The foundation of psyllid diversity in New Zealand may be the result of a combination of ancestral arrivals and subsequent species radiation within the country. Understanding this, and thus the evolutionary strategies adopted by the different families and genera, would be useful for many reasons. First of all, a better understanding of the origin of psyllids could inform the modelling of future routes or risks of invasion by pest species (Syfert et al. 2017). Moreover, the geographic origin of psyllids may help in understanding the ancestral psyllid-host plant associations, which could be useful to retrospectively understand the risks for future plant colonisations, possibly based on climatic similarities as it has been demonstrated elsewhere (Syfert et al. 2017). A much more complete taxonomic dataset, including that from potential ancestral sources, is necessary to enable these in the future. Nevertheless, based on the phylogenetic range here, a number of hypotheses can be generated that may form the basis for such future work.
Dating psyllid arrivals to New Zealand
Associating the phylogeny here with a molecular clock may allow the arrival times of different psyllid groups to be determined. Consequently, knowing the time of ancestral arrivals could contribute to answers as to their geographic origin. This especially considering that New Zealand was part of Gondwana and some archaic lineage could have originated from that time. For example, are some of the lineages, such as Anomalopsylla, relics of the super continent land mass of Gondwanan times, or
161 are they modern-day dispersers? Unfortunately, calibration methods based on fossils, geological events or mutation rates (Hipsley and Muller 2014) are not easily accessible here.
Fossils belonging to the family Psylloidea or its ancestors are scarce. The oldest crown group psyllid appear in Baltic amber during the Eocene (Klimaszewski 1996). While recent studies on fossils preserved in Mexican amber suggest that the Miocene fauna was quite similar to the contemporary one (Drohojowska et al. 2016). The superfamily Psylloidea, however, may well have had
representatives from the late Jurassic (Mesozoic). In fact, specimens dated back to that period have been assigned to the extinct families Liadopsyllidae and Malmopsyllidae (Bekker-Migdisova 1985). However, a fossil to confirm the split time between psyllids and other Sternorrhyncha such as aphids, or between families within the Psylloidea, is still missing.
In absence of a fossil a geological event such as Zealandia’s separation from Gondwanaland
83 Mya (Goldberg et al. 2008) or from New Caledonia 55 Mya (Schellart et al. 2009) has been considered elsewhere [see (Goldberg et al. 2008)]. However, an arc of volcanoes between New Zealand and New Caledonia along the Three Kings ridge may have provided a path between the two land masses (Schellart et al. 2009). We know that after separation from Gondwanaland starting ~80 Mya, the continent of Zealandia gradually submerged beneath the sea, and that modern New Zealand is primarily the product of tectonic activity initiated ∼25 Ma [e.g. (Campbell and Hutching 2007)]. It is not known how much land persisted, probably fragmented in a number of smaller islands; however, extreme reduction of the landmasses is likely to have caused biological bottlenecks (Cooper and Cooper 1995). Land connectivity, however, may not have been instrumental in
ancestral arrivals as small winged insects such as psyllids are known to be easily windblown
(Burckhardt et al. 2014). This could account for more recent movement amongst land masses, with wind dispersal from Australia still considered as one of the most probable means of arrival (Yen et al. 2014). Future research in this area may instead find that the use of the substitution rate of
mitochondrial DNA is the best option for determining evolution of psyllids in New Zealand. A specific substitution rate has not been applied before to this group. The closest has been for the
Metrosideros-hostedpsyllids of Hawaii where the psyllid arrival on the islands was estimated according to arrival time of their hosts. Such data could provide an approximate substitution rate that could be then compared with the some of the most recent estimates for mitochondrial DNA substitution rates [e.g. (Brower 1994, Papadopoulou et al. 2010)]. However, the limitation of any given substitution rate, due for example to rate variation among lineages and over time, must be considered and accounted for at all times (Ho and Lo 2013).
162 The phylogeny obtained in this study may also generate some hypotheses as to arrivals vs radiation. For example, the presence of species native both to New Zealand and Australia, such as
Ctenarytaina, may suggest a recent, post Gondwanan, split of this genus between the two countries. In fact, the position of the crown speciation of the New Zealand native species in the phylogenetic tree appears to be at the same depth of other genera, such as Trioza. On the other hand,
Anomalopsylla and Atmetocranium appear to have diverged much earlier than Ctenarytaina. The study of these two genera would particularly benefit from a molecular clock to support a pre- or post- Gondwanan split. However, while the hypothesis of a Gondwanan origin for Atmetocranium
and Anomalopsylla might explain the very distinct morphology of these psyllids, a pre-Gondwanan origin may be unlikely based on a recent compilation of plant and animal phylogenetic analyses revealed that only 10% of those could be dated back to the splitting of Zealandia from Gondwana (Wallis and Trewick 2009).
