Willows, midges, and moose

connections between tiny insects and big herbivores

The many species of willow are subject to chewing, nibbling, gnawing, and poking by a huge variety of consumers. Here are just of few of the complex interactions.

Most of us here have seen the ‘willow roses’ or rosettes that develop on the twigs or shoots of certain species of willows. The rosettes are galls, induced when a certain tiny fly called a midge lays eggs on the tip of the shoot. The normal elongation of the shoot is suppressed but leaves continue to develop and become crowded together, forming the rosette. The midge’s larva develops inside the rosette, feeding on the bases of the innermost leaves. The larva pupates inside the rosette gall, and the adult emerges the following spring, in time to lay eggs before leaves develop.

The rosette is formed of more leaves than would occur on normal shoots, perhaps forming a wall of defense against enemies of the midge larva (such as parasitoid wasps that would lay eggs on the larva). The inner portions of the rosette also have less photosynthetic capacity and more defensive compounds than the outer portions, which may deter parasitoids and pathogens. The midge larva is presumably is physiologically capable of dealing with the defensive compounds. However, I’ve not found out how well these deterrents work against such enemies. I’ve read that European titmice know how to open the rosettes to gobble up the larva; so of course I now wonder if our chickadees can do the same.

Photo by Bob Armstrong

Female midges are quite choosy about where to lay their eggs. Only some species of willow are susceptible to attack by this gall-forming midge; Barclay and Sitka willow are among them here. Experiments in other regions have shown that individual plants of the same species differ genetically in their susceptibility to these gall midges. And I have observed that rosette galls seem to be more common on shoots that are not shaded.

The rosette-bearing, stunted shoots cannot produce catkins, so the reproductive capacity of the willow plant is reduced in proportion to the number of rosettes. Eventually, the rosette kills the shoot, without apparently affecting neighboring twigs. The rosette, however, offers winter protection for spiders and beetles that shelter among the crowded leaves.

Willows are often heavily browsed by snowshoe hares, moose, and reindeer, and this activity can affect the abundance of various kinds of galling insects on the plant. Several studies have shown that some galls can be more abundant on heavily browsed stems. Unfortunately, I have found no such information for the rosette-forming midge specifically.

However, there is evidence for the reciprocal interaction: moose browsing is affected by the presence of rosette galls. Experiments with captive moose in British Columbia showed that moose clearly preferred to eat willow shoots that bore no rosettes. Although they sometimes bit a shoot with a gall, they soon spit out the rosette.

In the absence of rosette galls, browsing by mammalian herbivores, such as moose and hares, can have significant effects on willow growth and reproduction (by removing stems that would bear catkins). Some studies have shown that severe browsing, which leaves little more than a stump, leads to the production of so-called juvenile shoots and leaves. These often have a somewhat different shape from normal leaves and commonly have more defensive compounds, which reduce palatability and nutritional value; this protects the new shoot from further browsing, at least for a year or two. Moose and hares tend to avoid browsing twigs with lots of those defensive compounds.

However, moderate browsing may have very different effects: One study showed that winter browsing by hares on feltleaf willow twigs led to bigger, more nutritious leaves the following spring. In other cases, moderate browsing has elicited compensatory growth of the willow, but this is not feasible in habitats with low nutrient availability and poor growing conditions. The bottom line here is that the interaction between herbivorous mammals and willows varies a lot, depending on severity of browsing, growing conditions, the species of willow, and no doubt many other factors.

It is clear, at least, that herbivores selectively forage on different species of willow; even within a single species of willow, some plants are more palatable than others. Some such differences are genetic, while others have to do with growing conditions, such as the amount of shade. In either case, selective removal of favored kinds of leaves and twigs makes them unavailable for decomposers below the plants. Heavy browsing obviously reduces the amount of litter fall and can change the availability of soil nutrients that result from decomposition. So moose browsing can affect the soils, leading to changes in plant species composition and, potentially, the course of early plant succession below the browsed shrubs.

Visiting Gustavus

observations on geese, raptors, otters, and the effects of moose browse

One of the first things I noticed was a roadside muskeg in which the shore pines were in sorry shape, with many brown needles. Upon inquiry, I was told that these pines are infected by a fungus that causes needle blight. The fungus kills off mostly the older needles, so an affected pine has only small tufts of newer needles near the branch ends. A serious level of infection can kill the tree. Cool, wet weather favors the production of spores, which are spread by wind or rain-splash, and the spread of this disease. And that’s the sort of weather that has prevailed this spring, it seems.

A flock of several hundred white-fronted geese loafed peacefully in someone’s back yard, while an off-shoot group foraged nervously in a nearby field. These birds were on their way to the nesting ground on the tundra up north. They were enjoying an important way-station on their journey, where they can feed and rest.

A pair of Canada geese seemed to own a piece of pasture, apparently unruffled by grazing horses. But they reacted sharply when a little flock of other geese came in to graze. The small group of new arrivals was a mixed lot—mostly white-fronts, a couple of Canadas, and a sole snow goose. The residents paced back and forth, talking at the unwelcome visitors, who went right on grazing but did not venture much farther into the pasture.

The great outwash plain created by melting glaciers supports numerous stands of willows, most of which have been severely browsed by moose, such that they seldom exceed three or four feet in height. ADF&G has a long-term study of moose ecology and the effects of moose browsing, focused on a set of exclosures that prevent moose access. The willows inside the exclosures look markedly different from those outside: the inside willows are much taller and produce more catkins. The paucity of catkins on the heavily browsed willows may have extended effects, well beyond the diminished reproductive capacity of the willows themselves. Willow catkins are an important source of food (pollen, nectar) for queen bumblebees, which mate in the fall and overwinter in small burrows. In spring, they need to feed so they can produce broods of workers. Bumblebees are important pollinators of blueberries, beach pea, lupine, and many other flowers. So the obvious question is: Does moose browsing significantly reduce the spring food of queen bees, thus reducing their ability to produce healthy broods, and thus impairing the pollination success of various flowers?

Photo by Bob Armstrong

There may be other ripple effects of moose browsing. For example: Browsed willows produce fewer leaves and therefore less leaf litter below the shrubs. So we can ask if the reduced leaf litter affects the mosses, lichens, grasses, and herbs that constitute the ground cover. A lower number of leaves on the browsed shrubs should reduce their growth rates, because fewer leaves mean a reduced capacity to synthesize carbohydrates that provide energy. And one has to wonder if the cropped-off willows offer fewer nest sites and foraging surfaces for birds, leading to a lower density of birds than would occur in unbrowsed stands.

A male harrier coursed up and down the deeply incised channel of a tidally-influenced river that runs across the outwash plain; he was probably hoping to snag a shorebird or two. I heard snipe, high in the air, doing their flight display: as they swoop around, the rush of air over the spread-out tail feathers is modulated by the beating of the wings, producing a characteristic sound known as ‘winnowing’. Snipe winnow both on migration and on the nesting grounds, and most of these were probably still in migratory mode. I also think I heard a pygmy owl, tu-tu-tu-ing in the trees. A male robin carried a beakful of worms to his chicks, proving that the robins thought it was spring despite the raw, cold weather.

Another treat was the opportunity to watch three river otters exploring one of the inland ponds near Bartlett Cove. I suspect this was a family of a female with two of last year’s offspring, although there was not much difference in size. Otter families often stay together over a winter and into the next spring. These otters were well aware of our presence and kept bobbing up to peer at us. I was told that river otters are often seen as they travel considerable overland distances in Gustavus, visiting inland ponds (are there fish to be caught??) and returning to the ocean shores.