An annotated checklist of freshwater Copepoda (Crustacea, Hexanauplia) from continental Ecuador and the Galapagos Archipelago

Abstract An annotated checklist of the free-living freshwater Copepoda recorded in different regions in Ecuador (including the Amazon, the Andes, the coastal region, and the Galapagos Islands) is here provided. We revised all published records, critically evaluated the validity of each taxon and provided short taxonomic and biogeographical remarks for each one. A total of 27 taxa have been reported, including species and records at the generic level only. The species and taxa identified only up to the generic level belong to five families and 14 genera. The Cyclopoida is the most diverse group with 16 records belonging to species (or identified to the generic level only) and eight genera, followed by the Harpacticoida with six species, one identification to the generic level only, and four genera, and Calanoida with four species belonging to two genera. A total of 18 taxa are recorded for the Andes. Six have been recorded in the Amazon, two are recorded for the coastal region, and six for the Galapagos. One species is shared between the Amazon and the Andes. One species is shared between the coastal region and the Amazon. Seventeen are only reported from the Andes and four are only reported from the Amazon. At the current status of the knowledge, any attempt to analyze and generalize distributional patterns of copepods in Ecuador is premature due to the scarcity of available information, and evidently there is an urgent need for more extensive field collections. A few working hypothesis for future studies are identified.

Similarly to the freshwater Cladocera and Rotifera (López et al. 2018a(López et al. , 2018b, our knowledge of the Copepoda of Ecuador in comparison to other countries in tropical South America is relatively recent and very limited. This is in sharp contrast to the great habitat diversity in the country, ranging from Amazon rainforest, including uphills and the lowlands, to alpine tundra paramo (more than 4000 m a.s.l.) and to the inclusion of Ecuador as a hotspot of biodiversity for plant and vertebrate species (e.g., Myers et al. 2000;Brummitt and Lughada 2003;Rieckmann et al. 2011).
As part of an ongoing project dedicated to collecting and revising the Copepoda, Cladocera and Rotifera from inland water bodies of Ecuadorian mainland and the Galapagos Islands, we assembled a list of the inland water Copepoda known to date for the country and provide a short discussion of relevant nomenclatural issues and known geographic distribution of the species. Our goal is to identify the major information gaps and pave the way for future studies on the Ecuadorian freshwater copepods, which ultimately might allow better framing of the copepod fauna of Ecuador in the Neotropical region and understanding its origin and affinities.

Methods
The list of the copepods of continental Ecuador and Galapagos Islands compiled herein is based on literature data, including theses and taxonomic and ecological publications. The current valid species names and combinations are mostly based on Dussart andDefaye (2002, 2006) and the WoRMS database (http://www.marinespecies.org). Here we adopt the classifications of Boxshall and Halsey (2004) and Kohdami et al. (2017), who have included the Poecilostomatoida families within Cyclopoida.
The geographic distribution of the freshwater taxa within the country is described by dividing continental Ecuador into three subregions (Andean, coastal, and the Amazonian subregions; see Steere 1950) to which the Galapagos Islands are to be added (Fig. 1). References to other regions within South America follow the biogeographical classification proposed by Dussart (1984).
Abbreviations used in the text: enp1-3 first to third endopodal segment; exp1-3 first to third exopodal segment; P1-P5 first to fifth legs.

Results
Twenty-seven records have been reported in literature from the inland water bodies of Ecuador, including the Galapagos Islands. The species and taxa identified only up to the generic level belong to five families and 14 genera. The Cyclopoida is the most diverse group with 16 records belonging to species (or identified to the generic level only) and eight genera, followed by the Harpacticoida with six species, one identification to the generic level only, and four genera, and Calanoida with four species belonging to two genera (Table 1). Eighteen taxa are recorded for the Andes, and six for the Amazon. One cyclopoid species is shared by the Amazon and the Andes. One cyclopoid species is shared between the coastal region and the Amazon. Seventeen are restricted to Andes and 4 confined to Amazon. Two species are recorded for the coastal region, and six to the Galapagos Islands. Remarks. According to Löffler (1963) the type locality lies somewhere between the towns of Villarrica and Llanquihue (straight-line distance between the towns, 227 km), in the southern Chile. The color is distinctly violet and the length of the specimens from the type locality varies between 370-560 μm for males and 400-700 μm for females. In Ecuador, the specimens were larger, the females reaching a length of 900 μm and the males 700 μm. The Ecuadorian males are variable in the armature of the endopodite in P2 and P4. Löffler, 1963 Distribution. Andes (Löffler 1963).

