- Top of page
The term ‘puddling’ includes feeding on (dried) mud and various excrements and secretions of vertebrates, and carrion. It is thought to be a form of supplementary feeding, not targeted at obtaining energy. Although the natural history of the puddling phenomenon in herbivorous arthropods becomes better known, it is still largely unclear how puddling (in particular for sodium) affects fitness despite the growing knowledge of insect physiology at the cellular level. If we follow the definition used for puddling in Lepidoptera, representatives of a wide range of herbivorous and detrivorous terrestrial arthropods (Lepidoptera, Orthoptera, Blattodea, Hymenoptera, Hemiptera, Diptera, and Diplopoda) have been observed to puddle. It appears that those species with diets low in sodium (e.g., folivorous larvae) puddle for sodium whereas those with diets low in nitrogen (e.g., detritivores) puddle for nitrogen. Sex differentials in puddling behavior can usually be explained by transfers of nutrients from males to females during mating. Puddling is rare or absent in immature stages and there is some evidence that nutrients from puddles increase female reproductive success. Strong evidence for the widely cited hypothesis that sodium from puddles is used to enhance neuromuscular activity is still lacking. High mobility and long life spans could be associated with puddling behavior, whereas insects that are concealed or well defended are less likely to puddle (e.g., beetles). The role that risks of pathogen and parasite infection as well as predation at puddling substrates may play in the evolution of puddling remains virtually unexplored.
- Top of page
Nutrition is a central issue in biology at a variety of levels – much of organismal and evolutionary biology is concerned with how organisms extract nutrients from their environment, allocate nutrients to different goals, or avoid becoming food for other organisms, ecologists study how elements move through the landscape and food webs, and metabolism is a central process for cellular biology.
The term ‘puddling’ stems from the mud-puddling behavior commonly seen in butterflies and also includes feeding on various excrements and secretions of vertebrates or carrion, which is often observed in the same species (Downes, 1973), and may be extended to feeding on eyes, nose or blood of vertebrates (Bänziger, 2007, Bänziger et al., 2009). Puddling is thought to be a form of supplementary feeding, not targeted at obtaining energy, but at specific micronutrients. The simple question ‘why are there so many butterflies puddling near roads and streams in the tropics?’ may be understood as an ecological, evolutionary, or physiological question, and gives rise to a series of new questions. These include: which nutrients are limiting fitness and why? How do nutrients interact? And why do species follow different feeding strategies? Answers to these questions promise to provide deep insight into evolution and ecology. Yet, puddling in insects is still poorly understood and common explanations are still lacking experimental evidence. With some new developments in this field, I aim here to offer a broad perspective by giving a brief oversight of what is known, identify gaps in our understanding of puddling, and provide some testable hypotheses.
After Norris (1936) included rich information on puddling behavior in her extensive review of the feeding habits of butterflies, a wealth of additional observations have been added to the puddling literature (Boggs & Jackson, 1991; Sculley & Boggs, 1996; Beck et al., 1999; Hall & Willmott, 2000; Molleman et al., 2005; DeVries et al., 2009). Most puddling observations have been made in Lepidoptera (various butterfly and moth species), but they include other insect orders. Among the Hymenoptera there are some records of honey bees puddling (Butler, 1940) and feeding on sweat or tears is common among sweat bees (Halictidae) and stingless bees (Butler, 1940, Bänziger et al., 2009). Ants may also puddle as they readily visited salt baits (Kaspari et al., 2008). Some hemipterans are also known to puddle (Adler, 1982; Rakitov et al., 2005).
