Research Article |
Corresponding author: Anna E. Hjalmarsson ( annahjalmar@gmail.com ) Academic editor: Ralph Holzenthal
© 2018 Anna E. Hjalmarsson, Wolfram Graf, Sonja C. Jähnig, Simon Vitecek, Steffen U. Pauls.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Hjalmarsson AE, Graf W, Jähnig SC, Vitecek S, Pauls SU (2018) Molecular association and morphological characterisation of Himalopsyche larval types (Trichoptera, Rhyacophilidae). ZooKeys 773: 79-108. https://doi.org/10.3897/zookeys.773.24319
|
Himalopsyche Banks, 1940 (Trichoptera, Rhyacophilidae) is a genus of caddisflies inhabiting mountain and alpine environments in Central and East Asia and the Nearctic. Of 53 known species, only five species have been described previously in the aquatic larval stage. We perform life stage association using three strategies (GMYC, PTP, and reciprocal monophyly) based on fragments of two molecular markers: the nuclear CAD, and the mitochondrial COI gene. A total of 525 individuals from across the range of Himalopsyche (Himalayas, Hengduan Shan, Tian Shan, South East Asia, Japan, and western North America) was analysed and 32 operational taxonomic units (OTUs) in our dataset delimited. Four distinct larval types of Himalopsyche are uncovered, and these are defined as the phryganea type, japonica type, tibetana type, and gigantea type and a comparative morphological characterisation of the larval types is presented. The larval types differ in a number of traits, most prominently in their gill configuration, as well as in other features such as setal configuration of the pronotum and presence/absence of accessory hooks of the anal prolegs.
Caddisfly, GMYC, Hengduan Mountains, Himalaya, life stage association, PTP
With ~15,000 described and around 50,000 presumed species, caddisflies are one of the larger insect orders and the largest primary aquatic insect order (
Species of the genus Himalopsyche are a particularly interesting group of caddisflies. They primarily inhabit mountain and alpine environments in Central and East Asia, although the genus also radiated into the Nearctic where it is represented by a single species, H. phryganea (Ross, 1941). Himalopsyche larvae mostly inhabit highly turbulent, fast-flowing streams, where they live as ferocious predators.
Including recent species descriptions, the genus Himalopsyche currently comprises 53 known species, adding to the status of the last major treatment of Himalopsyche by
Caddisflies are good biological indicators (e.g.,
Molecular data have proven successful in facilitating the association of larvae with adults in Trichoptera (e.g.,
In this study, we associate larvae with adults based on molecular data from a nuclear and a mitochondrial gene, CAD and COI, respectively, and employ three different methods for life stage association to generate a consensus result. We then present the first comparative morphological characterisation of all known larval types of Himalopsyche.
Larval specimens used in this study were mostly collected in Nepal and China in 2011–2013; adult material was largely obtained from research collections (Suppl. material
For this study, we used partial sequence data of the single copy nuclear marker CAD and the mitochondrial COI. CAD has proved useful for insect phylogenetics (
DNA was extracted from legs using one of the following methods: HotShot protocol (
Primers used for PCR and sequencing. Fragment lengths refer to the primer pairs 743nF-ino & 1028r-ino, C1Fb & C7Ra, HCO1490 & LCO2198, and B1Fa & B3Ra.
Gene | Primer | Sequence | Tm (°C) | Fragment length (bp) | Reference |
---|---|---|---|---|---|
CAD | 743nF-ino | 5’-GGIGTIACIACIGCITGYTTYGARCC-3’ | 52.4 | 850 |
|
CAD | 1028r-ino | 5’–TTRTTIGGIARYTGICCICCCAT–3’ | 42.1 | 850 |
|
CADinternal | C1Fb | 5’–TGYGTTGTRAAGATTCCGAG-3’ | 51.8 | 736 | this work |
CADinternal | C7Ra | 5’–TGTCCATTACAACCTCGAATG-3’ | 62.3 | 736 | this work |
COI | HCO1490 | 5’-GGTCAACAAATCATAAAGATATTGG-3’ | 53.2 | 658 |
|
COI | LCO2198 | 5’-TAAACTTCAGGGTGACCAAAAAATCA-3’ | 51.0 | 658 |
|
COI internal | B1Fa | 5’-ATTGCDACWGATCAWACAAA-3’ | 54.9 | 367 | this work |
COI internal | B3Ra | 5’-AAYGTARTWGTWACWGCTCA-3’ | 47.2 | 367 | this work |
Protocols for 10 μL PCR reactions, using VWR peqGOLD Hot Taq DNA Polymerase kits. BSA = Bovine serum albumin. All numbers are given in μL.
