3. Biodiversity
The key to improving biodiversity in streams and riparian zones is habitat diversity and connectivity with other habitats. The greatest improvements in habitat diversity are likely to occur when riparian management involves planted trees or remnant forest.
Riparian planting effects on stream habitat for aquatic biota include:
- provision of woody debris as trees fall into streams over the long term, providing habitat diversity and cover for aquatic invertebrates and fish;
- increased shade and provision of terrestrial food sources (fallen leaves etc.) as riparian plants grow so that levels of instream productivity and trophic pathways resemble the natural state;
- reduced erosion and inputs of fine sediment from (1) exclusion of livestock, leading to an improvement in streambed and bank habitat and (2) interception of hillslope sediment over the long term, and (3) tree roots that stabilise the stream banks and provide habitat;
- reduced water temperatures if sufficient lengths of upstream shade exist, and lower air temperatures and humidities, and less wind exposure, in the riparian zone where the adult stages of some aquatic insects spend part of their lives and some native fish lay their eggs (banded kokopu, short-jawed kokopu).
Lack of stream shade appeared to be the most important factor affecting invertebrate populations in Waikato hill-country streams (Quinn et al. 1997). Quinn et al. (1997) concluded that shade effects on algal biomass were a major cause of the lower abundance of some invertebrate groups, notably midge larvae, in some Waikato forested streams. Reduced water temperatures can also be expected with riparian planting, particularly if the planted buffer zones extend over several hundred metres of shallow stream systems. Many New Zealand stream invertebrates (e.g., mayflies, stoneflies) are sensitive to water temperatures >20ºC, temperatures that are commonly exceeded in open pasture streams. Rutherford et al. (1999, 2000) used computer models to show how high water temperatures can release periphyton from control by temperature-sensitive invertebrates, like mayflies, resulting in algal proliferations.
Quinn et al. (1997) found that ‘stream health’, as indicated by invertebrate communities, was similar in pine plantation streams to that in native streams (and very different from the pasture streams) in the Hakarimata Range – despite the sedimentation and turbidity in the pine plantation streams from bank erosion. This suggests that shading benefits outweigh the sedimentation side-effects associated with channel widening. The reduced inputs of fine suspended sediment expected over the long term following bank stabilisation may also improve conditions for migrating fish such as banded kokopu whose juvenile migrations are adversely affected when turbidity increases above 25 NTU (Richardson et al. 2001).
Riparian trees add leaf litter and wood that are an important source of habitat diversity for invertebrates and fish, particularly in silt-bed streams. Recent work has demonstrated that stable bank habitat and the presence of riparian tree roots penetrating into those banks creates habitat for freshwater crayfish (Parkyn & Collier in press). Field investigations of Auckland stream plantings aged from 10-30 years showed that woody debris from fallen branches, wind damage to plants, and unsuccessful plantings had begun to accumulate in small stream channels (pers. obs.).
Furthermore there has been increasing recognition recently of the role of riparian vegetation in creating suitable microclimate conditions for the adult stages of some stream insects. Collier & Smith (2000) reported that 50% of female stonefly adults died within 4 days at constant air temperatures of 22-23ºC. These temperatures were exceeded 25% of the time in January next to a Waikato pasture stream. Davies-Colley et al. (2000) found that at least 40 m of forest habitat next to pasture was required before air temperatures became comparable to those in a large block of native forest in the Waikato. However, narrower buffer zones can give significant temperature control. Air temperatures measured in a clear cut pine plantation within a 5 m buffer of well-established native vegetation on one side of a stream were similar to those in a 30 m buffer on the other side of a stream (John Quinn, pers. comm.). Daily maximum temperatures during summer were reduced from about 30ºC in the clear cut area to 25ºC in the buffer zones.
3.1 Effectiveness of riparian management for habitat and biological diversity
Parkyn et al. (2003) studied a number of riparian restoration schemes in the Waikato region to determine whether riparian management was achieving improvements in stream health. The sites were grouped according to the stream substrate or land topography, e.g., cobble/gravel substrate, lowland (silty substrate), pumice substrate. The buffer zones had been fenced to exclude stock and tree species had been planted (or remnant vegetation was present). The age of planting ranged from ‘recent’ (c. 2 years) to “mature” (>20 years) within each substrate/hydrological grouping. Each buffer zone was compared to an unfenced and actively grazed stream section upstream of the buffer zone or in a neighbouring stream when no upstream control was available. In general, streams in buffer zones showed rapid improvements in clarity, bank stability, and nutrient contamination. Often channel widths decreased in buffered reaches where the plantings were young, presumably from a reduction in trampling by stock.