Flies (Diptera) are among the most common groups found on excrements, carrion, sweat and mammalian eyes (Bänziger et al., 2009), but this is rarely considered puddling, even though such substrates may only be primary resources for a subset of fly species. For example, tephritid fruit-flies feed on bird droppings, but fruits, yeasts, honey dew, and substances grazed from the surface of leaves are probably their main resources (Hendrichs et al., 1991, 1993). Shen et al. (2009) recently published a paper on a urine puddling locust and included an extensive literature review. They point out that this is the first published record of puddling in an insect with chewing mouthparts. During field-work in Uganda, I have encountered orthopterans on rotting fish baits (mainly catydids) and pictured locusts feeding on the carcass of a snake (Figure 1A & C), therefore, puddling is probably not exceptional in Orthoptera. In addition, locusts are known to regulate sodium intake (Trumper & Simpson, 1993) and crickets can cannibalize for salt (Simpson et al., 2006). Other hemi-metabolous arthropods with chewing mouthparts that puddle include millipedes that may typically feed on mosses (Figure 1A & B) and cockroaches (Schal & Bell, 1982). Note that the concept of puddling is concerned with supplementary feeding: insects that feed exclusively on dung or carrion, such as dung beetles and some species of Trigona bees (Roubik, 1982), are not considered ‘puddlers’.
Figure 1. Examples of puddlers from a variety of taxa in Kibale National Park, Uganda. (A) Millipede and locust near dead snake; (B) millipede with many mites grooming after feeding on dead snake; (C) Charaxes tiridates (Cramer) male and locust feeding on dead snake; (D) Charaxes etesipe (Godart) male feeding on bird dropping; (E) Cymothoe herminia Grose-Smith female mud-puddling; (F) Papilio butterflies (P. phorcas Cramer and P. bromius Stoll) feeding on the wounded tail of a puff adder that was still alive. These images show for the first time that millipedes that normally feed on plant materials such as mosses, and Orthoptera can puddle on carrion. They also illustrate that (1) visiting puddling substrates comes at the risk of picking up potentially harmful mites and pathogens (possibly avoided by the Papilionidae that are not attracted to rotten meat), (2) in some butterfly species females puddle, and (3) dry mud can be used as a puddling substrate.
Download figure to PowerPoint
What substances are sought?
To understand the function of puddling behavior, we need to know which substances are sought and how these affect fitness. Puddling substrates include moist ground, dry mud/concrete (Figure 1E), excrements (of predatory vertebrates or birds), carrion, sea, perspiration, and tears, all of which are typically enriched in salts and some of which can also be sources of nitrogen/amino acids. Several studies have addressed what substances are sought, and sodium appears the most common (Poulton, 1917; Arms et al., 1974; Barrows, 1974; Fraser, 1985; Smedley & Eisner, 1995; Beck et al., 1999; Boggs & Dau, 2004; Molleman et al., 2005; Bänziger, 2007; Shen et al., 2009). However, nitrogenous compounds have also elicited puddling in particular taxa (Schal & Bell, 1982; Beck et al., 1999; Boggs & Dau, 2004; Bänziger, 2007; Shen et al., 2009). Negative results with nitrogenous compounds (e.g., Molleman et al., 2005) are difficult to interpret because the particular form in which the nitrogen is offered matters. For example, butterflies that do not respond to amino acids may respond to proteins (Boggs & Dau, 2004), and in the yellow-spined bamboo locust feeding is elicited by NH4Cl and NH4HCO3, but not by CO(NH2)2 (Shen et al., 2009).
Sex and stage differentials in puddling behavior
Probably the largest void in our understanding of puddling is how the substances from puddles affect fitness. The data on the yellow-spined bamboo locust may provide some leads: puddling was most prevalent in adults and occurred infrequently in the last immature stage but not in other stages, and it occurred in both sexes (Shen et al., 2009). Puddling in Lepidoptera is also restricted to adults, but this could be explained by the low mobility of caterpillars that makes it difficult for them to move to puddling substrates and would also put them at greater risk as they are unable to quickly move away. In contrast to the locusts, in Lepidoptera often exclusively males puddle. In most cases this can be explained by males providing the females with sodium in the spermatophore (Sculley & Boggs, 1996; Smedley & Eisner, 1996; Molleman et al., 2005; Watanabe & Kamikubo, 2005) but not all (Molleman et al., 2005). This is corroborated by puddling in females in butterfly species where spermatophores are not sodium enriched (Scriber, 2002; Molleman et al., 2005). Nuptial gift giving may also explain why in some cases mainly older males puddle: these may have depleted their larval stores of nutrients for nuptial gifts and need to supplement them by puddling (Boggs & Jackson, 1991). Therefore, I conjecture that both male and female herbivorous insects may benefit from sodium intake. Similarly, nitrogenous compounds can act as nuptial gifts and sex differentials in puddling behavior can also be related to nitrogen transfers from males to females (Boggs & Gilbert, 1979; Boggs, 1981, 1990; Marshall, 1982; Schal & Bell, 1982). The reason why puddling behavior is concentrated in the adult stage could then be sought in processes that are unique to adults such as flight and reproduction.
