Flying with dinosaurs: the evolution of moths and butterflies

Serendipity plays a pretty important role in scientific advances; indeed, it was involved in the discovery of penicillin, microwaves and x-rays. And now, it seems a bunch of old moth scales can be added to that list.

Scientists drilling cores from lake sediments in Germany – hoping to learn about past ecosystems from ancient pollen grains – recently stumbled across a profusion of tiny scales from moth wings. This is significant as the sediments are a whopping 200 million years old, making it the earliest appearance of moths and butterflies in the fossil record. If the history of the planet was represented as a 24 hour clock, this means the first moths would appear some time before 11pm (modern humans only evolved a few seconds before midnight).

Was a Triassic landscape like this home to the very first moths? Image: Oxford Scientific/Getty (RF)

It hadn’t occurred to me that moths have been around for quite so long. Hopefully the process of writing a blog post will go some way to remedy my lack of knowledge on the early evolution of my favourite group of insects…

The very first moths

By using a suite of techniques including genomic sequencing, constructing evolutionary trees and radiocarbon-dating fossils, researchers estimate moths first evolved over 250 million years ago. Around this time, an enterprising group of individuals began to forge an identity separate from the ancestral group (now long extinct). These ancestors are presumed to have lived in wet habitats and gave rise to caddisflies, as well as the first moths.

About the size of a grain of rice, the earliest moths were probably found in ancient humid forests, where the caterpillars lived in damp soil or moss. The primitive adults did not sip nectar like their contemporary cousins but instead had mandibles (jaws) and fed on pollen grains or fern spores.

A living fossil: Microropterix are thought to be similar to the first moths that lived over 200 million years ago. Image: M. calthella, Patrick Clement/Flickr (CC BY 2.0)
Scaling the heights

Due to their delicate wings and buoyant scales, butterflies and moths typically fossilise very poorly. In the German lake cores, Timo van Eldijk’s team found about 70 individuals scales. This raises the question of whether there was anything special about this particular area, or have the tiny scales simply been overlooked in the past? I suspect the latter.

Butterflies and moths are characterised by the scales that cover their wings; indeed, the name of the taxonomic order they belong to – Lepidoptera – derives from the Greek word for scale (‘lepis’) and wing (‘ptera’). These scales, which are essentially flattened hairs, distinguish moths and butterflies from their closest living relatives: caddisflies (Trichoptera = ‘hairy-winged’).

Scales are central to the success of Lepidoptera; the infinite medley of intricate patterns and colour combinations they can generate is likely to have facilitated the divergence of the c.200,000 described species of butterfly and moth alive today.

Scales provide the canvas that is used to develop elaborate camouflage tactics or, conversely, enable gaudy colouration to advertise their unpleasant taste to potential predators. Scales may also afford stealth to moths by absorbing sound thus rendering bat echolocation less effective; and have been shown to improve aerodynamics in migratory butterflies. The scales that cover their bodies are also advantageous, providing a thick fur coat to keep flight muscles above the critical temperature required for flight (a rather toasty 30°c).

Scales tile the wings of butterflies and moths. Image: Papilio ulysses, Nikola Rahme/Shutterstock (RF)

While lepidopterous scales often contain pigments that confer colour, many brighter shades (especially iridescent blues and greens) are purely the result of structure; the micro-architecture of the scales scatters and reflects light to produce colour.

Broader variation in scale shape and structure is often distinctive, allowing individual scales to be assigned to particular families of butterflies and moths. A varied selection of scale types was noted in the ancient lake sediments, including ones similar to the primitive pollen-munching Microropterix still alive today. Significantly, hollow scales were also uncovered. These are exclusively found in Glossata, a group of Lepidoptera equipped with a tongue-like structure called a proboscis.

Prior to the discovery, this particular group was thought to have evolved over 70 million years later in the Cretaceous period, capitalising on the explosion of flowering plant diversity that occurred at that time. This presents something of a conundrum. What were these lepidopterans using their proboscis for, if nectar-bearing flowers were yet to arise? The authors propounded the idea it was an adaptation to the parched conditions in the Triassic, allowing thirsty individuals to suck up moisture.

The tight interactions formed between flowers and their insect pollinators often drive coevolution, so the finding that proboscis-possessing Lepidoptera were around before flowering plants also raises an interesting prospect. Could moths and butterflies have been a driving force behind the spectacular radiation of flowering plants?

Fluttering through the eras
A geological timeline of the last 250 million years with key events and the number of Lepidoptera families present in the fossil record illustrated.

250 million years of evolution have carefully fashioned different ecological niches for the 125 families of moths and butterflies alive today. While the number of families present in the fossil record remains low until the last 50 million years, the low fossilisation rate means it’s difficult to know for sure if the group diversified as late as this (Lepidoptera are usually thought to have radiated in the cretaceous period).

Whenever it occurred, the diversification of Lepidoptera was accompanied by pronounced shifts in feeding habits. The early caterpillars fed internally, within soils, stems or woody tissues. It wasn’t until later that some groups ventured above ground to feed on foliage – though, at first, only staying within the narrow tunnels they mined through leaves. Fossilised patterns produced by these leaf-mining moths date back 100 million years. Such conspicuous clues left by internal feeders could have easily attract predators and parasitoids; the prospect of being a ‘sitting duck’ may be why some groups have sacrificed shelter and embraced the mobility afforded by external feeding.

From the leaves to the roots, different groups of caterpillars have evolved the ability to exploit almost every part of a plant. This vast array of feeding strategies has enabled the divergence of many different families, but the spectacular diversity within families is largely due to specialisation in eating certain species of plant (many moths and butterflies only feed on a single type of plant). A handful of groups have shunned plants altogether, with some moth larvae preferring to gorge on fungi, other caterpillars, dead animals, or even sloth poo!

