Introduction

The science of Natural History is of so unbounded an extent, that perhaps I may be al- lowed, comparatively speaking, to say, that scarcely a day passes without an opportunity being afforded to zoologists of bringing to light unknown instances of its latent trea sures.

joshua brookes (1829:95), British anatomist and naturalist, in reference to his de- scription of the Plains Viscacha, Lagostomus.

n his introduction to the fi rst volume in this se- efforts of Jorge Crespo. In the intervening 50 years, the ries, Alfred L. Gardner (2008) detailed the history of recognized diversity of South American rodents at both the Ithe intention that he, Sydney Anderson, and James L. genus and species levels has increased by approximately Patton had to produce an updated reference to the mam- 160%. Perhaps not surprisingly, many of the taxa at all mals of South America, one that would provide a thorough categorical levels treated by Cabrera have also been re- review of knowledge about mammals dating from the mid- ordered in fundamentally different ways. To illustrate, we 1700s through today. The goal was to document the rich use taxon diversity trends for the largest group of South nomenclatural history for all taxa (families, genera, and American mammals: the rats and mice of the cricetid sub- species); describe morphological and other traits that aid family Sigmodontinae, which encompass some 62% of the in identifi cation, both in the fi eld and museum; delimit spe- generic diversity and 56% of the species diversity of rodents cies’ distributions by listing and mapping marginal records recognized in the accounts presented in this volume. Only confi rmed by voucher specimens and the published litera- about 34% of the genera and 39% of the South Ameri- ture; and summarize available observations on natural his- can rodent species we recognize today owe their discovery tory. Gardner’s volume 1, which covered marsupials, xen- and description to scholars of the nineteenth century. The arthrans, shrews, and bats, set an exemplary standard of fi rst two decades of the twentieth century saw an increase scholarship for this effort. This, the second volume in the of another 25% to the totals of both genus- and species- series, is restricted to the single order Rodentia, the most group taxa, followed by a 50- year rather quiescent period diverse of all mammalian groups worldwide, accounting (from 1921 until 1970) which saw only a modest growth in for more than 42% of species and 39% of genera (Wilson overall diversity at both taxonomic levels of less than 2% and Reeder 2005). All Recent and extant taxa are included; per de cade. Tracing these trend lines up to the time when those known only from the fossil record or now extinct are our project was initiated (in the 1970s), one might have not (e.g., Megaoryzomys from the Galapagos Islands, Ecua- surmised that most of the known taxonomic diversity of dor; Noronhomys, from Ilha Fernando de Noronha, Brazil; sigmodontines on the continent had already been discov- Dusimys, from Curaçao, Caribbean Netherlands), nor are ered and described. Such a conclusion, however, could not those introduced to the continent (such as beavers, Castor have been further from reality. Indeed, in the last 30 years, canadensis, and muskrats, Ondatra zibethicus, from North the generic diversity in South American sigmodontines has America, or the Old World rats [Rattus spp.] and mice [Mus increased by 27%, with 24 new genera added, and species spp.], family Muridae). diversity by 24%, with 56 new species described. And, in The fi rst substantive review of South American rodents the few months since this book went into production, ad- was Angel Cabrera’s seminal Catálogo, the rodent part of ditional species descriptions have appeared in the literature which was published in 1961 after his death, thanks to the (e.g., Gonçalves and Oliveira 2014). Fully one quarter of xviii Mammals of South America all genera and species of this group of rodents that are rec- sis on taxonomy and nomenclature. Our area of coverage ognized today have been described in the terminal three of includes nuclear South America as well as the continental the 25 de cades since Linnaeus (1758) established the earli- islands of Trinidad and Tobago and the Caribbean Neth- est currently accepted technical names, and the explora- erlands off the Venezuelan coast (Map 1). Along with ex- tion of the biota of South America began (see the historical tensive but not exhaustive synonymies, the accounts in this reviews of Neotropical mammalogy given by Hershkovitz volume include identifi cation keys and descriptions of each [1987b] and the Introduction in A. L. Gardner [2008]). order, family, genus, and species. Each species account also The remarkable rate of recent discovery and descrip- includes comments on distribution; lists of selected localities tion of South American sigmodontines is mirrored by most that are plotted on maps; comments on subspecies and, in other rodent groups (as well as for mammals in general a few cases only, lists of subspecies and their synonyms and worldwide [e.g., Reeder et al. 2007] or those limited to distributions; summaries of natural history information; and the Neotropical Realm [e.g., B. D. Patterson 2000, 2001]). discussions