How does sodium intake affect fitness?
Even though there is detailed information on the role of sodium in the digestive, excretory, and neuromuscular systems of insects (Nicolson, 1976; Dow, 1992; Zeiske, 1992; McLean & Caveney, 1993; Peyronnet et al., 2000), no physiological studies appear to have addressed how additional sodium intake would affect fitness of herbivorous insects. Arms et al. (1974) suggested that sodium is needed for neuromuscular activity and, therefore, male butterflies, that are the most active flyers, puddle. Hall & Willmott (2000) found that in riodinid butterflies puddling is concentrated in species with relatively small wings compared to their bodies, which is interpreted as ‘high levels of neuromuscular activity’. However, our study on fruit-feeding butterflies in Uganda did not corroborate this trend (Molleman et al., 2005). The fact that ants used sodium baits even though they do not fly (Kaspari et al., 2008) also suggests that the neuromuscular activity hypothesis may not be the explanation. Other possible explanations for puddling for sodium are that it may enhance sperm motility (Ciereszko et al., 2001), or facilitate amino acid uptake in the insect gut (McLean & Caveney, 1993), and thus make more amino acids available for growth and reproduction.
Some positive effects of sodium intake on reproduction have been documented in herbivorous insects. Most notably, in the skipper Thymelicus lineola (Ochsenheimer) larvae from eggs that were provisioned with extra sodium (provided to the females by males that had the opportunity to puddle) survived better (Pivnick & McNeil, 1987). Other examples are aphid populations that grew faster on sodium treated foliage (Braun & Fluckiger, 1984), and that ant population densities can be predicted from soil sodium content (Kaspari et al., 2008). However, Japanese beetles (not among the puddling insects) that were provided extra sodium laid fewer eggs and lived shorter (Stamp & Harmon, 1991), whereas we detected no significant effect of male sodium diet on female egg-production and fertility in the butterfly Bicyclus anynana (Butler) (Molleman et al., 2004).
How does nitrogen intake affect fitness?
There is much evidence that reproduction in herbivorous insects is often limited by lack of sources of amino acids in their diet (White, 1978; Dunlap-Pianka, 1979; Mevi-Schutz & Erhardt, 2005; Geister et al., 2008), and therefore, we may extrapolate that puddling for nitrogenous compounds can directly enhance fecundity. This has convincingly been shown in the cockroach Xestoblatta hamata (Giglio-Tos) where males forage for urates in bird and reptilian droppings prior to mating. Females then acquire these urates by actively feeding on male uricose gland secretions following copulation (Schal & Bell, 1982). This urate most likely increases female reproductive success (Nalepa, 1994). It is then tempting to hypothesize that Lepidoptera that appear to be especially keen on feeding on bird droppings, e.g., certain groups of skipper butterflies (DeVries et al., 2008, 2009) and Charaxes etesipe (Godart) (Figure 1D), also use urates as nitrogen sources. A small study using stable isotopes, however, did not detect usage of animal derived nitrogen in the bodies of butterflies that can be expected to have fed on carrion or predator excrement (Molleman & Midgley, 2009). On the other hand, Arms et al. (1974) demonstrated that labeled amino acids in mud can reach the eggs of butterflies that mated with males that puddled on it. Female tephritid fruit flies have shown elevated rates of reproduction when offered bird dung in combination with fruit, but did not benefit when urates were offered separately (Hendrichs et al., 1991, 1993). Apart from yolking eggs, nitrogenous compounds derived from puddling substrates may be used to maintain or grow (flight) muscles, and thus in many insect species, enhance male competitive ability (Anholt et al., 1991; Kirkton & Schultz, 2001).