Remarks.
Originally described from Ecuador. Of the studied males and females, Löffler (1963) found that the P2 enp and the P4 enp can be asymmetric in armature. The dorsal ornamentation of the urosome is also variable in the male, and it can be either dorsally absent or present on the 2 nd to 4 th urosomites. Males measure are 540-700 μm long, and females are 800-980 μm long. This species was found in a high mountain pond in the southern Colombian Andes (Gaviria and Defaye 2012).

Bryocamptus Chappuis, 1929
Distribution. Torres and Rylander (2006) mentioned Bryocamptus from Ecuadorian highland lakes. However, this genus is basically boreal, with a few representatives Table 1. Distribution of the taxa in the four geographical regions of Ecuador. "×" indicates the occurrence of a calanoid "resembling Notodiaptomus amazonicus" from Lake El Junco in San Cristobal island.  (Reid 1993), and a single species, Bryocamptus (Bryocamptus) campaneri (Reid 1993), from Central Brazil. Records of Bryocamptus from a lake in the state of Rio de Janeiro (Reid and Esteves 1984) are a misidentification of Attheyella (Chapuisiella) fuhrmanni (Thiébaud, 1914) (Reid 1993). Bryocamptus broiensis Rocha and Matsumura-Tundisi, 1976 described for the state of São Paulo is recognized by Reid (1993) as Attheyella (Delachauxiella) broiensis Reid, 1994. According to Löffler (1972), the North American species of Bryocamptus do not occur south of the northern limit of the Eocene-Miocene submergence of Central America.

Cletocamptus axi Mielke, 2000
Distribution. Collected from lagoons of the islands of Santa Cruz and Floreana, Galapagos Archipelago. Mielke (2000) referred to the type locality as "Floreana: lagoon behind the beach".

Cletocamptus ecuadorianus Löffler, 1963
Distribution. Andes (Löffler 1963, as C. deitersi ecuadorianus). Remarks. Originally described from Ecuador. Length of males reaching 620 μm long, females 750 μm long. Asymmetry is observed in the armature of the female P5 basoendopod. Both males and females show variability in the armature of the antennal exopodite. In males, P3 exp3 may be variable in armature. C. deitersi (Richard, 1897) has been recorded from Ecuador (Löffler 1963 as C. deitersi ecuadorianus), Venezuela (Escaravage and Castel 1989), Peru and Bolivia (Harding 1955), Haiti (Kiefer 1936), and USA (California; Dexter 1995). However, several authors (Dexter 1995;Suárez-Morales et al. 1996;Gee 1999;Mielke 2000Mielke , 2001 have suggested that C. deitersi consists of a number of morphologically indistinguishable sibling species (Gómez 2005). According to Gómez et al. (2004), C. deitersi is a species inquirenda, because Richard's (1897) original description is based on highly conservative features that are not useful for species separation. Future study of specimens from all of these localities is required to show if the records refer to C. ecuadorianus or to C. deitersi. Both species are in need of redescription, and C. ecuadorianus is considered to be a species inquirenda.

Cletocamptus schmidti Mielke, 2000
Distribution. Collected from lagoons of the islands of Santa Cruz, Galapagos. Type locality, Laguna de Puerto Núñez.
Remarks. According to Mielke (2000), C. axi and C. schmidti slightly differ from each other in their body ornamentation and in the chaetotaxy of the exopodites of P3 and P4. Although Mielke (2000) have provided a very detailed illustration of both C. axi and C. schmidti and a complete description of their anatomical details, Wells (2007) considered C. axi a species inquirenda but without giving an explanation. Both species fit well the range of variability of C. deitersi, yet co-occurrence of the two morphotypes and the lack of intermediate forms support that C. axi and C. schmidti are separate species rather than morphological variants of a single species (Mielke 2000). Löffler, 1963 Distribution. Andes (Löffler 1963).

