Insects constitute the most diverse animal group and represent one of the earliest lineages occupying the terrestrial habitat (Grimaldi and Engel 2005). Although they are mostly terrestrial, insects usually dominate inland waterbodies in terms of species number and biomass. Many terrestrial insects are resilient to eventual submersion in water (Chapman 1998), but true aquatic insects inhabit freshwater or marine environments at least for one life stage and have several morphological, physiological, and behavioural adaptations. Aquatic insects are usually abundant and important components of energy flow in freshwater habitats, and thus have been the focus of many ecological studies, as well as, the most frequent macroinvertebrate group used in biomonitoring (Rosenberg and Resh 1993, Merritt and Cummins 1996).

The diversity of aquatic insects is a result of several independent invasions of aquatic habitats by terrestrial lineages (more than 50 separate invasions, according to Dijkstra et al. (2014). Marine insects are relatively rare, most of them living in coastal areas and only very few species in the open sea (Cheng 1976). On the other hand, insects inhabiting freshwater are close to 100,000 described species, in twelve insect orders, five of them primarily aquatic: Ephemeroptera, Odonata, Plecoptera, Megaloptera, and Trichoptera (Dijkstra et al. 2014).

Regional knowledge of the aquatic insect fauna is essential to ecological surveys or biomonitoring. In the Neotropics, diversity and distribution of aquatic insects is still poorly known, with many species to be recorded and many others to be described, especially because of the small number of taxonomists and sparse faunistic studies, even near large urban areas. Currently, there are approximately 7,000 described species of aquatic insects in Brazil (Boeger et al. 2016), representing approximately 8% of known species in the world. Amongst these groups, only Odonata and Psychodomorpha reach numbers higher than 14% of the world’s fauna (Table 1). The number of species of some aquatic taxa recorded from Brazil, e.g., Trichoptera, is far below the average of 8% of the world fauna, very likely due to the many undescribed species waiting to be formally named or recorded from the country. Undersampling is one of the major biases for richness evaluations studies and conservation actions such as recently demonstrated for Brazilian dragonflies (Vianna and De Marco Júnior 2012). Therefore, local inventories should remain a priority, particularly in those poorly sampled areas, such as Northeastern Brazil.

The Caatinga Domain (Coutinho 2006) extends for an area of approximately 800,000 km2 in Northeastern Brazil (Fig. 1), including parts of the Brazilian states Piauí (PI), Ceará (CE), Rio Grande do Norte (RN), Paraíba (PB), Pernambuco (PE), Alagoas (AL), Sergipe (SE), Bahia (BA), and Minas Gerais (MG). It is a mosaic of different forest types, characterized by short trees and shrubs with xerophytic characteristics, influenced by climatic extremes when compared to other Brazilian formations: highest solar radiation, highest annual temperature mean, lowest rates of relative humidity, and lowest precipitation, which is limited to a very short period of the year (Prado 2003). It is estimated that over 28% of the original Caatinga vegetation has been altered by human activities, such as, slash-and-burn agriculture, harvesting of firewood, hunting, and herding (Castelletti et al. 2003). Currently, only less than 2% of its natural vegetation is enclosed under federal protected areas.

Figure 1.  

Map of Brazilian states colored by four major phytogeographical domains: Amazon forest, Cerrado, Atlantic forest, and Caatinga; the last one including Ubajara (blue star) and Sete Cidades (green star) National Parks.

Low diversity and low numbers of endemic species is traditionally assigned to the Caatinga, although recent studies suggest otherwise, highlighting it as an important component of Brazilian biodiversity (Leal et al. 2005). The Caatinga lies in between two of the world’s biodiversity hotspots (Myers et al. 2000, Mittermeier et al. 2011), the Atlantic Forest and the Cerrado, and shares many biotic components with these domains. Current assessments of the Caatinga’s biodiversity is certainly underestimated because few biological inventories have focused this region, e.g., for woody plants, 41% of its area has never been surveyed by scientists and 80% of what has been was surveyed poorly (Tabarelli and Vicente 2004). This scenario is certainly worse when it comes to general insect inventories.