of issues related to type localities, taxonomic This increased rate owes its origin to many factors, perhaps interpretations, matters of nomenclatural importance, and the most important of which has been the establishment karyotype, if known. As with volume 1, this one contains its of exceptionally strong academic programs in systematic own Gazetteer and Literature Cited sections, with the latter and evolutionary sciences, including mammalogy, in most including all citations in the text as well as those cited in the countries on the continent, a development that has accel- references relating to dates of publication. erated in the past two de cades. Naturally coupled to ex- Also, as was the case for volume 1, the scope of individual pansion of academic programs has been the development contributions and thus authorship assignments is uneven. of professional organizations wherein annual or biannual The authority and responsibility for a given contribution meetings bring together scholars of all academic stages to continue until a new author line is encountered; for exam- share knowledge and inspire further research. These two ple, some authors were responsible for all entries in a single critical components in the development of the professional family, while multiple individuals singly or in combination fi eld of mammalogy have been directly tied to other factors authored the accounts in other families. The Acronyms and underlying the increase in our knowledge of South Amer- Abbreviations, Literature Cited, and Gazetteer sections were ica mammals, including but not limited to (1) substantially compiled from information provided by each account author. increased fi eld research, especially into the remote areas of With a few exceptions, the taxonomy of higher catego- this ecologically diverse continent; (2) an increase in the ries follows McKenna and Bell (1997), and the sequence diversity of survey methods for securing samples (particu- of family, genus, and species group taxa is based on the larly owl pellet analyses in Argentina and pitfall trapping third edition of Mammal Species of the World, edited by in Brazil); (3) computer- driven analytical methodologies to Wilson and Reeder (2005). Documentation by reference evaluate characters, compare morphologies in multivari- to more recent literature is provided for deviations from ate space, and construct the trees upon which phylogene tic either of these two general references, or is noted when systematics depend; (4) the explosion of molecular-based based on the personal views of the contributing authors methodologies, particularly DNA sequencing, which can be from their own, often unpublished, research. A. L. Gardner applied to historic specimens in museum collections as well (2008:xvi) provided the guidelines followed herein for the as recently collected materials; (5) the requisite shift in sys- synonymies, including citation of the original description tematic philosophy to one based on phylogene tic principles, and any unique name combination subsequently applied with taxon defi nition appropriately based on relationships by any author for the fi rst time to that taxon. Interested determined by shared- derived characters rather than over- persons should see his written synopsis for details. We, as all similarity, and (6) the recognition, through both policy did Gardner (see his dedication in volume 1), relied heavily implementation and fi nancial support, by governmental on the “scrapbooks” compiled by Remington Kellogg, the and nongovernmental agencies of the importance of basic former curator of mammals and director of the U.S. Na- systematic research as a key component of biodiversity as- tional Museum, in constructing synonymies. Fortunately, sessment, which in turn underlies sound environmental pol- much of the early eigh teenth and nineteenth century litera- icy and establishes conservation priorities. We intend that ture is now available in digital form from the Biodiversity this volume and the others in our series will identify gaps Heritage Library (http:// www .biodiversitylibrary .org /), a in our current knowledge and encourage additional fi eld, truly remarkable resource. Finally, J. R. Ellerman is credited museum, and laboratory research to fi ll those gaps. in this volume with the name combinations cited from his As noted by A. L. Gardner (2008) in his introduction The Families and Genera of Living Rodents; however, if a to volume 1, we selected Cabrera’s (1958–1961) cata log as name came from one of the lists and was available before the primary point of departure because of its strong empha- 1937, in many cases either R. W. Hayman or G. W. C. Holt Introduction xix compiled the individual list on which those name combina- landscape modifi cation or other factors in recent decades. tions are based. Nor should these range maps be viewed as static into even This volume deviates from volume 1 in limited but im- the near future, as continuing fi eldwork in remote areas portant ways. First, we asked account authors to provide will invariably expand those ranges now known and global a description of each species beyond those characters that climate change continues to impact both distribution and comprise the keys. Most rodents have not been mono- presence (e.g., Malcolm