A Buzz in the Meadow

by Dave Goulson

  There is another way in which we can increase colonisation of nature reserves and so boost the number of species they are likely to have, and this is by providing habitat corridors along which animals can move. I have already mentioned that many animals are likely to be unwilling to cross water, but may be much more likely to cross land, even if it is inhospitable. However, this varies enormously between species. Some birds and butterflies will happily zoom across cereal fields, towns or busy roads to get from one meadow or woodland to another, but many will not. For example, some woodland bird species, such as long-tailed tits and dunnocks, are unwilling to leave their preferred habitat and fly out into the open, and it is these species that are likely to be most badly affected by habitat fragmentation. We can make life much easier for such creatures if we provide links between patches of habitats; in the case of woodland, hedges often provide routes for dispersal for creatures that shun open fields. Tunnels under roads have been used with some success to help creatures as diverse as elephants and badgers to get from one habitat patch to another.

  Other creatures are harder to accommodate. For instance, some butterflies found in flower-rich meadows are enormously sedentary – particularly blue butterflies such as the Adonis and small blue, both of which will often happily live out their entire life in a few square metres of downland, and will rarely (if ever) choose to head off into the unknown on a perilous adventure. One can hardly blame them these days; their chances of finding another suitable patch of habitat are remote, so they are probably wise to stay put. On the other hand, their sedentary lifestyle cuts them off from gene flow – each small population in a habitat fragment becomes entirely cut off from all the others, and this can doom it to eventual extinction.

  The problem with isolation is that it leads to inbreeding. Small populations tend to have less genetic variation than large ones to start with, simply because there are fewer copies of each gene. They also tend to lose genetic diversity over time, through a process known as ‘genetic drift’, whereby rare forms of genes drop out of the population by chance. With less genetic diversity, the population is less able to adapt to any change in the environment. What is more, inbreeding can lead to the expression of rare, harmful, recessive genes. Every individual has a few dud genes; humans have somewhere in the region of 50,000 genes (though probably far fewer that are absolutely vital to our well-being), of which on average about four are likely to be faulty, but because we have two copies of every gene, and as long as at least one copy is okay, we are fine. Fortunately your partner is unlikely to have the same faulty genes that you do, so there is little chance of your offspring getting two non-functioning copies of the same gene, although it does occasionally happen. If they do, they may be severely ill, disabled or even die early in development. However, if you are related to your partner, then you are likely to have some of the same faulty genes. This is not good, for then, for every faulty gene you have in common, there is a one-in-four chance that your offspring will inherit both non-functioning copies. As I have mentioned, in a small population it takes only a few generations before everybody becomes cousins of one another, sharing grandparents or great-grandparents. In lab studies using insects, forced inbreeding over a few generations quickly results in a population of feeble individuals with low fertility and reduced life expectancy, a phenomenon known as ‘inbreeding depression’. Even without this, small populations are much more likely to go extinct than large ones just through bad luck, but when these small populations are also subjected to inbreeding they are unlikely to persist for long.

  Although the effects of inbreeding are well established from lab studies, until fairly recently there were few good examples of inbreeding hampering the survival of insect populations in the wild. The first – and still one of the most elegant – demonstration of the negative effects of inbreeding on insects was provided by Finnish ecologist Ilkka Hanski’s work on populations of the Glanville fritillary, one of the most common butterfly species in my meadow. In Finland, Glanville fritillaries are found only in the Åland archipelago, a cluster of more than 6,000 small islands in the Baltic Sea, halfway between mainland Finland and Sweden. On these islands the Glanville fritillary is widespread and fairly common, occupying dry, grassy meadow areas where its food plant, ribwort plantain, is to be found. The lovely adult butterflies with their orange-and-black chequerboard wings emerge in May and June and lay batches of eggs which, when they hatch into small black spiny caterpillars, live in gregarious clusters within protective webs that they spin among the plantains.

  Hanski’s team (which consists largely of undergraduate students) has repeatedly surveyed about 4,000 of these meadows every year since 1991, recording whether or not the butterfly is present in each patch – an astonishing effort, but presumably involving many happy hours of splashing about in boats to get between the mostly uninhabited islands. Their work has become a classic study in what is known as ‘metapopulation dynamics’, and their findings have a lovely symmetry with MacArthur and Wilson’s theory.

  Hanski’s group has been studying just one species while MacArthur and Wilson were trying to understand what determined the number of species in entire communities, but the processes involved are exactly the same. Hanski found that the majority of the 4,000 meadows had no Glanville fritillaries in any particular year, with only 300–500 occupied ones. Some of these populations are tiny – just a few butterflies. Populations in any particular patch go extinct quite frequently, but these extinctions are roughly balanced by colonisation events whereby butterflies discovered and established a new population in an empty patch. A large regional population composed of lots of small fragmented populations, each of which blink in and out of existence from time to time, is what is known as a ‘metapopulation’.