Elaphoidella humboldti
Remarks. Originally described from Ecuador. The male reaches 620 μm and the female is unknown.
According to Gaviria andDefaye (2015: 1026) the "Diversity of Elaphoidella Chappuis, 1929 in Colombia (5 species) is lower than in Cuba (10) and Brazil (9), but higher than in Suriname (2) and Argentina (2). However, these data are not the result of extensive research and sampling of all biomes and environments. Thus, we cannot draw any biogeographical pattern from this study. Only one species is known from each of the following Neotropical countries: Mexico, Costa Rica, Venezuela, French Guiana, Ecuador, Peru and Paraguay. The French islands Bonaire and Martinique are also inhabited by one species each." Groundwater, benthic habitats of high Andean lakes, and aquatic habitats within rainforests are potential habitats for harpacticoid copepods and particularly for Elaphoidella Chappuis, 1929. Other still poorly investigated biotopes are phytotelmata and semiterrestrial habitats, which would no doubt yield new species of copepods (Gaviria and Defaye 2015).
Remarks. Löffler (1963) noted that all Acanthocyclops Kiefer, 1927 specimens from Ecuador possessed a spine formula of the "vernalis type" (2.3.3.3). However, two, three, three, and three spines on the terminal exopodal segments on P1 to P4, respectively, may occur in both A. robustus and A. vernalis Fischer, 1853, which are currently considered distinct from one another.
Acanthocyclops robustus is supposedly restricted to the northern Holarctic region (Mirabdullayev and Defaye 2002). All records of A. robustus from the southern hemisphere need verification, although introduction outside the native range by human activities cannot be excluded. The morphology and taxonomic relationships of A. robustus have been revised by Defaye (2002, 2004), but see Miracle et al. (2013) for an alternative opinion on the taxonomy of the A. robustus group.
The genus is most diversified in the northern temperate region. So far reported from South America, there are only two species, here not including the southern South American A. michaelseni (Mrázek, 1901) and A. skottsbergi Lindberg, 1949 Reid & Suárez-Morales, 1998); the geographic distribution is still poorly understood of these species . The occurrence of Acanthocyclops in South America may raise intriguing questions of the taxonomic identity and evolutionary origin of these taxa. (Fischer, 1853) Distribution. Coastal (Quimi 2014).

Acanthocyclops vernalis
Remarks. This species, which was originally described from the neighborhood of St Petersburg, Russia (Fischer 1853), needs redescription. The actual distributional area is likely confined to the Palearctic region (Einsle 1996), and all South American records need verification (see also A. robustus). (Kiefer, 1926) Distribution. Coastal and Amazon (Napo river valley) (Löffler, 1963).

Mesocyclops meridianus
Remarks. The species range is likely restricted to South America. Mesocyclops meridianus (Kiefer, 1926), which was described from San Bernardino, Paraguay, is morphologically highly similar to M. pseudomeridianus Defaye & Dussart, 1988   The genus is distributed worldwide and is represented by 13 native species in South America, most of which (10 of 13) are endemic to the continent. This number included M. aspericornis (Daday 1906) but excludes M. ogunnus Onabamiro, 1957, which is a supposedly recently introduced species. More extensive sampling will likely reveal more species in Ecuador. Kiefer, 1927 Distribution. Metacyclops are the dominant cyclopoid taxa in glacial lakes in the tropical Andes (Van Colen et al. 2017). (Kiefer, 1927) Distribution. Glacial lakes, 3800-4000 m a.s.l. in Páramo de Guamaní, Andes (Torres and Rylander 2006).

Metacyclops leptopus leptopus
Remarks. Metacyclops leptopus leptopus was originally described from Lake Huarón and Lake Naticocha in Region Pasco in the High Andes of Peru (Kiefer 1926(Kiefer , 1927 Kiefer, 1956 (Mariposa Reservoir, Caracas, Venezuela, ca 985 m a.s.l.). The latter subspecies was considered by Dussart (1984) and Dussart and Defaye (2006) to represent M. mendocinus rather than a lineage within M. leptopus. For more comments on the taxonomic relationships of M. leptopus, see M. mendocinus.