et al. 2006; Urban et al. 2012). Fi- graphed in recent decades or ever, and many species are nally, because the shapes of species’ ranges are diverse, often known only from their initial description. Moreover, an in- linear or geo graph ically restricted, we have not followed the creasing number of species have been defi ned in the past few uniform clockwise listing of localities, starting with N on a decades by karyotypic data (diploid number and arm num- compass, employed in volume 1. Rather than explaining de- ber of both autosomes and sex chromosomes) and/or DNA viations from this scheme in so many individual cases, local- sequences. As important as these characters are, neither the ities are simply listed alphabetically by country, department/ karyotype nor sequence are available for the vast major- province/state, and specifi c locality in all accounts. Virtually ity of specimens in museum collections nor for live animals all locality rec ords are based on vouchered specimen records captured during fi eld studies. As a consequence, reliance where the identifi cation could be verifi ed, either literature on keys alone for identifi cation in the fi eld or museum can records from revisionary accounts or the actual examination be problematic. We thus believe that descriptions of exter- of a museum voucher. We note, however, that the advent of nal and/or craniodental characters of each species is an widespread camera traps and archives of verifi able digital important addition to this volume, one enhanced by mul- photos—which have been used so successfully to document tiauthored accounts for large, geograph i cally widespread, new rec ords for large mammals— are beginning to provide and particularly vexing taxa, such as the sigmodontine insight even into smaller and often very secretive rodents genus Akodon or the echimyid genus Proechimys. Where (e.g., Blake et al. 2010). Although photos are unacceptable possible, the authors have provided the standard external replacements for vouchered specimens in systematic research, mea sure ments of head and body length, tail length, hind- a few accounts have included localities derived from unam- foot length, ear length, and/or mass, taken either from the biguous recent photographs of rare taxa with few records. literature or from labels of museum vouchers they exam- Inset maps depicting the range of the genus are also provided ined as well as characters of the pelage (color, color pattern, in the maps for each species. details of hair types), appendages, and both soft and cranio- Third, there is also substantial unevenness in coverage dental anatomy. However, these descriptions vary greatly in of natural history and other information on population bi- depth of detail owing to both the level of knowledge of in- ology, ecology, reproduction, and other attributes across dividual and geographic variation available and an author’s accounts. Some of this simply results from the actual quan- direct familiarity with specimens in museum collections. In tity of information available, as so many species of ro- some cases, our knowledge of a par tic ular species is so poor dents have never been studied in their natural habitats, but that little beyond the original description is available. some also stems from whether existing or relatively recent Second, we have allowed authors liberty in delineating synopses have already been published (e.g., the Mamma- locality records beyond just those circumscribing the edge lian Species accounts produced by the American Society of of a species’ range. Many species of rodents are so poorly Mammalogists, http:// www .mammalsociety .org /publications known that we believed it valuable to list and map virtu- /mammalian -species) . ally all known records in those cases. However, we caution Finally, the construction and use of vernacular names is a users of metadata, such as marginal locality maps in global vexing problem in mammalogy, as there is no standard set analyses of biodiversity, not to confuse a shaded range of rules that apply to their formation (but see Duckworth bounding locality records with actual occurrence through- and Pine 2003). This issue is exacerbated by the common out. Many species of rodents are habitat specialists and are application of En glish names (as given, for example, in certainly not evenly spread across the geographic landscape Wilson and Cole [2000] for mammal species worldwide) indicated by that shaded range. Shading surrounding the rather than use of the larger number of common terms ap- mapped localities is provided primarily to highlight those plied to those species by South American biologists, the locality placements and not to identify actual geographic lay public in each country, and/or indigenous groups. The boundaries, although the maps are drawn with as much common names employed herein are those of the account precision as possible given their scale. The ranges mapped authors and may refl ect either or both English and local also include locality records assembled over many decades; terms or equivalents. Many of these follow Wilson and as a result, they do not accommodate any range shift or Cole (2000) and/or those of account authors in Wilson and fragmentation that may have resulted from anthropogenic Reeder (2005).