Which butterflies have the longest life span

Life flutters by quite quickly for these beautiful winged insects. The longest lived adult butterflies live for nine months to a year, but the average lifespan for the majority of species is just two to four weeks.

For example, monarch butterflies typically live between weeks, but it varies from generation to generation. Some generations also migrate and hibernate, which can more than double the lifespan. Other migratory species like the painted lady live for only two weeks. Some measure of phylogeny, although not part of best models for both LAB and MMR, was part of most models from the confidence sets, which scored high in Akaike weights.

For further details, see Material and methods; data presented only for life span maxima. In effect, measures represent the occurrence of variables in models, weighted by model probability. This can be seen as an assessment of the importance of particular variables in predicting life spans. MRR model did not include Warning coloration and host plant chemical syndromes. Our data compilation was of immense heterogeneity with regard to the quality and reliability of individual species data, and the methods used to measure them.

Apart from random sampling errors, we may assume that studied local populations might have specific reasons for average or maximum longevities attained and, for scoring of our explanatory traits, we often had only coarse classifications available. Against this background, it is unsurprising that the explanatory powers of our statistical models are not particularly high. Nevertheless, we were able to extract some clear, consistent contingencies of adult life spans and other species traits, which allowed more precise formulations of evolutionary hypotheses to be addressed in further e.

Life span measures derived from the literature may be biased in various ways. Time of death can always be assured, and also the time of eclosion from the pupa was precisely known in many, but not all Beck, studies. Keeping conditions may, on the other hand, be closer to optimal in some species, but not in others. The MRR data, however, suffer from various interrelated sources of error.

Measurements of the residence time from the first to the last sighting systematically underestimate true longevity because such data do not allow disentangling mortality and emigration. As a result, median MRR life spans were smaller than LAB medians in those species where both types of data were available, whereas with life span maxima this bias vanished see Differences in data types.

The underestimation of longevity from the MRR data could be mitigated by applying more sophisticated models Clobert, However, relevant data were not reported in the large majority of our data sources and could therefore not be used in that regard. Using a model selection approach in multivariate analysis, we were able to narrow down attention from a large number of possible interactions in the system to a handful of variables of which we may assume relevant simultaneous effects on life spans.

This does not mean that no other, more complex interactions may play a role, but our analytical results suggest the factors identified thus far are the primary phenomena that need to be understood in the course of further research. Among potentially relevant factors not included in our analysis are body size and related morphological and physiological measures Brown et al.

Explicit inclusion of phylogenetic information into models did not affect general conclusions inferred from univariate assessments and multivariate AIC-best models. This indicates that either direct phylogenetic inertia of traits do occur or, unsurprisingly, that we have failed to include all important traits in our analyses or did not have relevant data available , most of which would probably be related to phylogeny to some extent.

We propose and discuss five hypotheses that may have promoted the evolution of the patterns recovered in the present study. Seasonality and other features of environmental stability should affect life spans and other life-history traits, but we found support for this idea only in the MRR data.

Here, tropical species are living longer than those from presumably more seasonal temperate areas. We expect that this effect would increase in strength with a more detailed account of habitat conditions and climate for each species. We believe that lack of support for this hypothesis from the LAB data indicates an artefact of less suitable keeping conditions in most available studies on tropical butterflies. We cannot assess, however, whether or not it is a genuine effect.

Qualitative differences in adult feeding resources are clearly linked to adult longevity. Various lines of evidence point towards amino acids as a crucial nutrient. Such direct evidence is so far lacking for fruit feeders, but butterfly-utilized, over-ripe fruits were shown to be rich in amino acids Molleman et al.

Beck showed that those species that could utilize adult feeding-derived amino acids from fruits or nectar were among the most long-lived taxa in a Bornean forest assemblage. We predict that long-lived nectar feeding species feed mainly on flowers with high essential amino acid content, whereas short-lived taxa would be less discriminating. It is tempting to interpret access to adult-derived amino acids as a conditio sine qua non for long life spans and associated traits , rather than an evolutionary cause.