Metacyclops mendocinus (Wierzejski, 1892)
Distribution. Löffler (1963) reported this species from numerous sites in the Andes, and Steinitz Kannan (1979) found it in Lake Cuicocha, Chicapan (= San Pablo), and Yaguarcocha. Peck (1994) reported it in the in the Galapagos from temporary pools on Isla Santa Cruz.
Originally described from northern and western Argentina (Jujuy and Mendoza Provinces) (Wierzejski 1892), this species is widely distributed in both South America (Bolivia, Brazil, Chile, Colombia, Paraguay, Peru, Uruguay, and Venezuela), and Middle America (Cuba, Haiti, Nicaragua, and Puerto Rico) (Dussart and Defaye 2006).
Remarks. The remote mid-Atlantic islands of the Azores harbour a subspecies, M. mendocinus insularis Defaye & Dussart, 1991, which suggests that this species has good capacity for dispersal. Metacyclops mendocinus, along with M. leptopus, belongs to a group of species that are predominantly Neotropical in distribution. They share the 12-segmented state of the antennule and two terminal spines on the terminal endopodal segment of P4 (Herbst 1988). The relationships of M. mendocinus to the M. leptopus-complex need to be revised. Reid et al. (1990) proposed the use of the relative length of the inner terminal caudal (longest) seta as the main distinguishing character between M. mendocinus (seta less than twice as long as caudal ramus) and members of the M. leptopus-complex (seta 2.6 or more times longer than caudal ramus). Reid et al. (1990) also mentioned ecological differences between the two species: M. mendocinus appears to be eurytopic, while M. leptopus apparently inhabits relatively pristine lakes at mostly high altitudes. Accordingly, the records from Andean Ecuador might refer to M. leptopus rather than M. mendocinus (Reid et al. 1990).

Microcyclops alius
Remarks. This species was originally described from Santa Lucia, Southern Uruguay. Rocha (1998) supposed that M. alius is a junior synonym of M. dubitabilis (Kiefer, 1934) (type locality: Trou Caiman Lake, near Port au Prince, Haiti). In a revision of the American Microcyclops, Gutiérrez-Aguirre and Cervantes-Martínez (2016) confirmed the conspecificity of these taxa, and for a redescription of M. dubitabilis (Kiefer 1934 The genus, which has approximately 54 species or subspecies, is distributed worldwide, yet most diversified in the tropics, where there are 42 species. South America harbours about 12 species, and we expect more taxa occur in Ecuador.

Microcyclops anceps (Richard, 1897)
Distribution. Amazon (Napo river valley) (Löffler 1963). (Richard 1897). Two subspecies are distinguished. The range of the nominotypical subspecies extends from Mexico throughout Central and South America as far as Chubut Province, Argentina (Menu-Marque 2001; Dussart and Defaye 2006). Microcyclops anceps pauxensis Herbst, 1962 is known from its type locality at Lago Pauxís in the Brazilian Amazon. Reid (1985) synonymized the form M. anceps var. minor (Dussart 1984) from the Unaré river valley, northern Venezuela with the Amazonian M. anceps pauxensis. Given the current knowledge of the morphology of the American Microcyclops, and M. anceps s. s. in particular (see Gutiérrez-Aguirre and Cervantes-Martínez 2016), the taxonomic position of M. anceps pauxensis and the Venezuelan form need to be revised, as they may represent distinct species rather than subspecies of M. anceps.
Remarks. Eucyclops agilis (Cyclops agilis in original combination), which has as its type locality Regensburg, Germany, is a nomen dubium, and its use should be avoided (Alekseev et al. 2006). In the past, the name E. agilis was often applied to E. serrulatuslike copepods, and in the Americas some of these records might refer to E. pectinifer (Cragin, 1883) (Dussart and Defaye 2006). The identity of Eucyclops Claus, 1893 from the Galapagos and Lake Cunro in the Andes need verification. Löffler, 1963 Distribution. Andes (Löffler 1963).

Eucyclops breviramatus
Remarks. The terra typica of this species is the Lake Papallacta region in the Ecuadorian Andes (3920 m a.s.l.). The general distribution of this species is poorly understood. Records from Mexico are instead another species (Mercado-Salas et al. 2016).

