Feature

Of Deer and Detritivores

Browsing mammals in native forests may affect not just plants but also soil invertebrates and the ecosystem services they provide.

David A. Wardle, Gary M. Barker, Gregor W. Yeates and Karen I. Bonner

Several species of browsing mammals, including deer and goats, were introduced to New Zealand in the late 19th and early 20th century, and have subsequently spread throughout most of the forested landscape of New Zealand.

In the absence of the population regulation mechanisms that operated in their natural environment, their numbers rapidly increased to plague proportions and soon exacted enormous damage on a vegetation understory that was ill-equipped to deal with high levels of browsing mammals. (Although moas did browse forest vegetation until the 1400's, little is known about the effects that they exerted in the forest, other than that the habit and intensity of their browsing was probably quite different to that of the introduced browsers.)

While hunting, live capture and other control measures have reduced deer and goat densities in several areas in the past three or so decades, there are still many areas throughout New Zealand in which their populations are sufficient to maintain a large impact on vegetation composition.

The effects of introduced browsing mammals on vegetation are most effectively demonstrated through the use of "exclosure plots". These are fenced areas of vegetation in the forest, usually 20 x 20 m, which were established by the former New Zealand Forest Service (NZFS) mainly in the 1950s-1980s to keep out browsing mammals. With the dissolution of the Forest Service in the late 1980s, maintenance of the vast majority of these plots was discontinued and many are no longer effective at excluding browsing mammals, although it is likely that over 100 still remain in adequate condition.

Recent renewed interest in introduced animal impacts by the Department of Conservation and others, has seen several of these exclosures restored and new ones established. In many forest types, particularly those of high fertility, the effects of browsing on vegetation are immediately apparent -- inside the exclosures there is often a dense, highly diverse community of large-leaved palatable shrubs, while outside the exclosure the understory is either extremely sparse or dominated with unpalatable ferns, shrubs and monocotyledonous species.

Underground Effects

Although browsing mammal effects on vegetation are readily visible, much remains unknown about how browsing mammals affect the "unseen" components of forested ecosystems -- those organisms which occur below ground and which are responsible for maintaining the decomposition pathways and nutrient cycling processes required for sustaining the forest in the long term.

Three main mechanisms have been proposed to occur, all with different implications for the soil biota and the processes that it regulates.

Firstly, browsing animals return organic matter and nutrients to the soil as dung and urine, which are in highly labile forms and are more readily accessible to both soil organisms and plants than are other sources of nutrients in the forest such as dead twigs, leaves and roots. This has positive effects for soil processes.

Secondly, browsing animals can induce physiological effects in plants, both by increasing allocation of carbon and nutrient to roots (with positive effects on soil processes) and by inducing increases in concentrations of defense compounds such as phenolics in leaves (with negative effects on soil processes).

Thirdly, browsing mammals will preferentially feed upon palatable plant species, causing their reduction, and therefore favouring domination of the forest understory by unpalatable species. Those plant species which produce unpalatable foliage also tend to produce litter which is less favourable to soil organisms, retarding soil processes.

These three mechanisms all have different implications for belowground organisms. While we know little about how introduced browsing mammals affects native belowground invertebrates, we would expect the nature of the response to depend upon the relative importance of each of the three mechanisms.

Invertebrate Responses

Our work has focused on 30 long-term fenced exclosure plots established between 1961 and 1985, which represent most of the major forest types throughout New Zealand, ranging from Waipoua Forest in Northland to Stewart Island, and from near sea level to just below treeline. We made a number of assessments, both above- and below-ground, and both inside and outside each exclosure plot.

The vegetation measurements all showed predictable responses; in most locations vegetation density was significantly reduced, and browsing often favoured dominance of unpalatable plant types at the expense of palatable types. However, these unidirectional effects above ground were not necessarily matched by corresponding effects belowground. The nature of effects of browsing on soil fauna was very much dependent on the body size of the soil animal group under consideration.

For the soil microfauna (animals with body width of .2 mm, e.g. nematodes, tardigrades, rotifers, copepods), and for the microorganisms upon which they feed, there were no consistent trends across the 30 locations. Although browsing mammals had strong indirect effects on populations of these organisms at several locations, as many positive as negative effects were detected. These idiosyncratic patterns suggests that the relative importance of the three mechanisms by which browsing may affect soil organisms may differ across locations and forest types.

In contrast, nearly all larger soil organisms were adversely affected by browsing; this included all the main components of the mesofauna (body width of 0.2 - 2.0 mm, e.g. mites, springtails) and macrofauna (body width of 2.0 mm, e.g. insects, gastropods, spiders, harvestmen, millipedes, centipedes, amphipods, isopods). Frequently the magnitude of these reductions was very large; it was not uncommon for browsing mammals to reduce densities of these fauna by over 70%.

The mechanism by which browsers reduce populations of these invertebrates is unlikely to include shifts in resource quality because these fauna were often influenced in sites in which resource quality was not detectably affected. Rather this is more likely to reflect disturbance effects of browsing mammals on the soil microhabitat through scuffing and trampling. The microfauna, which live in the aqueous pores of dead leaves and humus, are less likely than the mesofauna and macrofauna to be affected.

Across the 30 locations, we found significant reductions of vegetation diversity in the browse layer outside the exclosures relative to inside; on average browsing reduced species richness by 41%. These reductions were also matched by reductions in the species richness of litter-dwelling gastropods and the number of families of millipedes present. This suggests that the diversity of at least some soil faunal groups is adversely affected by browsing mammals at least at the spatial scale at which we were operating. Richness of other soil faunal groups, such as nematodes and beetles, were not consistently affected by browsing, but in several instances the community composition of these groups was still significantly influenced.

Ecosystem Considerations

Below-ground ecosystem processes such as decomposition, soil carbon dioxide release and release of nutrients required for plant growth are performed primarily by bacteria and fungi, but their activity is profoundly affected by the soil animals which feed upon them and improve their physical environment. Although browsing mammals affected soil CO₂ release for relatively few locations, there were frequently strong effects of browsing on soil storage of both carbon and nitrogen. However, roughly equal numbers of locations showed strong positive and negative effects of browsing on carbon and nitrogen sequestration.

Our study provides clear evidence that the magnitude of effects observed belowground does not necessarily match those observed aboveground. There were some instances, particularly in nutrient poor sites dominated by such species as kauri or mountain beech, in which there were no detectable effects aboveground but very strong effects on many components of the soil invertebrate fauna. This suggests that even low levels of browsing mammals can sometimes exert severe effects on the decomposer subsystem, and that what we see in terms of effects aboveground may not give a reliable indication on the extent of effects on the "hidden" components of the ecosystem.

New Zealand's native forests also host a wide range of other introduced alien organisms, including such pervasive pests as possums, wasps, vertebrate predators and weeds. We are becoming increasingly aware of the damage that these aliens exert on the "visible" components of ecosystems, such as plants and bird life.

However, the belowground component of natural ecosystems, frequently ignored in conservation assessment, is of fundamental importance on two fronts.

Firstly, most terrestrial biodiversity occurs belowground, not above it. Secondly, the soil biota carries out the decomposition and nutrient release processes upon which the aboveground component is ultimately dependent.

Much remains to be discovered about how introduced alien organisms affect both the biodiversity of the belowground subsystem and the essential ecosystem services that it provides.

Gary M. Barker is a scientist with Landcare Research.
Karen I. Bonner is a scientist with Landcare Research.
David A. Wardle is a scientist with Landcare Research.
Gregor W. Yeates is a scientist with Landcare Research.