Gene | Primer pair | Buffer S | Buffer Y | dNTPS 2mM | BSA 20 mg/ml | Forward primer 10μM | Reverse primer 10μM | Taq | DNA | H2O | Cycler |
---|---|---|---|---|---|---|---|---|---|---|---|
CAD | 1028r-ino, 743nF-ino | 1 | 0 | 1 | 0 | 0.25 | 0.25 | 0.1 | 1 | 6.4 | 5’ 95°C, 35x (45’’ 95°C, 45” 55°C, 60” 72°C) 5’ 72°C |
CAD | C1Fb, C7Ra | 1 | 0 | 1 | 0 | 0.25 | 0.25 | 0.1 | 1 | 6.4 | 5’ 95°C, 35x (30’’ 95°C, 30” 50°C, 45” 72°C) 5’ 72°C |
COI | HCO1490, LCO2198 | 1 | 0 | 1 | 0 | 0.25 | 0.25 | 0.1 | 1 | 6.4 | 5’ 95°C, 5 x (30’’ 95°C, 1’ 44°C, 1’ 72°C), 15x (30’’ 95°C, 30’’ 48°C, 1’ 72°C), 20 x (30’’ 95°C, 30’’ 50°C, 1’ + (10’’ * n) 72°C), 5’ 72°C |
COI | B1Fa, B3Ra | 1 | 0 | 1 | 0.4 | 0.25 | 0.25 | 0.1 | 1 | 6.0 | 5’ 95°C, 35x (30’’ 95°C, 30” 45°C, 45” 72°C) 5’ 72°C |
Adult males were identified to species based on morphology. The dataset included 38 adult species based on males, including four putative new species (Hjalmarsson submitted, Kuranishi et al. unpublished data). We used and compared the results of three phylogenetic association criteria: Poisson Tree Process (PTP;
Gene trees of all available specimens were reconstructed in MrBayes v3.2.6 (
Specifications of alignments used for gene tree reconstruction with BEAST and MrBayes.
Alignment | Number of sequences | Length (bp) | Variable sites | Parsimony informative sites | Missing data | Analysis | Substitution model |
---|---|---|---|---|---|---|---|
CAD | 353 | 736 | 37,6% | 28,0% | 1.5% | MrBayes | 1: GTR+I2: F813: HKY+G |
CAD haplotypes | 136 | 736 | 31.8% | 26.4% | 1.6% | BEAST | 1: BMod2: BMod3: BMod |
COI | 451 | 658 | 40,0% | 37,7% | 18.5% | MrBayes | 1: SYM+I+G2: F81+I3: GTR+I+G |
COI haplotypes | 183 | 658 | 39.2% | 35.0% | 15.9% | BEAST | 1: BMod2: BMod3: BMod |
The GMYC method uses a haplotype-based ultrametric gene tree to determine the transition from inter- to intraspecific branching patterns (
Haplotype-based chronograms require that identical sequences be removed from the alignment, leaving an alignment consisting only of unique sequences. Identical haplotypes were removed from the original alignments using collapsetypes_v4.6 (
For this paper, we define the use of the terms ‘species’, ‘putative new species’, ‘cluster’ and ‘OTU’ (operational taxonomic unit) as follows: ‘Species’ refers to formally described morphological taxa, following established taxonomy. With ‘putative new species’ we mean morphologically distinct taxa that are still unknown in the literature. The term ‘cluster’ refers to specific results from one of the two analyses, outputting delimited GMYC and PTP ‘clusters’, respectively. For our consensus result from morphology, PTP, and GMYC based on CAD and COI we use the term OTU, which can represent single species or groups of species referred to as species complexes.
Comparative morphological analysis of larvae followed a standard procedure. We screened all larvae of each OTU for consistent morphological characters. Instar differentiation and thus assignment of most larvae to different instars is not possible with the currently available material. Therefore, general features commonly represented by all OTUs within one phylogenetic clade were considered as synapomorphies. Some characters were found present across all size classes of single OTUs and phylogenetic clades, e.g., distolateral accessory hooks on lateral plates of anal prolegs were consistently present in even the smallest instars.