Only taxa that, for whatever reasons, switched their feeding habit towards sources rich in amino acids could evolve long life spans if this posed evolutionary advantages Beck, We would expect to find still a wide variety of life spans even among taxa feeding on a rich resource. Consistent with this prediction, reported maximum life spans among pollen-feeders vary between 21 Heliconius sappho and days Heliconius hecale. Such an explanation may be convincing in the case of pollen feeding, where a complete clade switched to this derived food source.

However, fruit-feeding may have evolved several times among the Nymphalidae as evident also by different proboscis adaptations; Molleman et al. The clade Mycalesina, for example, comprises fruit-feeders within the otherwise nectar-feeding Satyrini.

If such back- and forward shifts of adult resource use are common, it might be important to understand why they occurred and whether they relate to life-span and associated life history trade-offs.

Such a mechanism would also be reasonable as an evolutionary explanation of interspecific variation in adult food resource use, although Beck did not find supporting evidence in a relatively crude approach.

Species feeding on poor larval hosts could make up for this by using high-quality adult food. This requires more adult time to either find high-quality nutrients or to utilize these resources e. Species that are able to balance the risk of this strategy premature mortality , for example, by efficient predator avoidance, would benefit from being able to exploit poor larval hosts less competition, less time to be invested in search for larval hosts.

If this hypothetical mechanism were true, one would predict that butterfly species feeding on larval food plants poor in nitrogenous compounds are more likely to utilize adult foods rich in amino acids. We predict that such patterns would be explicable in a broader life-history assessment, as discussed here, and that the amount or concentration of essential amino acids available in the nectar of different flowers would play a crucial role in such an explanation.

Traits leading to reduced predation offer a straightforward mechanistic explanation for the evolution of longevity. They are of advantage already to the individual assuming that even in short-lived taxa predation is a major cause of premature death and, once established, they allow the evolution of fitness-enhancing traits such as prolonged ovary function and life span.

The effects of Batesian mimicry were much less conclusive, indicating that, in many species, mimicry-derived predator avoidance does not work efficiently enough over evolutionary time to permit consequences such as the evolution of longer physiological life spans. Based on this reasoning, we would predict more effective anti-predator functions with regard to their natural predators of mimicry or eye-spots in long-lived species e. Caligo sp. Conversely, with regard to aposematic taxa, we predict that short-lived species e.

Danaus affinis , Danaus gilippus are relatively inefficient in predator deterrence, possibly due to complex intra- and interspecific mimicry relationships Ritland, Mud-puddling behaviour is negatively associated with adult life spans. Based on similar results from a smaller data set, Beck hypothesized that mud-puddling is a particularly risky behaviour, due to elevated predation outside a camouflage-suited background, or because of increased risk of e.

However, mud-puddling is almost exclusively a male behaviour. We would expect, therefore, that females of mud-puddling species live longer than their males. If high male mortality due to predation or other factors were a general feature of mud-puddling butterflies, we might even expect that in some species females might exploit nuptial gifts derived from mud-puddling of their mates as indicators of male quality, irrespective of the fitness value of the gifts themselves.

Apart from water as a potential puddling stimulus in arid landscapes Launer et al. However, Beck et al. This was, in particular, observed in ten lycaenid genera [ Acytolepis puspa , Anthene emolus , lycaenina , Caleta elna , manovus , Discolampa ethion , Ionolyce helicon , Monodontines musina , Nacaduba beroi , Prosotas bhutea , gracilis , pia , Udara dilecta , dilectissima , Zeltus amasa ], seven nymphalid genera Athyma larymna , Charaxes bernardus , Chersonesia intermedia , Cyrestis maenalis nivea , Phalanta alcippe , Polyura athamas , Vindula dejone , erotus , in two Papilio helenus , memnon , one Eurema indet.

Fiedler, previously unpublished data. If access to amino acids facilitated by mud-puddling has similar effects as suggested above for fruit feeding, we would expect that these species live longer than mud-puddlers searching for anorganic compounds. Unfortunately, we only had life spans for three of these species available, which did not allow any meaningful comparison.

Jervis et al. On the other hand, we found strong evidence that the growth form of host plants may be related to life spans. We find it difficult to interpret whether this indicates a causal link between growth form and butterfly life history e. Possibly, our host plant chemistry classification contains a lot of information irrelevant to butterfly life span, which are therefore better captured by our much simpler growth form classification.