Eucyclops serrulatus (Fischer, 1851)
Distribution. Löffler (1963) reported this species from numerous sites in the Ecuadorian Andes, and Steinitz Kannan (1979) identified a cyclopoid as probably this species from Lake Yambo, Cotopaxi Province. However, these records likely refer to other species; in fact, all records of E. serrulatus from the Americas need verification. Alekseev et al. (2006) revised the taxonomy of this species based on classic morphological charac-ters and integumental pore pattern. In a geographically large-scale overview of the E. serrulatus-complex, Alekseev and Defaye (2011) found E. serrulatus s. s. to be restricted to the Palearctic region. Mercado-Salas et al. (2016), in revising the Mexican fauna, failed to find E. serrulatus, which provides further support that the native range of this species does not include the New World. Mercado-Salas et al. (2016) demonstrated the diagnostic value of several previously overlooked morphological structures (i.e., the surface ornamentation of P4 and antennal coxobasis) in the American Eucyclops.
Remarks. Macrocyclops albidus s. s. is considered to be cosmopolitan (but see Karanovic and Krajicek 2012) and have been reported from several countries in South America, including Argentina, Chile, Colombia, Ecuador, possibly Paraguay, Uruguay, and Venezuela (Löffler 1981;Reid 1985;Rocha and Botelho 1998;Dussart and Defaye 2006;Gaviria and Aranguren 2007). The other South American subspecies, M. albidus principalis Herbst, 1962, differs from the nominal subspecies, among others, in the full development of the inner distal seta on the terminal endopodal segment of P4 (seta reduced to short element in M. albidus s. s.). Macrocyclops albidus principalis is endemic to the Brazilian Amazon, Venezuela, and Colombia (Herbst 1962;Dussart and Defaye 2006;Gaviria and Aranguren 2007) and might perhaps represent a distinct species rather than subspecies. Löffler (1963) reported M. albidus from Ecuador without reference to a subspecific name. He noted that M. albidus did not occur in the High Andes.
Remarks. This is one of the few truly cosmopolitan species in the Cyclopidae (Karaytug 1999). Paracyclops chiltoni also occurs in remote islands, such as New Zealand (terra typica), the Azores in the Atlantic, Crozet and Amsterdam islands in the southern Indian Ocean, and Tahiti and Easter Island in the Pacific (Lindberg 1958;Karaytug and Boxshall 1998b). This suggests that this species could also occur in the Galapagos Islands.
Remarks. The valid name is P. hardingi for the Paracyclops originally described by Lindberg (1958) from Peru and also reported by Löffler (1963) from Ecuador. Paracyclops fimbriatus andinus Lindberg, 1957 is a junior homonym of P. andinus Kiefer, 1957, and P. hardingi was proposed as a replacement name (Karaytug and Boxshall 1998a). Outside of Ecuador, P. hardingi is known from several localities in the High Andes of Peru: Lake Conococha (Ancash) and Lake Huampucocha (Junín) (the type localities of Paracyclops fimbriatus andinus), as well as from various water bodies near Lake Titicaca (Karaytug and Boxshall 1998a).
Remarks. Ergasilidae are parasitic copepods, parasitizing mainly freshwater sometimes marine coastal fish. The overwhelming majority of the South American ergasilid species (69 of 75) are known from Brazil (Marques et al. 2017).
Remarks. According to Löffler (1963), in spite of the numerous collections in Peru, this species has so far been found around the Titicaca Lake in the south of the country. Conversely, it is quite common and widespread in Chile and Argentina, especially in southern areas. Gaviria (1989) found this species in the Cordillera Oriental of the Colombian high Andes. The Ecuadorian population is variable in the segmentation and setation of the female P5 endopodite, which has an asymmetrically distally fused segment in some specimens, and both endopodites have one seta less. In some cases, both P5 endopodites are 2-segmented. Additionally, the left and right endopodite of the male is also variable in length, shape, and segmentation. Such pronounced variability is worthy of further study. Marsh, 1906 Distribution. Andes (Löffler 1963, Brehm 1924Delachaux 1928, as Pseudoboeckella godeti;Dussart 1984;Bayly 1992;Gaviria 1989;Dussart and Defaye 2002;Van Colen et al. 2017).