The dataset comprised 525 Himalopsyche individuals (205 adults, 313 larvae and 8 pupae), and R. polonica as outgroup for MrBayes. We generated 352 Himalopsyche sequences of CAD (736 bp) and 450 Himalopsyche sequences of COI (658 bp). After haplotype reduction of the alignments, the CAD alignment had 136 unique haplotypes and the COI had 183 unique haplotypes. The total CAD alignment had 37.6% variable sites; the total COI haplotype alignment had slightly more with 40% (Table
PTP delimited 29 clusters with CAD, and 62 with COI for the ingroup. GMYC delimited 27 clusters with CAD and 46–48 clusters with COI. The results from separate runs were stable except for GMYC with COI. This instability did not affect larval association and we hereafter only refer to COI run 1 which delimited 48 GMYC clusters (Table
Number of clusters delimited by PTP and GMYC. Number of PTP clusters refer to the ingroup only.
Gene | Method | Run | Chain length | Burn-in | ESS | Clusters |
---|---|---|---|---|---|---|
CAD | MrBayes & PTP | 1 | 1* 107 | 25% | All >200 | 29 |
CAD | MrBayes & PTP | 2 | 1* 107 | 25% | All >200 | 29 |
CAD | BEAST & GMYC | 1 | 1* 109 | 50% | All >200 | 27 |
CAD | BEAST & GMYC | 2 | 5 * 108 | 10% | Most >200 | 27 |
COI | MrBayes & PTP | 1 | 5* 107 | 25% | All >200 | 62 |
COI | MrBayes & PTP | 2 | 1* 107 | 25% | Most >200 | 62 |
COI | BEAST & GMYC | 1 | 1* 109 | 10% | All >200 | 48 |
COI | BEAST & GMYC | 2 | 1 * 109 | 10% | All >200 | 46 |
We could unambiguously associate 239 larvae and eight pupae to the following nine species: H. acharai, H. anomala Banks, 1940, H. digitata (Martynov, 1935), H. gregoryi (Ulmer, 1932), H. phryganea, H. sylvicola, H. tibetana, H. 677, and H. 685 (Table
Overview of life stage association results for each OTU. Asterisks denote. The OTUsmartynovi-complex, H. 683, and H. 685, formed a single GMYC cluster in CAD, here referred to as ‘martynovi-clade’. Key: † indicates larvae without DNA data, * indicates conflicts in COI and CAD gene tree topologies.
OTU | Samples | Species | Larvae | Larval type | CAD units (PTP/GMYC) | COI units (PTP/GMYC) | Association with CAD (PTP/GMYC) | Association with COI (PTP/GMYC) | Reciprocal monophyly (both genes) | Number of criteria fulfilled |
---|---|---|---|---|---|---|---|---|---|---|
H. acharai | 4 | H. acharai | yes | japonica type | 1/1 | 1/1 | yes/yes | yes/yes | yes | 5 |
H. anomala | 6 | H. anomala | yes | tibetana type | 1/1 | 4/3 | yes/yes | no/yes | yes | 4 |
H. auricularis | 2 | H. auricularis | no | unknown | 1/1 | 2/1 | – | – | not applicable | – |
H. biansata | 2 | H. biansata | no | unknown | no data | 1/1 | – | – | not applicable | – |
H. diehli | 1 | H. diehli | no | unknown | 1/1 | 1/1 | – | – | not applicable | – |
H. digitata | 34 | H. digitata | yes | tibetana type | 2/1 | 2/2 | yes/yes | no/no | yes | 3 |
H. eos | 2 | H. eos | no | unknown | no data | 1/1 | – | – | not applicable | – |
excisa-complex | 31 | H. excisa, H. placida, H. maitreya | yes | tibetana type | 1/1 | 5/2 | no/no | no/no | no | 0 |
H. gigantea | 2 | H. gigantea | no | gigantea type | no data | 1/1 | – | – | not applicable | – |
H. gregoryi | 77 | H. gregoryi | yes | tibetana type | 1/1 | 1/1 | yes/yes | yes/yes | yes | 5 |
H. gyamo | 1 | H. gyamo | no | unknown | 1/1 | 1/1 | – | – | not applicable | – |
H. horai | 5 | H. horai | no | unknown | 1/1 | 1/1 | – | – | yes | – |
japonica-complex* | 8 | H. japonica, H. sp. n. 1529 | yes† | japonica type | 2/1 | 4/3 | – | – | no | – |
