135
range of present-day hosts (De Baets et al. 2015). Any inferences concerning the
origins and evolutionary histories of parasitic lineages should therefore acknowledge the possibility of a potentially rich complexity of host-parasite interactions
that may have been impacted over time by gradual and mass extinctions. Indeed, it
is fascinating to contemplate what parasites may have been lost during mass extinction events and what host lineages may have survived and subsequently diversified
having lost those parasites (van Dijk and De Baets 2021). It should be noted that
there is a growing appreciation that co-extinction of parasites (affiliate taxa) along
with taxa they depend on (hosts) may be the most common form of species loss
(Dunn et al. 2009; Strona 2015). However, as far as we are aware, the modelling
approaches developed so far to estimate parasite extinctions are constrained to estimating the number of extinctions of affiliate species as a function of host extinctions
retrospectively, based on knowledge of present-day patterns (Colwell et al. 2012),
or by modelling dynamics of digitally evolving organisms (reviewed in Strona
2015). Parasite extinctions have also been inferred by detection of ancient DNA
(e.g. host-specific heterokoid nematodes in coprolites of moa; Boast et al. 2018), but
this technique cannot be applied to reveal extinctions over deep time.
In addition, limited sampling of the diversity of many parasites will constrain
insights about patterns of host use. Current undersampling of both host and parasite
diversities may thus compromise current views, for example that all major myxozoan clades are characterised by different invertebrate host groups. This limitation
is particularly relevant when cophylogenetic analyses attempt to characterise patterns of parasite co-evolution relative to focal host groups. Such analyses require
reliable phylogenetic trees—otherwise congruence of phylogenies may be compromised by poor phylogenetic resolution.
4.6.2.2 Endocnidozoan Origins and Host Use Over Time
The first attempt to characterise the timing and divergence of endocnidozoans
(Holzer et al. 2018) provides a useful platform for interpreting endocnidozoan evolution as well as to illustrate how inferences may be compromised by the various
pitfalls of such analyses. We variously elaborate on these issues below.
To estimate divergence times, Holzer et al. (2018) analysed concatenated alignments of six protein-coding genes in endocnidozoans, their hosts, and other metazoans (10 myxozoans, Polypodium and 127 other metazoan taxa from Erwin et al.
2011). Molecular clock analyses estimated divergence times of endocnidozoans and
myxozoans to be in the late Cryogenian at 651 Ma (700–601 Ma) and the Ediacaran
at 588 Ma (642–540 Ma), respectively. As outlined earlier, such divergence times
are consistent with other molecular clock analyses which predict cnidarian origins
in the Cryogenian (720–625 Ma) and the presence of crown cnidarians in the uppermost Ediacaran or earliest Cambrian. However, one of the main conclusions, that
endocnidozoans diverged when invertebrates were incorporated as first hosts in the
Cryogenian, is not well supported. Parasitism of stem hosts other than invertebrates
is also possible. Although crown group vertebrates are estimated to be much
4 Evolution, Origins and Diversification of Parasitic Cnidarians
range of present-day hosts (De Baets et al. 2015). Any inferences concerning the
origins and evolutionary histories of parasitic lineages should therefore acknowledge the possibility of a potentially rich complexity of host-parasite interactions
that may have been impacted over time by gradual and mass extinctions. Indeed, it
is fascinating to contemplate what parasites may have been lost during mass extinction events and what host lineages may have survived and subsequently diversified
having lost those parasites (van Dijk and De Baets 2021). It should be noted that
there is a growing appreciation that co-extinction of parasites (affiliate taxa) along
with taxa they depend on (hosts) may be the most common form of species loss
(Dunn et al. 2009; Strona 2015). However, as far as we are aware, the modelling
approaches developed so far to estimate parasite extinctions are constrained to estimating the number of extinctions of affiliate species as a function of host extinctions
retrospectively, based on knowledge of present-day patterns (Colwell et al. 2012),
or by modelling dynamics of digitally evolving organisms (reviewed in Strona
2015). Parasite extinctions have also been inferred by detection of ancient DNA
(e.g. host-specific heterokoid nematodes in coprolites of moa; Boast et al. 2018), but
this technique cannot be applied to reveal extinctions over deep time.
In addition, limited sampling of the diversity of many parasites will constrain
insights about patterns of host use. Current undersampling of both host and parasite
diversities may thus compromise current views, for example that all major myxozoan clades are characterised by different invertebrate host groups. This limitation
is particularly relevant when cophylogenetic analyses attempt to characterise patterns of parasite co-evolution relative to focal host groups. Such analyses require
reliable phylogenetic trees—otherwise congruence of phylogenies may be compromised by poor phylogenetic resolution.
4.6.2.2 Endocnidozoan Origins and Host Use Over Time
The first attempt to characterise the timing and divergence of endocnidozoans
(Holzer et al. 2018) provides a useful platform for interpreting endocnidozoan evolution as well as to illustrate how inferences may be compromised by the various
pitfalls of such analyses. We variously elaborate on these issues below.
To estimate divergence times, Holzer et al. (2018) analysed concatenated alignments of six protein-coding genes in endocnidozoans, their hosts, and other metazoans (10 myxozoans, Polypodium and 127 other metazoan taxa from Erwin et al.
2011). Molecular clock analyses estimated divergence times of endocnidozoans and
myxozoans to be in the late Cryogenian at 651 Ma (700–601 Ma) and the Ediacaran
at 588 Ma (642–540 Ma), respectively. As outlined earlier, such divergence times
are consistent with other molecular clock analyses which predict cnidarian origins
in the Cryogenian (720–625 Ma) and the presence of crown cnidarians in the uppermost Ediacaran or earliest Cambrian. However, one of the main conclusions, that
endocnidozoans diverged when invertebrates were incorporated as first hosts in the
Cryogenian, is not well supported. Parasitism of stem hosts other than invertebrates
is also possible. Although crown group vertebrates are estimated to be much
4 Evolution, Origins and Diversification of Parasitic Cnidarians