of terrestrial crocodiles (Mekosuchinae). They disappeared sequentially in Sahul (Sobbe et al., 2013), New Caledonia (Mead et al., 2009), Vanuatu (Mead et al., 2009), and Fiji (Irwin et al., 2011) with dates that corresponded to human arrival on each of these islands, around 53 and 3 thousand years ago in Australia and elsewhere, respectively (Roberts et al., 1994; Worthy et al., 1999; Stevenson et al., 2001; Kinaston et al., 2014). Although Sahul had a terrestrial predator fauna with large carnivorous mammals and reptiles, on smaller islands terrestrial mekosuchines were the top predators. In addition, humans caused the extinction of the Malagasy horned crocodile (Voay robustus) in the Holocene (Brochu, 2007), and the dramatic range contraction of the (now) Cuban crocodile (Crocodylus rhombifer), which occurred after human colonization of the Caribbean islands (Steadman et al., 2015). Together, these form part of the series of lost top predator roles played by reptiles. The unique ecological roles filled by Crocodilians continue under threat, with nine of the surviving 25 modern species considered to be threatened with extinction, seven of which are recognized as “Critically Endangered” under the IUCN categorizations.

Squamate reptiles, although capable of sustaining high cropping rates (Natusch et al., 2016), have also lost several species. Besides the widespread extirpation of the relictly insular Komodo dragon, the largest lizard species (Varanus priscus, averaging 320 kg and toping 7 m in length; Molnar, 2004) was eliminated in Sahul during human colonization (Flannery, 2002). Anthropogenic extinctions of giant snakes are also known. The last species of the Matsoiid snake family, Wonambi naracoortensis—a species reaching five meters long—and two other large species, W. barriei and Yurlunggur camfieldensis, also became extinct following aboriginal colonization (Flannery, 2002). Finally, the current status of some giant carnivorous turtles is far from good (Rhodin et al., 2011); hopes for the survival of the largest of their kind, R. swinhoei, is currently dependent on the last captive pair. Several other examples of top consumers roles performed by extinct and extant reptiles and the deart of knowledge on them can be seen in Table 3.

An ecosystem engineer is any organism that creates or significantly modifies habitat structure, and as a result is important for maintaining the biodiversity of the environment in which it lives (Samson and Knopf, 1996). Some herbivorous reptiles can have a profound effect on vegetation. The grazing and browsing impact of large tortoises, for example, can intensely impact vegetation composition and structure (Hamann, 1993). Giant tortoises on oceanic islands can have a great impact on their environments and exert top-down pressure, and have been described as ecosystem engineers (Gibbs et al., 2010). Considering that giant tortoises were—and in some places are—conspicuous elements of continental faunas the possible impacts of their loss must be assessed.

A similar relationship of habitat engineering is known from marine ecosystems. The seagrass (Thalassia testudinum) is heavily grazed by the green turtle (Chelonia mydas). As a result, the seagrass productivity and composition is positively affected by grazing (Moran and Bjorndal, 2005). In a similar way, the hawksbill turtle has a role as a sponge predator, with preference for overabundant aggressive sponges such as Chondrilla nucula. The predation by Hawksbill Turtles affects reef diversity and succession by influencing space competition (Bjorndal and Jackson, 2003). As the Green and Hawksbill turtles are severely overexploited, their regulatory impact on marine ecosystems has been much reduced.

Another means by which reptiles can act as ecosystem engineers is via their burrows, which can provide shelter for other species. Burrowing reptiles are recognized in several dry environments (Kinlaw, 1999), and their burrows are used by a wide diversity of animals: over 300 species have been recorded using those of the gopher tortoise (Gopherus polyphemus, Kinlaw and Grasmueck, 2012), including burrowing owls (Athene cunicularia), eastern diamondback rattlesnakes (Crotalus adamanteus) and marsh rabbits (Sylvilagus palustris). While the gopher tortoise is the best-studied burrow-providing reptile, ecosystem engineering through burrows can potentially occur in several other species, including the burrowers bolson tortoises (Gopherus flavomarginatus), Mohave desert tortoises (G. agassizii) and Sonoran desert tortoises (G. morafkai) in North America and red-footed tortoises in South America. Spurred tortoises in Africa can dig burrows that are both extensive and complex, but whose effects on other species are still unknown (Kinlaw and Grasmueck, 2012). Spurred tortoise burrows have great ecosystem engineering potential, as they are huge, up to 15 m long (Devaux, 2000) and occur in Sahel semi-desert landscapes, where cover is otherwise scarce. Ecosystem engineering also occurs with smaller reptilian species: <1 kg spiny tailed lizards dig burrows in coverless arid environments and have been shown to be ecosystem engineers by providing shelter to a variety of other organisms (Wilms et al., 2010).

Crocodilians can act as providers of shelter, digging extensive burrow systems in those areas of Asia, Africa, and North America where avoiding seasonal harshness requires aestivation (Whitaker et al., 2007; Brito et al., 2011). While such burrows may include galleries several meters deep, their use by other species is little known. During the dry season in the Everglades, wallow holes created by alligators (Figure 1A) act as a refuge for several species of fishes, birds, and aquatic reptiles, providing areas of deeper water that sustain populations until they can colonize more extensive water bodies at the onset of the next high water season (Palmer and Mazzotti, 2004).

Another way in which crocodilians can act as ecosystem engineers is via the mounds of rotting vegetation they assemble for egg laying. The subtropical american alligator nest mounds are used by other reptiles as nesting habitats. They include eight turtle species, two snakes and two lizards, which all lay their eggs in alligator mounds (Deitz and Hines, 1980; Kushlan and Kushlan, 1980; Elsey et al., 2013). While some species do this facultatively, the Florida red-bellied turtle (Chrysemys rubriventris) is a specialist mound-nester, and a single alligator nest mound may hold up to 200 of their eggs (Deitz and Hines, 1980). This phenomena is mirrored by amazonian yellow-spotted river turtles (Podocnemis unifilis), which sometimes nest in black caiman (Melanosuchus niger) mounds (Maffei and Da Silveira, 2013), and by basilisks (Basiliscus vittatus) and red-eared sliders (Trachemys scripta) which lay eggs in the nest mounds of Morelet's crocodile (Crocodylus moreletii; Platt et al., 2008). The apparent reduced richness of commensals in the nests of species other than alligators may be related to the fact that alligators are the best studied crocodilian species, suggesting that the subject merits further research in other crocodilians.

Several species with potential to act as ecosystems engineer are now threatened or extinct and severe knowledge gaps remains (Table 4). Do the mounds and holes of the chinese alligator (Alligator sinensis) have similar effects on the aquatic communities in the Yangtze River Basin? We may never know, because these 1.5 m, 36 kg animals are functionally extinct in the wild (Ding et al., 2001). This fate is yet another to be added to the long list of vanished or functionally extinct reptiles already cited in this text, that continues to grow every year (Turvey et al., 2017).

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.