H. kuldschensis | 2 | H. kuldschensis | no | unknown | 1/1 | 1/1 | – | – | yes | – |
H. lanceolata | 4 | H. lanceolata | no | unknown | 1/1 | 1/1 | – | – | yes | – |
H. lepcha | 5 | H. lepcha | no | unknown | 1/1 | 2/2 | – | – | yes | – |
H. lua | 5 | H. lua | no | unknown | no data | 2/1 | – | – | not applicable | – |
martynovi-complex | 30 | H. martynovi, H. epikur | yes | tibetana type | martynovi–clade | 1/1 | no/no | no/no | no | 0 |
H. navasi | 4 | H. navasi | no | unknown | 1/1 | 4/4 | – | – | not applicable | – |
H. phryganea | 4 | H. phryganea | yes | phryganea type | 1/1 | 2/2 | yes/yes | yes/yes | yes | 5 |
platon-complex | 27 | H. platon, H.. sp (F), H.. sp (L) | yes | tibetana type | 1/1 | 4/3 | no/no | no/no | yes | 1 |
H. sp. 1196 (L) | 17 | NA | yes | tibetana type | 1/1 | 2/1 | – | – | yes | – |
H. sp. 1254 (L) | 4 | NA | yes | tibetana type | 1/1 | no data | – | – | not applicable | – |
H. sp. 1338 (L) | 1 | NA | yes | gigantea type | 1/1 | 1/1 | – | – | not applicable | – |
H. sylvicola | 33 | H. sylvicola | yes | gigantea type | 1/1 | 6/3 | yes/yes | no/yes | yes | 5 |
H. tibetana | 58 | H. tibetana | yes | tibetana type | 1/1 | 2/2 | yes/yes | yes/yes | yes | 5 |
H. todma | 4 | H. todma | no | unknown | 1/1 | no data | – | – | not applicable | – |
triloba-complex* | 67 | H. triloba, H. hageni, H. macilipennis, H. malenanda, H. efiel, H. yatrawalla | yes | gigantea type | 2/2 | 5/3 | no/no | no/no | no | 0 |
H. yongma | 1 | H. yongma | no | unknown | 1/1 | 1/1 | – | – | not applicable | – |
H. 677 | 41 | H. 677 | yes | tibetana type | 1/1 | 1/1 | yes/yes | yes/yes | yes | 5 |
H. 683 | 8 | H. 683 | no | tibetana type | martynovi–clade | 1/1 | – | – | yes | – |
H. 685 | 34 | H. 685 | yes | tibetana type | martynovi–clade | 1/1 | no/no | yes/yes | yes | 3 |
Synapomorphic larval characters of Himalopsyche according to
• Mandibles with prominent lateral protuberances (Figure
• 2nd and 3rd leg with anterodorsal single coxal gills (e.g., Figure
• Abdomen with prominent and complex gills consisting of a multitude of gill filaments positioned on several bases or a single base which can be slightly to distinctly protuberant (e.g. Figure
• Anal proleg with two proximal accessory hooks fused with lateral sclerites (e.g. Figure
Based on these characters, larvae of Himalopsyche can easily be differentiated from the closely related genera Philocrena and Rhyacophila. The larvae of the monotypic genera Fansipangana and Phoupanpsyche are unknown. Within Himalopsyche, the four different larval types can easily be differentiated based on distinct character states.
Himalopsyche phryganea is the only species known with this larval type. Larvae of H. phryganea were described and illustrated by
Thorax. Pronotum with a single row of long dark setae along the entire anterior margin; short light recumbent setae concentrated at anterolateral pronotal edges; Sa1 present as a transversal band of 4–5 setae, Sa2 absent (Figure
Mesolegs and head of Himalopsyche larvae. 11 Mesoleg of H. gregoryi 12 Mesoleg of H. sylvicola, arrows indicate dorsal fringe of setae on legs 13 Mesoleg of H. sylvicola. Arrows indicate pennate setae on coxa and femora 14 Head of H. phryganea, arrow points to lateral protuberances on mandible.
Thorax of Himalopsyche larvae in dorsal and lateral view. 15, 16 H. phryganea 17, 18 H. gregoryi, arrow a points to anterodorsal single coxal gill 19, 20 H. japonica, arrow a indicates anterodorsal single coxal gill; arrow b indicates the ventral gills 21, 22 H. sylvicola. Arrows A points to anterior.
