Speculations about caddisfly relationships, based on morphological evidence and hand calculations have been published for many years (reviewed by Morse 1997). More recently, molecular techniques have accelerated this field of enquiry (Malm et al. 2013; Kjer et al. 2016; Thomas et al. 2020; Ge et al. 2022), motivated in part by increased appreciation of the ecosystem services provided by caddisflies, especially the use of these insects in assessment of water quality (e.g. Morse et al. (2019)). An understanding of these historical relationships is useful for interpreting evolution of known functional traits of caddisflies relevant for those ecosystem services and for proposing and testing hypotheses of yet-unknown traits. In addition to re-assessing the phylogenetic relationships of caddisfly families with six independent genes and > 10 000 nucleotides, sequencing 185 species representing 49 families, Thomas et al. (2020) also estimated the ages of the resulting clades using BEAST, concluding that “the most recent common ancestor of Trichoptera appeared in the Permian (ca. 275 Ma) in line with the first appearance of Trichoptera in the fossil record and that vicariance explains the distribution of most trichopteran taxa.”

The identity and dates of many fossils were re-assessed by Nicholson et al. (2015), providing updated information for recalibrating the age estimates by Thomas et al. (2020). We report here our revised estimates resulting from those recalibrations.

Ages of lineages were estimated by BEAST as described by Thomas et al. (2020), but recalibrated with fossil dates provided by Nicholson et al. (2015) and more recently available dates from The Paleobiology Database (2023). The phylogenetic estimate is based on the multiple sequence alignment from Thomas et al. (2020), except an unresolved node was manually constrained with findings using genomic data from Frandsen et al., in prep. Fossil families are included in the phylogeny, based on morphological evidence from Wichard and Müller (2022, for Palleptoceridae), references cited by Nicholson et al. (2015) and The Paleobiology Database (2023).

Historical relationships of Trichoptera families with estimated minimum ages and fossil evidence are shown in Table 1 and Fig. 1. Thin lines indicate relationships only; their lengths are not supported by fossil evidence and are arbitrary and uninformative. Thick lines indicate the extent of fossil evidence (Nicholson et al. 2015). Solid lines show relationships according to Thomas et al. (2020) with a constraint from Frandsen et al., in prep. Broken horizontal lines show relationships based on morphological evidence only. Dots indicate re-estimates of minimum age. The re-estimated maximum age is at least 174.0 Ma for all Phryganides lineages and at least 201.3 Ma for all others.

According to Nicholson et al. (2015), amongst the family-level lineages, in Annulipalpia, the extant philopotamid lineage is represented confidently by a fossil from the Pliensbachi (J1, median age 186.3 Ma, with our estimated age greater than 174.1 Ma). Fossil families †Electralbertidae, †Protobaikalopsychidae and †Vitimotauliidae persisted during the Cretaceous. Estimated minimum ages for extant families Polycentropodidae, Dipseudopsidae, Ecnomidae, Psychomyiidae, Pseudoneureclipsidae, Hydropsychidae and Stenopsychidae were greater than 145.0, 125.5, 125.45, 113.0, 93.5, 46.2 and 33.9 Ma, respectively.

Again, according to Nicholson et al. (2015), in Integripalpia, the oldest lineage, dating from the Norian (T3, median age 222.6 Ma), is of †Necrotauliidae, a basal lineage of Integripalpia that survived to at least the Berriasian (K1, median age 142.85 Ma). Fossils of extinct families †Dysoneuridae, †Baissoferidae, †Burmapsychidae and †Cretapsychidae, †Taymyrelectronidae and †Yantarocentridae date from 173.6, 162.95, 105.8, 84.65 and 44.5 Ma, respectively. The oldest fossil of extant family Rhyacophilidae is from the Oxfordian (J3, median age 158.4 Ma., compared with our estimated age greater than 83.5 Ma). Similarly, the oldest fossils of Hydrobiosidae and Phryganeidae are from the Tithonian (J3, median age 148.15 Ma, compared with our estimated ages greater than 33.9 Ma and 93.5 Ma, respectively). Estimated minimum ages for other extant integripalpian families range from at least 145.0 Ma for Glossosomatidae to 33.9 Ma for Atriplectididae, Beraeidae, “Goeridae,” Hydrobiosidae and Odontoceridae. The similar estimated age of these five integripalpian families (and annulipalpian Stenopsychidae) is curious, inviting investigations about geological and ecological conditions that may have accelerated diversification about this time.

Ge et al. (2022) provided an alternative phylogeny to that shown in Fig. 1. It was based on mitochondrial genomes and notably places Hydroptilidae at the base of the Annulipalpia clade. Confirmation of this alternative awaits further supporting evidence.

Additional information about fossil caddisflies in Burmese amber was recently provided by W. Wichard (2023).

To serve as a basis for interpreting functional traits of caddisflies, Fig. 1 depicts not only our current understanding of relationships of extant and fossil caddisfly families, but also general understanding of their habitats, habits and trophic relationships, based mainly on the summaries by Holzenthal et al. (2007, 2015) and Cummins et al. (2019). A manuscript interpreting the evolution of these functional traits is in preparation.

Conceptualisation: JAT, PBF, JCM. Literature review: JAT, PBF, JCM. Investigation: JAT, PBF, JCM. Writing-Original draft: JCM. Writing-Review/editing: JAT, PBF, JCM.

The authors declare no conflicts of interest. No other persons or entities had a role in the design of the study; in the collection, analyses or interpretation of the data; in the writing of the manuscript; or in the decision to publish the results.