Due to the lack of collecting efforts in the Caatinga and Northeastern Brazil, and consequent scarcity of available voucher material from Ceará and Piauí states in insect collections, any information on insect diversity from inventories in these regions will probably constitute new records from these Brazilian states. An up to date list of Coleoptera (Hydrophilidae), Ephemeroptera, Hemiptera (Gerromorpha and Nepomorpha), Odonata, and Trichoptera recorded from Brazilian states of Ceará (CE) and Piauí (PI) are given in Table 2. This table includes species records previously published based on material collected during this project. Previous to this study, there were no records of Elmidae, Epimetopidae, and Torridincolidae (Coleoptera), Leptophlebiidae (Ephemeroptera), Gelastocoridae, Naucoridae, and Pleidae (Hemiptera), and Ecnomidae, Polycentropodidae, and Xyphocentronidae (Trichoptera) for these Brazilian states. Part of distributional data were initially compiled based on available information at Moreira et al. (2011), Moreira (2016), Salles et al. (2016), and Santos et al. (2016). Curiously, the first record of an aquatic insect from either state was of Nephepeltia phryne from Piauí described by Perty (1834) (Fig. 2) and, amazingly, a second dragonfly species was only recorded to this state after 160 years in 1994.

Figure 2.  

Detail of plate 25 modified from Joseph Anton Maximilian Perty’s work from 1834, with a water color illustration of the holotype of the small skimmer dragonfly Libellula phryne (Fig. 3, Odonata: Libellulidae; currently Nephepeltia phryne) described by him from "Provincia Piauhiensi", Brazil.

In this paper, we provide a preliminary checklist of selected taxa of aquatic insects from Ubajara and Sete Cidades National Parks in Northeastern Brazil (Fig. 1), an undersampled region in terms of biological inventories. Additionally, species richness was compared between the two National parks. As far as we know, this is the first published aquatic insect inventory of the selected National Parks.

Identification of aquatic insects depended on the availability of specialists, therefore we focused on Ephemeroptera; suborders Nepomorpha and Gerromorpha of Hemiptera; families Elmidae, Epimetopidae, Hydrophilidae, and Torridincolidae of Coleoptera; Hemerodromiinae (aquatic Empididae) of Diptera; Odonata; Plecoptera; and Trichoptera. These focal taxa were identified at least in genus level, given that in some cases species-level identification was not possible because collected individuals were not adult males (immatures, subimagoes, or adult females), and, in other cases, because of the lack of comprehensive taxonomic revisions for particular genera. Higher-level classification followed the "Catálogo Taxonômico da Fauna do Brasil" (Lecci and Froehlich 2016, Monné et al. 2016, Pinto 2016, Rafael and Câmara 2016, Salles and Boldrini 2016, Santos et al. 2016).

Material examined in this paper were obtained during two collecting expeditions to the study areas, from April 18^th to 21^st of 2012 and February 7^th to 13^th of 2013 at Parque Nacional de Sete Cidades (PNSC, Piracuruca municipality, PI) and from Abril 21^st to 25^th of 2012 and February 13^th to 20^th of 2013 at Parque Nacional de Ubajara (PNU, Ubajara municipality, CE). Although both parks are included in the Caatinga Domain (Fig. 1), areas sampled within these parks seem to represent islands dominated by Cerrado (PNSC) or montane humid forest (PNU). The two parks are distant by approximately 110 km.

Sampling localities within these parks were in or very close (less than 100 m) to bodies of water and are listed in Table 3. Bodies of water ranged from lentic and lotic systems and some streams sampled during the 2012 expedition to PNSC were completely dry in 2013 (see Figs 3, 4).

Figure 3.  

Cachoeira do Riachão (PNSC-02) at Parque Nacional de Sete Cidades, PI, Brazil. (A) April 2012. (B) February 2013.

Figure 4.  

Riacho da Piedade (PNSC-04) at Parque Nacional de Sete Cidades, PI, Brazil. (A) April 2012. (B) February 2013.

The following insects traps were used: 6-meter intercept Malaise traps (Figs 5, 11, Gressitt and Gressitt 1962), suspended intercept traps (Fig. 14, Rafael and Gorayeb 1982), light traps using a white sheet (Fig. 15) and Pennsylvania traps (Fig. 12, Frost 1957), yellow pan traps (Fig. 16), in addition to manual collecting using forceps or entomological nets (Fig. 10). Besides individuals collected during both expeditions, material was also collected by Malaise traps that were left in both parks with samples removed every month between expeditions. Material collected was divided and deposited in the following institutions: Coleção Entomológica Prof. José Alfredo Pinheiro Dutra, Departamento de Zoologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro (DZRJ); Coleção Zoológica do Maranhão, Universidade Estadual do Maranhão, Caxias (CZMA); and Coleção de Invertebrados, Instituto Nacional de Pesquisas da Amazônia, Manaus (INPA).