One species with larvae of the tibetana type has been described in the larval stage: H. tibetana (
Thorax. Pronotum with two rows of setae along the anterior edge, anteriormost row of setae short, light, recumbent and posterior row setae longer, black; Sa 1 present as a transversal band of 4–5 setae, Sa2 absent (Figure
Abdomen lateral, ventral. 29, 30 H. phryganea, arrow points to ventral medial sclerite 31, 32 H. gregoryi, arrow points to ventral medial sclerite 33, 34 H. japonica, arrow points to ventral protuberance 35, 36 H. sylvicola, arrow a indicates small rounded protuberance on lateral process, while arrow b indicates the ventral medial sclerite. Roman numbers indicate abdominal segments. Arrow A points to anterior.
Two species of the japonica type have been described in the larval stage: H. acharai by
Thorax. Pronotum with a single row of setae along the anterior edge; Sa1 present as a transversal row of 2–4 dark setae medially, Sa2 present as a group (sometimes arranged as sagittal band) of 2–4 setae (Figure
Anal prolegs, lateral, caudal. 37, 38 H. phryganea 39 H. gregoryi 40 H. tibetana 41, 42 H. japonica 43, 44 H. sylvicola. Key: arrows a distolateral accessory hook. arrows b protuberance/hook on dorsal plate. arrows c dorsal spine on basal anal claw. arrows d proximal accessory hooks fused with lateral sclerites. Arrow A points to anterior.
Larvae of this type have been described by
Thorax. Pronotum with a single row of setae along the anterior edge; Sa1 present as a transversal band of 2–4 setae, Sa2 present as a sagittal band of 7–9 dark setae, prominent (10); legs with dorsal fringe of setae (Figure
We used several life stage association strategies based on two genes in a comparative setting: PTP, GMYC and reciprocal monophyly. All strategies acknowledge a successful association in cases of sequence identity, but differ in that PTP and GMYC use branch lengths to estimate which clades represent distinct units, and that reciprocal monophyly requires congruence of gene trees. The importance of using more than one gene (
We observed clear morphological differences in morphology between the larval types. Within larval types, however, the examined material did not show any stable and reliable morphological characters to delimitate larvae at species-level. More material, particularly of last instar larvae, and ideally from numerous sites is required to better assess interspecific, intraspecific and ontological variation in the here described as well as other morphological characters. Only then will it be possible to assess if the observed morphological variation is useful for delimiting species in the larval stage.
In this study, we present the characteristic morphological differences of four larval types of Himalopsyche. Life-stage association based on molecular data enabled us to do this, as it provided an OTU assignation for over 300 larvae. We found little or no morphological differences among species within the same type. Once we are able to better discern organisms at lower taxonomic rank (by morphology or molecular association, e.g., in high-throughput barcoding studies), aquatic insects and other benthic invertebrates can become much more valuable for biological monitoring in poorly studied regions. More generally, it is essential to be able to distinguish taxa at the lowest taxonomic resolution (i.e., to species) to understand their ecology and evolution. This is also of great relevance when developing tools to assess ecological status to ensure sustainable use and management of natural resources.
AEH, SCJ, SP, and WG developed and planned the paper. AEH, SCJ, and SP conducted field work. AEH generated molecular, geographic and elevational data and performed molecular association analyses. AEH, SV and WG studied morphology of larvae; AEH drafted larval type descriptions with support from SV and WG. AEH and WG photographed larvae. AEH wrote the initial manuscript and all authors contributed substantially to further versions of the manuscript.
We thank Hans Malicky (Lunz am See), Wolfram Mey (Berlin), Bob Wisseman (Oregon), R. B. Kuranishi (Chiba), and Hiroyuki Nishimoto (Komaki) for providing valuable specimens for the study. Field work and molecular work were funded by DFG Grant PA1617/2-1. WG was financially supported by the EC-funded Synthesys Access programme (DE-TAF-5887: Morphological characterisation and description of molecularly associated and differentiated clades of larvae of genus Himalopsyche (Trichoptera, Rhyacophilidae)). SCJ acknowledges funding from the German Federal Ministry of Education and Research (BMBF) for the “GLANCE” project (Global change effects in river ecosystems; 01LN1320A). We thank Ram Devi Tachamo Shah, Deep Narayan Shah, Fengqing Li, Subodh Sharma, Qinghua Cai, Xiaoli Tong, and Felicitas Erzinger (née Hoppeler) for invaluable contributions in the field, as well as the Department of National Parks and Wildlife Conservation (DNPWC) Nepal for providing the research permits.
List of specimens
Figure S1. Life stage association results based on CAD
Figure S2. Life stage association results based on COI