Figure 4 shows the frequency of chemical syndromes in different growth forms. Based on this, and the effects of chemistry in multivariate models not shown , we reclassified host plant chemistry using only the most prominent chemical categories.

However, a model containing this new classification did not surpass i. Speculatively, lianas might have a tendency for higher toxicity, independently of the class of chemical compounds involved or phylogeny. In the tropics, they are often among the few plants that produce buds and fresh leaves i. Liana-type growth forms might thus have evolved more often among toxic plant clades, which in turn may allow those herbivores able to feed on lianas to also utilize these plant secondary metabolites for their own protection.

However, we were unable, at the present state of knowledge, to substitute growth form with the relevant chemical traits that might be ultimately responsible for the observed effect. Alternatively, our finding might be due to a bias in our data.

Most liana feeders for which we were able to collate life span information i. Although the best model generally accounts for these supposedly independent effects on life spans, multiple parameter collinearity might have affected variance partitioning in our models Graham, In a data-mining approach on a mostly literature-derived data set, we lay out broad patterns of trait contingencies as far as available with butterfly adult longevity. We see at least three avenues of further research on this topic.

We present some hypotheses to explain recovered patterns, and suggested approaches of testing or supporting them. In particular, linking those to data on habitat choice and behaviour may be rewarding for scientific understanding as well as for conservation Tudor et al. A broader extension of this research agenda to nocturnal moths which make up the large majority of diversity of the Lepidoptera , or at least to the taxonomically better known Macrolepidoptera, would make available for analysis a broader spectrum of life history strategies and traits variables.

Three anonymous referees made useful comments that helped with presenting and discussing our results thoroughly. Abouheif E A method for testing the assumption of phylogenetic independence in comparative data. Evolutionary Ecology Research 1 : — Google Scholar. Why do male butterflies visit mud puddles? Canadian Journal of Zoology 60 : — Community heterogeneity and the evolution of interactions between plants and insect herbivores.

The Quarterly Review of Biology 81 : — Assessment of higher insect taxa as bioindicators for different logging-disturbance regimes in lowland tropical rain forest in Sabah, Malaysia. Ecology Research 22 : — Botanical Journal of the Linnean Society : — Sodium: stimulus for puddling behavior by tiger swallowtail butterflies, P.

Science : — Beck J The importance of amino acids in the adult diet of male tropical rainforest butterflies. Oecologia : — Diet breadth and host plant relationships of Southeast-Asian sphingid caterpillars. Ecotropica 12 : 1 — Effects of habitat disturbance can be subtle yet significant: biodiversity of hawkmoth-assemblages Lepidoptera: Sphingidae in Southeast-Asia.

Biodiversity and Conservation 15 : — Wallace's line revisited: has vicariance or dispersal shaped the distribution of Malesian hawkmoths Lepidoptera: Sphingidae? Biological Journal of the Linnean Society 89 : — Beck J , Linsenmair KE Feasibility of light-trapping in community research on moths: attraction radius of light, completeness of samples, nightly flight times and seasonality of Southeast-Asian hawkmoths Lepidoptera: Sphingidae.

Journal of Research on the Lepidoptera 39 : 18 — The long-life champion is a tiny yucca moth that feeds on Banana Yucca Yucca baccata. Jerry Powell of the University of California at Berkeley has found this moth's caterpillar may be able to wait up until 30 years to form a pupa and emerge as an adult.

Skip to main content. It works like a straw and, when not used, is recoiled under the head. Some species also get the food from sap flows on the tree as well as pollen, rotting fruit, dung, and even bird dropping.

They help the butterflies to blend in with the environment and avoid attacks from predators. It is always male butterflies that you see playing in mud puddles, where they take in salts and minerals to strengthen their sperm and encourage breeding. The nutritions in these puddling sites get more concentrated as the water evaporates, and the butterflies will continue visiting them until they are nearly dry. Learn about our giant swallowtail butterfly rearing journey. Cart 0. Leopard Stickers.

Back About Editorial Process. How Long Do Butterflies Live? By Species. When you see a butterfly flying over your garden, have you ever wondered how beautiful their life has been?

The Amazing Life Cycle of Butterflies Before emerging into butterflies, caterpillars will experience three life stages, the egg, the larva caterpillar , and the pupa chrysalis.