Figure 5.  

6-meter Malaise flight intercept trap over Rio Gameleira crossing Trilha Samambaia (PNU-01) at Parque Nacional de Ubajara, CE, Brazil.

Figure 6.  

Rio das Minas crossing Trilha Araticum (PNU-02) at Parque Nacional de Ubajara, CE, Brazil.

Figure 7.  

Rio Cafundó crossing Trilha Araticum (PNU-04) at Parque Nacional de Ubajara, CE, Brazil.

Figure 8.  

Cachoeira do Cafundó (PNU-07) at Parque Nacional de Ubajara, CE, Brazil.

Figure 9.  

Riacho da Bananeira (PNSC-01) at Parque Nacional de Sete Cidades, PI, Brazil. Photo by A. Somavilla.

Figure 10.  

Manual collecting with entomological net near border of Riacho do Bananeira (PNSC-01), Parque Nacional de Sete Cidades, PI, Brazil.

Figure 11.  

6-meter Malaise flight intercept trap over seasonally dried stream below Cachoeira do Riachão (PNSC-02) at Parque Nacional de Sete Cidades, PI, Brazil.

Figure 12.  

Pensylvania light trap near Olho d'água dos Milagres (PNSC-07) at Parque Nacional de Sete Cidades, PI, Brazil.

Figure 13.  

Poço do Bananeira (PNSC-08) at Parque Nacional de Sete Cidades, PI, Brazil.

Figure 14.  

Suspended intercept trap over stream at Parque Nacional de Ubajara, CE, Brazil.

Figure 15.  

Manual collecting using white-sheet light trap at Parque Nacional de Ubajara, CE, Brazil.

Figure 16.  

Yellow pan trap (YPT) near border of stream at Parque Nacional de Sete Cidades, PI, Brazil.

Species complementarity index for the two National parks were calculated with Cjk = Ujk/Sjk, where Ujk is the number of unique species in both sites and Sjk total richness in both sites combined (Colwell and Coddington 1994).

During the expeditions 27 samples (14 at PNSC and 13 at PNU) of Pennsylvania light traps placed over streams were collected. Each sample was the result of an approximately 6-8 hour effort of 15W fluorescent or UV light turned on at dusk. All individuals of Ephemeroptera, Plecoptera, Trichoptera, and Hydrophilidae (Coleoptera) were identified or morphotyped and counted. Adult females or subimagoes impossible to be identified into species level were treated as conspecific as males identified of the same genus on the sample (6 morphotypes from PNSC and 8 morphotypes from PNU of Ephemeroptera and Trichoptera). However, in cases where more than one species of the same genus was identified from the same sample (Helicopsyche and Oxyethira from PNU and Chimarra, Neotrichia, Smicridea, and Oxyethira from PNSC), the number of female individuals were equally divided among different species.

Quantitative samples of Pennsylvania light traps totalled 227 individuals of 23 species from PNU and 511 individuals of 54 species from PNSC of Ephemeroptera, Plecoptera, Trichoptera and Hydrophilidae (Coleoptera). Quantitative data served for comparisons of species richness between the two National Parks based on a rarefaction curve by individuals calculated in PAST 3.0 (Hammer et al. 2001).

Approximately 7,000 individuals of focal taxa were identified, being 5,472 from PNU and 1,539 from PNSC (Fig. 17).

Figure 17.  

Number of individuals of aquatic insects identified from Ubajara (PNU, blue) and Sete Cidades (PNSC, green) National Parks.

These individuals were identified into 96 species from PNU and 112 species from PNSC, while 20 of these species were found in both parks (Fig. 18). Species complementarity between both parks was of 89%.

Figure 18.  

Percentage of total species of aquatic insects found only at Ubajara (PNU, blue), only at Sete Cidades (PNSC, green), and in both National Parks (red).

The number of species per insect order in PNU and PNSC, respectively were: Coleoptera, 16 and 20; Diptera 10 and 1; Ephemeroptera, 1 and 9; Hemiptera, 9 and 20; Odonata, 21 and 21; Plecoptera, 1 and 1; and Trichoptera, 38 and 40. Species checklists for both National Parks are given below. Species marked with an asterisk (*) were found in both National Parks and distributional records (country or state) marked with and exclamation mark (!) are new records published herein. Species described or first taxa records based on material collected during this project are cited under Notes.