  Just as small and remote islands have fewer species in them in part because they are less likely to be bumped into by wandering creatures, so Hanski’s habitat patches were less likely to be colonised by butterflies (and so tended to remain unoccupied for longer) if they were small, or a long way from other patches. Similarly, he found that small populations were more likely to go extinct. Extinctions were caused by a variety of events – outbreaks of parasitoid wasps, overgrazing and trampling by cattle, human disturbance, and so on, but all of these were more likely to prove fatal to a local population if it was small. What is more, populations were found to be more likely to go extinct if they lacked genetic diversity – the first time that inbreeding had been shown to be harmful to wild populations of any insect. Populations with low genetic diversity were found to have reduced survival of both caterpillars and adults, and a low hatching rate of eggs, factors that are obviously likely to push a small population over the edge to extinction.

  This finding was followed up by one of Hanski’s students named Marko Nieminen, who created new wild populations from either the offspring of unrelated individuals or a mating between a brother and sister. The latter proved much more likely to die out, mainly because the low egg-hatching rate led to smaller groups of the gregarious caterpillars, and smaller groups are known to have a much-reduced chance of surviving through the winter.

  There are reasons to believe that social insects such as bumblebees might be particularly prone to inbreeding, and this might partly explain why some bumblebee species have fared so poorly in recent years. In butterflies such as the Glanville fritillary – indeed, in most creatures – every adult individual tries to breed, and although quite a few fail, many pass on their genes to the next generation. In bumblebees this is far from the case. The large majority of the population are workers, most of which will not produce any offspring. Save for a few sneaky workers that try (largely unsuccessfully) to produce sons, reproduction is the preserve of the queen. This means that the breeding population – the gene pool – is much smaller than you might at first think. Each nest has just one breeding individual (who also carries inside her the sperm of her mate). So ten nests, which might collectively contain several thousand bumblebees at th
eir peak, actually represent just twenty breeding individuals. A viable population, one that will be big enough that it does not suffer from inbreeding, needs to contain several hundred breeding individuals – let’s say at least 100 nests. We don’t know for sure how big an area of habitat would be needed to support 100 bumblebee nests of any particular species – it would depend on how good the habitat was, and would probably vary between bumblebee species – but it is likely to be in the region of hundreds of hectares. In the UK, indeed in most of western Europe, most nature reserves are smaller than this. For example, most of the patches of flower-rich grassland that survive consist of single meadows, often of no more than a dozen hectares or so, surrounded by arable crops. In other words, the habitat islands that remain are likely to be too small to support a viable bumblebee population. Lots of such small islands, close together, could support a metapopulation just like that of the Glanville fritillaries in the Åland archipelago. But we don’t have lots of patches of flower-rich grassland, so it is likely that fragmented populations of our rarer bees are suffering from inbreeding.

  In bumblebees, inbreeding can lead to some rather peculiar gender-bending side-effects. To understand them requires a little diversion into how sex is determined in bumblebees. In many animals (including ourselves) sex is determined by which sex chromosomes we inherit. If we get an X and a Y, we are male; two Xs and we are female (some animals do it the other way round). In bees, it is quite different; sex is determined by a single gene. If an individual has two different copies of this gene, it is female. If it has two identical copies, or just one copy, it is male. Female bees, like us, have two copies of each chromosome – to use the technical term, they are diploid. In a genetically healthy population there are usually lots of different versions of the sex-determining gene, so the chances are that diploid individuals will have two different copies and thus will be female. Male bees, typically, have just one copy of each chromosome – they are haploid. To produce a son, a female bee has just to lay an unfertilised egg; the haploid gamete develops into a healthy son. This means that male bees have no father. To produce a daughter, the female bee fertilises her egg using sperm from a male; in bumblebees this sperm had been stored inside the queen since she mated the previous summer. So long as the copy of the sex-determining gene in the sperm is different from each of the two different copies held by the mother, then these diploid offspring will all be female.

  The problem arises when the population is small. As I have said, small populations tend to lose genetic diversity through drift, so the number of versions of the sex-determining gene declines over time. Also, individuals soon become related to one another, so the chances that one of the queen’s sex-determining genes and the male’s match becomes more and more likely. If they do, then half of the queen’s diploid offspring will have two identical sex-determining genes. This is a disaster for the nest, for these individuals develop as ‘diploid males’, which do no work for the colony. Essentially half of the colony’s workforce is useless – worse than useless, since they place a burden on resources, eating the food stores and not replacing them. This might not be so bad if the diploid males went out and mated with new queens, but diploid males seem to have lower fertility than normal males, and in any case the nest starts to produce them right at the beginning of the season, when there are no virgin queens around to mate with. With half their workforce sitting around idly scoffing food, we might expect such nests to die out long before other nests start to produce new queens in high summer.