The origin of the ancestral psyllids that colonized New Zealand: dispersal and radiation in
the Pacific region.
The phylogenetic information obtained here on the current New Zealand fauna can also contribute to a better understanding of the origin and pathways that led to the arrival of ancestral psyllids to New Zealand. Similarly, comparisons between the New Zealand psyllids and those present in other Pacific Islands may cast some light on the present distribution of psyllids in the Pacific region. These analyses, together with the most recent information on the geological history of New Zealand, could then enable new hypothesis on the psyllids origin to be formulated. For example,
thirty years ago the dispersal of psyllids was considered unlikely: “it cannot be assumed that they did
[disperse] just because they are small and have wings” (Dale 1985). Today, on the other hand, the presence of psyllids on recently emerged Pacific islands such as the Hawaiian Islands [estimated origin around 28 Mya (McDougall and Swanson 1972)] suggests that dispersal can be the only reason Hawaii is home to more than a 100 psyllid taxa (Ouvrard 2017). In fact, the arrival of the triozid genus Pariaconus in Hawaii has been dated after the arrival of its host, Metrosideros, about 3.9-6.3 Mya (Percy et al. 2008, Percy 2017). Therefore, the fact that the Hawaiian Islands emerged from the sea leaves no doubt on the present psyllid fauna must have originated via dispersal.
The mechanism and pathway of this dispersal, however, remain uncertain. Recent studies confirm that insect wind dispersal is feasible, at least between Australia and New Zealand [e.g. (Yen et al. 2014)]. Similarly, evidence of the trans-oceanic dispersal of plants has been known for a long time (Davis 1950, Gillespie et al. 2012); this might explain the arrival of Metrosideros to Hawaii probably from Australia (Tarran et al. 2016, Tarran et al. 2017) and not from New Zealand as previously thought (Percy et al. 2008). Oceanic drift of host plant material is in fact well known [e.g.
163 (Winkworth et al. 2002, Gillespie et al. 2012, Percy 2017)] and may have directly connected Australia or New Zealand with Hawaii. Alternatively, a psyllid wind-mediated dispersal may have been
facilitated by an “Oceanic pathway”, with Pacific Islands as stepping stones to accommodate the large distances. This would support the hypothesis of an initial plant radiation followed by a psyllid colonization of the plant as suggested for Hawaiian Islands (Percy et al. 2008, Percy 2017).
Phylogeographic evidence of this would require inclusion of the triozid faunas of other Pacific Islands, including Australia (and Tasmania), New Caledonia, Fiji, Vanuatu, up to the Marshall Islands. Observation of genetic variation correlated to inter-island proximity may consequently suggest an
establishment “pathway” between them.
As a first step in the comparison between the New Zealand psyllid species and those present in other countries, COI sequences available from previous work were able to be included;
unfortunately, a complete set of 18S sequences were not available. This cursory comparison of the
COI barcode sequences between New Zealand’s most basal triozid species of T. curta, plus the Australian triozids species analysed here (except T. eugeniae) and Hawaiian triozids (Percy 2017) suggests that the New Zealand species are more closely related to the Hawaiian than the Australian species. This is in contrast to the hypothesis of a pathway between Pacific Islands originating from Australia, where it might be anticipated that more closely located islands, such as New Zealand and Australia, would have more closely related species. Moreover, while insect wind dispersal has been confirmed between New Zealand and Australia (Yen et al. 2014), this would be less realistic for the more distant New Zealand and Hawaii islands, especially considering that the southern hemisphere trade winds are predominately from the south east (https://en.wikipedia.org/wiki/Prevailing_winds).
Thus, the development of a different hypothesis may be required to account for psyllid dispersal in the Pacific.
Psyllid biological habit was also considered as a possible facilitator of oceanic dispersal. In fact, considering the gall-forming guild of psyllids, it appears plausible that psyllids encased in their galls at the nymphal stage may be dispersed via oceanic drift of their host plants. Obviously, in order to confirm this hypothesis, ecological experiments on the survival rate of psyllid nymphs exposed to salt water while within their galls would be useful. If confirmed, this theory would be consistent with the idea of ancestral species arrivals potentially being gall-formers as is indicated by the
phylogenetic positions of those present in New Zealand: T. curta and T. eugeniae being basal to the New Zealand triozids, T. “Price’s Valley” basal to the monophyletic group of the New Zealand Trioza, and the Aphalarid Atmetocranium myersi basal to the New Zealand Aphalaridae. In keeping with this, the gall-forming habit may be an ancestral feature, as has been considered for the triozid genus
164 common in the present day for Asian and Hawaiian psyllids (Crawford 1918), could be a residual characteristic of their ancestors. For example, more than 50% of the triozid species of Taiwan and Japan are reported to consist of gall-formers (Yukawa and Masuda 1996, Percy et al. 2015).