A simplified evolutionary tree showing some of the major Lepidoptera families (based on phylogenies from Kawahara & Breinholt, 2014 and Mitter et al., 2017). Note how the distinction between micro moths (orange) and macro moths (green) has no basis in evolutionary history.

In the adults, there have been multiple switches between favouring daytime activity and flying at night. The earliest evidence of a butterfly is dated from the mid-Paleogene, though they probably evolved tens of millions of years earlier. Intriguingly, many adult moths have lost the ability to feed, relying on food stores accumulated as larvae. Body size has gradually increased and many of the more spectacular groups (including hawk moths and silk moths) only came into existence relatively recently.

In their 250 million years on this planet, lepidopterans have witnessed the emergence and diversification of dinosaurs, mammals and flowering plants; seen the breaking up of the Pangaea supercontinent; and survived two mass extinction events. The chances are they’ll still be around long after mankind has gone extinct. Pretty good going for a fragile group of insects I’d say.

Sleepwalking towards Armageddon? We need more long-term ecological studies

Widely-reported research has led some to suggest we are “on course for ecological Armageddon”. Behind these headlines: an analysis of a German dataset spanning nearly three decades detected a 76 percent plummet in flying insect biomass. So is now the time to be building our apocalypse bunkers?

Forming the base of most food chains and providing vital services such as pollination, insects play a unique role across terrestrial habitats. Due to their sensitivity to environmental change, they are the ‘canary in the coal mine’. If the research findings from Germany are representative of wider insect populations across Europe, the implications for ecosystems and human wellbeing are likely to be catastrophic.

The windscreen phenomenon: anecdotal evidence for flying insect declines has come from a reduction in the bugs splattered on the front of cars. Image: RiverNorthPhotography/iStock (RF)

The recent research from Germany paints a much bleaker picture than previous studies of insect populations. Data from a large network of standardised light traps spread across the UK revealed that macro moth numbers had ‘only’ declined by a third since 1968. A smaller network of suction traps operated since the 1970s by the same organisation, Rothamsted Research, showed that total insect biomass was only declining in one of the four sites studied.

Ignorance is bliss?

The idiosyncrasy of these findings demonstrates the need for further long-term ecological studies. Are these divergent trends suggesting that insect declines have been worse in some countries? Are different groups of insect faring differently? We simply don’t know.

The design of long-term studies may contribute to the uncertainty. To detect accurate abundance changes, sampling should aim to minimise confounding variables, such as location. Surveys should ideally occur over consecutive years in the same place, allowing individual trends to be calculated (which can then be averaged over multiple locations). However, over half of the sites visited in the German study were only sampled for a single year during the 27 year study period. Did this imperfection affect the findings of this study? We need further long-term studies to be sure.

Malaise traps, which funnel flying insects into a collecting vessel, were used to measure biomass in German protected areas. Image: Hallmann et al., 2017/PLOSONE.

The assumptions made in these studies are also important to consider. Biomass is not the same as abundance (although it is usually a good proxy). In theory, the declines in insect biomass could just be due to the loss of a handful of very large species. Only with additional research are we likely to be able to understand exactly what’s going on.

Good science often takes time

Systematic ecological monitoring over extended periods of time is expensive; shortage of funding is a key barrier to good science in this regard. Long-term continuity is vital; however, most research grants are for less than five years. Publicly-financed long-term projects are heavily underfunded and are often seen as dispensable when times are hard.

In Australia, the decision was recently made to axe funding for a nationwide, biodiversity and ecosystem monitoring project; the network, which was established in 2012, consisted of over 1100 plots and was intended to run until at least 2025. It’s ironic that a project designed to inform billion dollar land management decisions was terminated for the sake of saving less than a million dollars a year.

It seems intuitive that long-term ecological studies are beneficial; however, the advantages have also been demonstrated quantitatively. Trends become more obvious and predictability is improved as more data is amassed. Furthermore, long-term studies tend to have more impact, and frequently influence environmental policy.

Long-term studies allow trends to be discerned in noisy data. Figure from The State of Britain’s Larger Moths, showing how the abundance of macro moths has changed in Britain.

The value of long-term monitoring studies may also extend beyond simply being descriptive. With careful analysis, it might be possible to tease out the mechanisms responsible for change. A study using three decades of data from the UK’s Butterfly Monitoring Scheme showed declines were steepest in areas of high neonicotinoid pesticides use. While it is impossible to prove causation from such studies, they can highlight factors for future experimental study.

On the way out?

Just like the organisms they monitor, long-term ecological studies seem to be in decline. The study of natural history has fallen out of fashion, while simplified predictive models are in vogue. There is no substitute, however, for well-designed observational studies. Citizen science approaches have become trendy but their power to examine long-term trends in populations is likely to be limited in most cases; meaningful and rigorous analysis of population trends typically requires standardised methodology and high-quality data.

Once funding is pulled from existing long-term studies, we lose the ability to track the pervasive effects of contemporary environmental change. This is a scary thought when we consider the diminishing flying insect populations in German reserves. How many similar biodiversity trends are going undocumented? We may not be headed for ecological Armageddon, but being in the dark about the health of our ecosystems should be just as worrying.

Well, this is new…

Phew! After a couple of days battling with PHP and trawling through CSS code, I have finally added a blog to my website. Please let me know if you spot any awry formatting or if something doesn’t work as it should!

I’m not exactly sure how I’m going to use this space yet but do check back – I’ll try to post something once every month or two.