The Ethiopian caddisfly fauna comprises 85 species, including 10 Oecetis species and three Lepidostoma species. In this context we provide the first species-level descriptions of Ethiopian caddisfly larvae. We describe and illustrate the larvae of O. mizrain and L. scotti, with additional notes on their habitats and distribution.

caddisfly larvae, distribution, ecology, Afrotropical Region, diversity, ecological management

Caddisflies are one of the most diverse aquatic insect groups, distributed all over the world. The order Trichoptera comprises approximately 15,000 species (including 685 fossils) in 616 genera and 49 families (Morse 2017). The Oriental region harbors the highest number of species (4865 species), followed by the Neotropical region (2562 species) (de Moor and Ivanov 2008; Holzenthal et al. 2015; Morse 2011). Species numbers in other biogeographic regions such as the Australasian (1439 species), the East Palearctic (1372 species), the Nearctic (1604 species) and the West Palearctic (1888 species) are lower in comparison (Morse 2011). The poorly explored Afrotropical region is currently represented by about 1200 species only, belonging to 28 families and 111 genera (Tobias and Tobias 2008). The Antarctic region is the only biogeographic area where no caddisflies have been recorded (Holzenthal et al. 2015; Morse 2011).

As in many Afrotropical countries, the Ethiopian caddisfly fauna is poorly studied. In a first effort to characterize the almost unknown African caddisfly fauna, Ulmer (1930) described and listed eight Ethiopian Trichoptera species. Three decades later, Kimmins (1963) reported 51 species, of which 17 were described as new. At present, the number has marginally increased to 85 species of 26 genera and nine families (Tobias and Tobias 2008; Morse 2017). Two families, Leptoceridae (35 species in 10 genera) and Hydropsychidae (22 species in six genera), represent over 60% of the currently known Ethiopian Trichoptera diversity.

Caddisflies of the widely distributed genera Lepidostoma Rambur, 1842 and Oecetis Mclachlan, 1877 comprise 453 and 539 extant species worldwide, respectively (Morse 2017). From Africa and its adjacent islands, 51 Lepidostoma and 103 Oecetis species are known (Tobias and Tobias 2008). The Checklist of East Africa (Johanson 1992) includes two Lepidostoma and 17 Oecetis species. Until now, three Lepidostoma and 10 Oecetis species have been recorded in Ethiopia (Table 1).

Within the genus Lepidostoma, the first species described from Ethiopia was L. scotti (Ulmer, 1930). Ulmer (1930) described the species based on the material obtained by Hugh Scott and Omur-Cooper during their expedition to the central highlands of Ethiopia. Since then, only two additional species have been described from Ethiopia, L. missa and L. zepho (Malicky & Graf, 2012). In the genus Oecetis, the species described first from Ethiopia was O. montana (Ulmer, 1930), based on specimens collected in the central highlands. Later, Kimmins (1963) described O. tjonnelandi, O. ghibensis, O. brevis and indicated the presence of O. brunnescens (Ulmer, 1923), O. montana (Ulmer, 1930), O. pangana Navás, 1930 and O. setifera (Ulmer, 1922) in Ethiopia. Most recently, Malicky and Graf (2012, 2015) described O. armaros Malicky & Graf, 2015, O. mizrain Malicky & Graf, 2012 and reported O. portalensis Mosely, 1939 from rivers in the Ethiopian highlands.

Caddisflies are frequently used along with other aquatic fauna as bioindicators in ecological assessment systems as they are sensitive to organic pollution and stream degradation (Barbour et al. 1999). In particular autecological characterization provides vital information for freshwater bioassessment by relating species to ecological conditions (Jones 2008). In this context, species-level identification of bioindicators is of great importance to fully reap the power of ecological analysis (Cranston 1990; Lenat and Resh 2001; Malicky et al. 2008). Species-level determination is achieved for Trichoptera and many other taxonomic groups in Central and Northern European countries where this information is used for stream quality assessments (Hering et al. 2003; Schmidt-Kloiber et al. 2006). Most studies in tropical Africa including Ethiopia, however, are restricted to family-level identification due to the lack of taxonomic knowledge (Aschalew and Moog 2015). Therefore, to support efforts to maintain the general ecological health of fresh waters and thus ensure sustainable use of water resources in Ethiopia the compilation of taxonomic and autecological databases for all potential bioindicators is essential. Hence, in this contribution, we describe and illustrate the final instar larvae of L. scotti and O. mizrain from Ethiopia.

Larval and adult material was collected at small highland streams (9°4'N, 38°8'E) within Chilimo Forest, in the upper catchment of the Awash River. Association of larvae and adults was enabled by the exclusive occurrence of these Lepidostoma and Oecetis species at this site and the presence of mature pupae in the case of O. mizrain. Chilimo Forest is a dry afromontane forest, located about 97 km west of Addis Ababa and 7 km north of Ginchi town. It covers an area of more than 2500 hectares within an altitudinal range from 2170 m a.s.l. to 3054 m a.s.l. (Teshome and Ensermu 2013). The forest and its surrounding areas receive little precipitation from March to April, while precipitation is highest from June to September with a mean annual rainfall of >1000 mm (Aschalew 2015). The dominant trees species in this forest are Juniperus procera, Podocarpus falcatus, Prunus africana, and Olea europaea (Kassa et al. 2009; Teshome and Ensermu 2013).