Costs of puddling
Although it is not yet clear how insects benefit from puddling, the costs associated with this behavior have rarely been investigated. There is one study showing that butterflies have higher chances of surviving bird attacks when puddling in larger groups (Burger & Gochfeld, 2001). Butterflies feeding on carrion or dung are usually very focused on their feeding and may be easy prey (they can readily be picked up by human observers), however, predation at these substrates has not been reported. Possibly, the foul juices that these insects imbibed make them temporarily unpalatable to most predators. That potential pathogens are contacted during puddling is evident from the fact that eye-frequenting insects can transmit diseases among humans (Bänziger et al., 2009). Studies of microbes (Zaspel & Hoy, 2008), mites and nematodes associated with puddling insects are needed. Pathogen infection at puddling substrates might also be inferred from Figure 1B, in which a millipede is grooming itself extensively and many mites can be seen on its body after it fed on the carcass of a snake. Studies of microbes (Zaspel & Hoy, 2008), mites and nematodes associated with puddling insects are needed. Furthermore, the digestive system of those puddlers that use carrion or dung probably needs to be capable of neutralizing microbes and their products (Roubik 1989). This capacity is probably not present in swallowtail butterflies: I have never caught any on carrion or dung even though they feed readily on fresh meat (Figure 1F).
- Top of page
Puddling is a widespread phenomenon among herbivorous arthropods (with the notable exception of beetles), but it is still largely unclear which substances are sought and to what extent it constitutes supplemental feeding, making it a somewhat problematic concept. Sodium and more rarely nitrogen have been shown to elicit puddling behavior, but it is hard to exclude various other substances, and the response to these two components may have evolved through their association with each other or with other nutrients.
Even though the natural history information on puddling and our understanding of the nutritional ecology of insects is still too haphazard to identify clear trends, some possible relations can be speculated upon. The information suggests that insects with low-nitrogen diets (e.g., detritivorous cockroaches) puddle for nitrogen, whereas those with low-sodium diets (e.g., Lepidoptera with folivorous larvae) puddle for sodium. The study of puddling could, therefore, shed more light on which nutrients are limiting to insects. Puddling may also be concentrated in species that have the mobility needed to locate widely dispersed puddling substrates, and are non-territorial (Molleman et al., 2005). Species that are well protected from predators and have long development times (e.g., stem borers or armored beetles) may not puddle because they can take time to gather all needed nutrients from their regular substrates. Species with short adult life spans may not have enough time to puddle, but in Lepidoptera, puddling does not seem to be associated with long life spans (Beck & Fiedler, 2009) even though in some species mainly older males puddle (Boggs & Jackson, 1991).
For a better understanding of the fitness effects of puddling behavior, we need to investigate whether and how puddling is enhancing neuromuscular activity (suggested in Arms et al. (1974) but still not proven), reproduction, or growth in a wide variety of herbivorous insects that normally puddle. Physiological work suggests that it is possible that sodium aids amino acid uptake (McLean & Caveney, 1993). Therefore, puddling for sodium may lead to enhanced fitness through a higher availability of amino acids. Sodium facilitating absorption of amino acids provides a further example of post-ingestive nutrient-nutrient interactions that can be modeled using the ‘geometric framework’ in the same way as performed in other contexts in previous studies (Simpson & Raubenheimer, 2001; Raubenheimer & Simpson, 2003, Raubenheimer & Simpson, 2004; Lee et al., 2004; Raubenheimer et al., 2009).
Apart from the physiological factors, ecological costs of puddling are potentially important. For a better understanding of the puddling phenomenon, integration over multiple levels is necessary (Sculley & Boggs, 1996; Molleman et al., 2005; Boggs, 2009; Raubenheimer & Boggs, 2009). It is particularly lamentable that those studying the role of sodium in insect physiology did not refer to puddling or attempt to understand how additional sodium intake would affect insect fitness, and that field biologists who often cite the neuromuscular activity hypothesis have not published results of experiments that rigorously test this idea.