Boeckella occidentalis
Remarks. According to Löffler 1963, the Ecuadorian populations of this species, which is abundant in the Peruvian Andean regions, differ slightly from the type as de-scribed by Marsh (1906). The enp-3 of the female P5 bears six setae. In the male, the P5 is very similar to the drawings provided by Delachaux (1928), Harding (1955), andLöffler (1955). In both the Peruvian and Ecuadorian populations the distal portion of the right P5 endopodite is recurved, scythe-shaped, and with tuberculi along its outer margin (each tuberculum with a small seta). Gaviria (1989) found this species in the Cordillera Oriental of the Colombian high Andes. Torres and Rylander (2006) and Araujo et al. (2014) mentioned the subspecies Boeckella occidentalis intermedia, yet neither WoRMS (2019) nor Dussart and Defaye (2002) include this taxon. This is probably not a valid name.
Remarks. Löffler (1963) described the subspecies N. amazonicus occidentalis based on a few mature males collected in the Napo river valley of northeastern Ecuador. No mature females were available to Löffler, so that the morphology of the female is currently unknown.
According to Löffler (1963), this taxon is so closely similar to N. amazonicus (S. Wright, 1935) and N. nordestinus (S. Wright, 1935) that they could be considered as variations of a single polymorphic species. However, this subspecies is currently considered to be a valid taxon (Dussart and Defaye 2002;WoRMS, 2019). According to Dussart (1984) the distribution of N. amazonicus s.l. includes the Andean, Amazonian, Orinoco-Venezuelan, Guyanean, and the Paranean regions.
The morphological characters discussed by Löffler (1963) seem too vague and incomplete to soundly allow the establishment of a subspecies, and a taxonomic revision of N. amazonicus s.l. is desirable. Steinitz-Kannan (1979) reported from Lake El Junco, San Cristobal Island, Galapagos, a calanoid resembling N. amazonicus, which was the most abundant zooplanktonic organism in the lake. However, Steinitz-Kannan did not offer drawings or a detailed description that could establish with certainty the identity of this record. Verification of this record is needed.
Remarks. This species is only known from the type locality, the Mazar reservoir on Paute River, Cañar Province, southern Ecuador. The river is eutrophic, belongs to the Amazon basin, and is 2127 m a.s.l. This species is recorded as the most abundant crustacean in the water column of the reservoir, and, considering its restricted known distribution, it is presumably endemic to the region. Notodiaptomus cannarensis has a mean length of 1.4 mm and is a remarkable species among calanoid copepods for its symmetrical aliform projections, which are laterally inserted on the female genital somite, and the presence of a conspicuous lamella on the exp-2 of the right P5 in males. All information available to the organism comes from its original description in Alonso et al. (2017).