  Oddly enough, this is not how it seems to work out, and we don’t yet know why. The presence of diploid males can be used as a warning sign of inbreeding in bees, and my PhD student Ben Darvill screened populations of some rare species, such as the moss carder bee, to see which populations are becoming inbred. He focused on the Hebrides, another island system where agriculture has changed relatively little and patches of flower-rich grassland are common. On the smaller, more remote islands of the Hebrides such as the Monach Isles, a tiny cluster of beautiful, sandy-shored islands off the west coast of North Uist, he found that diploid males were quite common. The unexpected thing was that some of the remaining bees were triploids – they had three sets of chromosomes.

  Triploids look and act like female bees. They are presumably produced by diploid males successfully mating with a normal, diploid queen, against all the odds. This requires a nest producing diploid males to survive long enough to produce these males in late summer when new queens are on the wing, and then for the diploid males to outcompete normal, healthy haploid males in finding and courting a virgin female, and finally to produce viable (but diploid) sperm, despite their supposed low fertility. On the Monachs one in five of the female bees turned out to be triploids. Clearly diploid males are not quite as hopeless as was thought. Nonetheless, we now have other evidence that the inbreeding that leads to diploid male production also has other harmful side-effects. Small, isolated bumblebee populations have measurably lower genetic diversity than the larger populations, and rare species tend to have much lower genetic diversity than the common ones. What is more, Penelope Whitehorn has found that the more inbred populations suffer from higher loads of the internal parasite Crithidia bombi, presumably a sign of their generally low vigour. We do not yet have a data set to compare with Hanski’s butterfly study, but I would hazard a guess that these small, isolated bumblebee populations, with little genetic diversity, high levels of diploid males and high parasite load, are probably not long for this world. However, if the population on the Monach Isles disappears it would not be a disaster as there are still plenty of bigger, less-isolated habitat patches on the larger Hebridean islands. The Monach Isles might well be recolonised in the future, and we would predict that the moss carder bee population there will continue to blink in and out of existence every few years, just as many of Hanski’s butterfly populations do.

  There are important lessons to be learned from all of this. The survival of the metapopulation as a whole depends upon colonisation events balancing extinctions. If the extinction of populations occurs faster than new ones are created, then the number of occupied patches will slowly decline over time, until eventually there are none. In the Åland islands, where the human population is low and agriculture is not intensive, there are still lots of habitat patches, and many of them are close together. Hence it is easy for unoccupied patches to be recolonised, and so new populations spring up as fast as old ones become extinct. Overall, Glanville fritillaries are doing just fine in the region. Similarly, moss carder bees are doing okay in the Hebrides. But imagine what would happen if a few of the habitat patches are destroyed. The average distance between patches of meadow will then be a little greater, slightly reducing the colonisation rate. If there are fewer meadows in total, there will be fewer occupied patches, and hence fewer sources of colonists. Hence the proportion of the remaining meadows that have butterflies or bumblebees at any one time will be lower. Remove a few more meadows, and it will become lower still. The remaining populations will be few and far between and will begin to become inbred, because they will receive few visitors from outside, and hence extinction rates will begin to climb. Remove too many meadows and you will reach a tipping point, beyond which the number of occupied patches will gently but steadily decline to zero. This is known by biologists as ‘metapopulation collapse’.

  Unfortunately it is probably a very common process. Most natural habitats in Europe are now highly fragmented, and the fragments are often not close together. Hence many populations of sedentary habitat specialists – be they snails, butterflies or bumblebees – are separated from one another, with levels of inbreeding slowly increasing. Many might be doomed to extinction by what we have done in the past, so that even if no further habitats are lost we can expect them to slowly, inexorably disappear. This is sometimes described as an ‘extinction debt’, and in truth we do not know how many species in the world are in this situation.

  Just as it is true that no man i
s an island, so no habitat island is truly isolated. It cannot survive on its own, for it depends on other islands for sources of colonists, for gene flow to keep populations healthy. Chez Nauche is far more diverse that it was when it was a cereal field, but how many more species might have arrived 100 years ago, when there were many flowery meadows nearby? There are no large blue butterflies, or corncrakes, because these species have largely gone from the landscape – there aren’t enough patches left for them to survive. No matter how carefully a nature reserve or habitat patch is managed, it is really only as good as the network of patches of which it is a part.

  What we do know is that it is possible to help. We can create new habitat islands, filling in some of the gaps, making the network of patches a little more connected. We can plant new woodlands, link them with new hedgerows, and regenerate flower-rich grasslands such as the meadow at Chez Nauche. It takes time, and we need lots of them, so all the more reason to start now. There is an old Chinese proverb: ‘The best time to plant a tree was twenty years ago. The second-best time is now…’

  CHAPTER FIFTEEN

  Easter Island

  1 August 2013. Run: 39 mins 38 secs. I was up early this morning; the mist hadn’t cleared as I set off and as I ran up the drive one of the barn owls swooped silently past, neither of us seeing each other until we almost collided – a wonderful, ghostly creature. People: only Monsieur Fontaneau junior, inspecting his cows. Dogs: 6. Butterfly species: 8, including a large tortoiseshell nectaring on a spear thistle. This butterfly is sadly extinct in the UK, wiped out by the demise of the elms – its larval food plant – at the hands of Dutch elm disease.