The collected larval specimens were preserved in 70% ethyl alcohol. Morphological characteristics of specimens were examined and photographed using a Zeiss StereoLumar V.12 equipped with an AxioCamErc5s camera and the Zeiss-native image processing software ZEN. Image series at different focus levels were obtained and stacked via CombineZP (Hadley 2008; Brecko et al. 2014) to create single extended-depth-of-focus images. Larval morphological features and nomenclature of primary setae and setal areas follows Wiggins (1996) and Waringer and Graf (2011).

Historically, the most comprehensive faunistic African studies were done in South Africa, Madagascar and West Africa. These studies yielded hundreds of Trichoptera species to the Afrotropical region, of which 253 species are known from South Africa (de Moor and Day 2013), more than 500 species from Madagascar (Benstead et al. 2003), and 343 species were reported from the West Africa region (Tobias and Tobias 2008). However, other regions of the African continent are not as well represented in the available caddisfly literature. In Ethiopia, 85 species are known currently, and the discovery of several new species within the next decades is likely. Particularly, on-going projects on water resource management and capacity building will foster faunistic surveys and thus provide highly relevant baseline data for both taxonomy and ecosystem monitoring.

However, most of the data available on diversity of the African Trichoptera fauna were compiled based on adult specimens, and most species are not known in the larval stage. According to Scott (1986) only 105 of 747 Sub-Saharan species of African Trichoptera were known in the larval stage at that time. Thus, development of more precise biomonitoring tools is severely hampered. Moreover, description of these larval stages could foster development of taxonomic expertise as crucially needed for biomonitoring and sustainable development of freshwater resources in the Afrotropical region. Consequently, faunistic surveys and taxonomic treatment of yet undescribed larval stages are of great value for the development of knowledge and human resources as well as for the documentation of freshwater biodiversity.

In general, larvae of the genus Lepidostoma inhabit springs and cool streams that usually have slow water flow and substantial input of allochthonous organic material (Wiggins 2004). Indeed, it is evident from the mandibles of L. scotti larvae (Fig. 13) that this species is, like most Lepidostoma, a detritivorous shredder that feeds on leaf litter and decaying woody debris. Moreover, larvae of Lepidostoma are known to exploit even recalcitrant resources, including the leaves of coniferous plants (Grafius and Anderson 1980; Wiggins 2004). The riparian zone of Chilimo Forest streams is usually dominated by the coniferous plants Juniperus procera and Podocarpus falcatus (Kassa et al. 2009; Teshome and Ensermu 2013). The leaves of these species most likely represent the primary food source of L. scotti larvae in these habitats. Other studies in Ethiopia indicated that Lepidostoma larvae exclusively live in forested streams instead of areas with intensified land use (Aschalew 2015; Ferengi 2016).

Larvae of the genus Oecetis are found in a wide range of freshwater habitats and are either carnivorous, or behave as detritivores or shredders (e.g., Waringer and Graf 2011). Larvae of carnivorous Oecetis species have elongated single-blade mandibles (Waringer and Graf 2011). Based on the elongated bladelike mandibles observed in O. mizrain, larvae of this species (Fig. 2) most likely have a predatory feeding behavior.

In addition to information on feeding ecology, stream zonation preference, sensitivity to organic pollution, or sediment load, knowledge on the flight periodicity of potential bioindicators is crucial to determine sampling seasons for biomonitoring approaches. According to Wright et al. (2013), caddisfly flight periodicity is likely controlled by a combination of innate behavior and environmental factors, most notably temperature. However, despite repeated sampling efforts in Ethiopia, the flight periods of single species could not yet be defined. Recent collections of adult L. scotti and O. mizrain revealed that these species were active in October and November (Malicky and Graf 2012, 2015), but their complete flight period remains unclear due to the lack of consistent faunistic surveys. In the eastern Afrotropical Region, these months are considered as the regeneration period for most macroinvertebrates following the heavy rainfall and high flooding season that extends from late June to mid-September (Aschalew 2015). However, flight periodicity of any Ethiopian caddisfly taxa has yet to be studied throughout the year, and further investigations on the annual flight period are currently under way.

Caddisfly larvae are widely used as indicator taxa in freshwater assessments as they exploit a wide range of ecological niches, often are found in abundance and cover a wide sensitivity range (Aschalew and Moog 2015; Barbour et al. 1999; Chakona et al. 2009; Hering et al. 2003). Ideally, assessment of the ecological condition of aquatic ecosystems is based on the identification of macroinvertebrates to species-level (Sharma et al. 2008). Jones (2007) also emphasizes the importance of species datasets for a better interpretation of bioassessment results, as well as testing ecological theories and evaluating threats of extinction to aquatic taxa. In particular, lack of species-level information leads to underestimating the actual differences in community structure (Benstead et al. 2003). Therefore, species-level resolution is fundamental for freshwater biomonitoring (Lenat and Resh 2001; Malicky et al. 2008; Corbi and Trivinho-Strixino 2006).

To achieve a better understanding of Ethiopian freshwater biodiversity and the biogeography of African freshwater fauna in general, description of species and preparation of species-level keys is imperative. Here we provide some data that might be useful for future studies to characterize the Ethiopian caddisfly fauna.

We are most grateful to Hans Malicky for confirming identification results. Fieldwork was conducted under the auspices of the LARIMA – Sustainable HighLAnd RIvers MAnagement in Ethiopia – project (Project Number 106) funded by the Austrian Partnership Programme in Higher Education and Research for Development (APPEAR) of the Austrian Development Cooperation (ADC) and the Austrian Agency for International Cooperation in Education and Research (OeAD).