Discussion
The identification of European-like species from other parts of the world was a tendency during the nineteenth century (Boxshall and Defaye 2008), and the same applies to the first half of the twentieth century. In the second half of the twentieth century, revisionary studies based on fine-scale taxonomic resolution have recognized numerous species complexes in place of so-called cosmopolitan species (Boxshall and Defaye 2008). Similar to the Cladocera (Lopez et al. 2018a), some species in this checklist may belong to undescribed species or to groups of species with unresolved taxonomic status in the Neotropics and worldwide. For example, Acanthocyclops robustus, Acanthocyclops vernalis, and Eucyclops serrulatus are considered to be cosmopolitan and obviously distributed in the Americas. Further studies may reveal that these species do not occur in the New World or that they have a much more restricted distribution than what has been reported, as for exemple Eucyclops serrulatus according to Mercado-Salas et al. (2016). We cannot, however, exclude the possibility that tropical high-altitude aquatic habitats could harbor relict populations originating from northern latitudes (Van Damme and Eggermont 2011) or that some records might be human-mediated introductions (Lopez et al. 2018a).
Our knowledge on the free-living freshwater copepod fauna from continental Ecuador and Galapagos Archipelago, in comparison to other countries in tropical South America, is relatively recent and rather limited. Countries with ecosystem diversity similar to that occurring in Ecuador have their biodiversity much better documented. For example, a checklist of the free-living copepods of the continental waters of Colombia (Gaviria and Aranguren 2007) reported 69 species and subspecies (14 Calanoida, 41 Cyclopoida, and 14 Harpacticoida). Having examined only 38 crustacean samples, Dussart (1984) increased the number of copepod species known to Venezuela from 28 to 66. From a single Colombian coastal lagoon Fuentes-Reinés and Suárez-Morales (2015) reported 15 copepod species, 10 of which typically live or can occur in freshwater. In a study along La Plata basin, Perbiche-Neves et al. (2014b) found 32 cyclopoid species.
Ecuador is a region with high species richness and high rates of endemism (Myers et al. 2000). Dussart (1984) provided a list of the South American copepod species and showed their distribution among the nine biogeographic zones of the continent. By being situated in both the Andean and the Amazonian biogeographic zones, Ecuador might be home to a significant part of the copepod fauna of both regions. Hence, the current low species richness of the Ecuadorian copepod fauna is most likely the effect of the scarce sampling effort rather than a real biogeographic pattern. We expect that geographically large-scale collections that take the extraordinary diversity of the habitats and strong altitudinal gradients in Ecuador into account will reveal a biodi-versity at least a magnitude greater than what is currently known for the Copepoda. The recommendations of Lopez et al. (2018a) for more extensive collections of the Cladocera, including specimens suitable for molecular studies, also hold true for the copepods. From among the four lake provinces [Paramo, Andean (under the Paramo, 2000-3500 m a.s.l.), Amazonia, and coastal plains and Andean foothills; Steinitz-Kannan et al. 1983], the lake-poor coastal region might be the greatest challenge to explore, albeit marsh-lakes and ephemeral swamps can harbor rich fauna of copepods (Reid 1987). Special attention must be directed at various altitudinal and latitudinal zones within the country to transitional or cryptic habitats such as littoral zones, temporary pools, mosses, phytotelmata, hyporheic zones, wetlands, cisterns, and other habitats.
A taxonomic and zoogeographic revision of the inland water copepods of Ecuador, using both morphological and genetic information, might allow us to test of some major questions of copepod biogeography and evolution (Table 2). To date, any attempt to infer local as well as broad biodiversity and biogeographic patterns of copepods within Ecuador would be premature due to the scarcity of data, dubious records and unsolved taxonomic problems. A better understanding of the biogeography, biodiversity and phylogenetic relationships of the Ecuadorian fauna, can only be reached if the taxonomic and faunistic data are interpreted within a broad geographic frame. To achieve this goal, we need a network of collaboration, with limnologists and taxonomists from both Ecuador and outside the country. Table 2. Some questions about the biogeography, biodiversity, and evolution of the New World Copepoda that could be answered with extensive taxonomic exploration of the Ecuadorian inland water fauna.

Main topics Quesions Dispersal corridor
Might the American Cordillera act as dispersal corridor between North and South America for temperate-or cold-adapted copepods (e.g., see the Acanthocyclops robustus-vernalis complex)? Biogeographical barrier Are the Andes an insurmountable barrier for the dispersal of lowland/thermophilic copepods (i.e. how does the copepod fauna of the Coastal and Amazonian regions differ from each other)? Comparisons might be made between copepods living in the benthic and in the hyporheic zones of rivers, semiterrestrial and cryptic habitats such as mosses, phytotelmata, forest litter, etc., as well as in temporary collections of water (i.e. ponds, pools and marshes), rather than limnetic copepods, as the coastal region has no natural lakes (Steinitz-Kannan et al. 1983). Patterns of speciation within islands Have inland water copepods undergone an evolutionary radiation similar to those found in the terrestrial organisms (Parent et al. 2008) in the Galapagos archipelago, which apparently has a shortage of the fresh surface water bodies (Steinitz-Kannan et al. 1983;López et al. 2018b)? To date, except for the records of the harpacticoids Cletocamptus axi and Cletocamptus schmidti, the cyclopoids Eucyclops agilis, Microcyclops sp., and Metacyclops mendocinus in Santa Cruz island (Peck 1994), and a calanoid resembling Notodiaptomus amazonicus from Lake El Junco in San Cristobal island (Steinitz-Kannan 1979), we have no information on the freshwater copepods of the archipelago. Dispersal capacity, biodiversity and biogeography How do the diversity and geographic distributional patterns change in Copepoda with different dispersal ability (Cyclopidae are considered to be good dispersers, while Diaptomidae are poor dispersers; Canthocamptidae are good but most of the Parastenocarididae studied so far seem to be very restricted geographically)? Diversity and endemism Are copepods less diverse, but with higher rates of endemism in high altitudinal lakes and rivers?