However, significant changes to macroinvertebrate communities towards “clean water” or “native” communities did not occur at most of the sites over the time-scales that were measured in this study. The lack of improvement in QMCI scores and taxa richness may indicate (1) a lack of source areas of colonists, (2) lack of suitable microclimate for adult invertebrates, (3) time-scales of recovery are large, or (4) that buffers were not achieving habitat goals. However, one stream with a wide buffer of >50 m, 25 year old plantings, and the whole stream length planted did show significant improvement in invertebrate communities compared to a nearby pasture stream. Improvement in invertebrate communities appeared to be most strongly linked to decreases in temperature suggesting that restoration of in-stream communities would only occur after canopy closure and after protection of headwater tributaries. This was particularly evident in lowland streams where catchment influences had a greater impact than local riparian influences.
Quinn et al. (2004) studied the effect of native forest buffers within plantation forestry on stream invertebrate communities in the Coromandel Peninsula, New Zealand. Clearcut reaches had the lowest diversity and taxon richness of 28 stream sites, while sites that had been logged leaving continuous buffers did not differ from those in intact native or mature plantation forest, indicating that buffers greatly reduced disturbance associated with logging. Logging impacts were strongly related to increases in periphyton biomass, water temperature, fine sediment, and channel instability.
In North American streams, Weigel et al. (2000) found that the macroinvertebrate community response suggested higher organic pollution in continuously grazed sections compared to woody buffered sections. However, they also found that catchment differences produced greater overall differences in the invertebrate communities than between different grazing treatments along the same stream. This variability between streams is a common problem with interpretation of riparian buffer zone studies, and can mean that the same management technique can have variable outcomes in different stream systems (Belsky et al. 1999). Sovell et al. (2000) found that faecal coliforms and turbidity were greater at continuously grazed stream sections than at rotationally grazed sites. However, they were unable to show associated changes to the macroinvertebrate or fish communities.
Biodiversity in streams with riparian plantings may be affected by being in a “transitional” state. As plantings mature and shade is introduced and water quality changes occur, some species characteristic of pasture streams may be lost, while there is a time lag before the buffer zone matures or until connectivity of riparian patches allows recolonisation of native forest species to occur. Scarsbrook & Halliday (1999) found that aquatic invertebrate community composition had recovered fully within native forest patches that were about 60 years old, in hill-country pastoral catchments in the Waikato.
Collier et al. (2000) found that streams draining catchments entirely in pasture or native forest had similar percentages of total taxa. Biodiversity of open, pasture streams can equal that of shaded, forest streams at the reach scale. However, if you stand back and look at a larger catchment scale, biodiversity of the entire stream system would have been reduced. Deforestation of headwater streams has enabled species characteristic of more open conditions, which would have been present in the lower, wider reaches of stream systems where canopy closure was not possible, to establish themselves further upstream. Therefore the reduction in habitat diversity over the whole stream system has led to a homogenising of species diversity.
3.2 Habitat in lowland streams
A potential problem associated with riparian plantings shading out macrophytes in soft-bottomed lowland streams is that these macrophytes (particularly submergent species) can provide important stable substrates for invertebrate colonisation at certain times of the year (Collier 1995, Collier et al. 1999) and increase habitat heterogeneity through their influence on water velocities (Champion & Tanner 2000). The highest number of invertebrate taxa in a lowland stream south of Auckland was found in macrophyte patches with intermediate biomass leading to the recommendation that patchy shade conditions should be maintained in soft-bottomed streams to enable moderate quantities of submerged macrophytes to grow (Collier et al. 1999; Champion & Tanner 2000). In many lowland streams submerged wood can also provide an important stable habitat for invertebrates (Collier et al. 1998), but riparian plantings would not be expected to contribute considerable amounts of woody debris to streams for many years after planting. However, growth of trees large enough to shade lowland streams will also take some time resulting in low levels of shading for many years, and fallen branches and failed plantings or even plantings lost once channel widening has begun will accumulate in the streams, particularly once early successional trees become mature (e.g., manuka).
3.3 Buffer Width
In Australia, Davies & Nelson (1994) found that small buffers (<10 m wide), retained after forest harvesting, did not significantly protect streams from changes in algal, macroinvertebrate and fish biomass and diversity. Buffer widths of >30 m appeared to provide protection from short-term impacts in a variety of forest types and geomorphology. However stream temperatures were only increased when buffer widths were below 10 m. The buffer width required to decrease stream temperatures may be less than that required to provide a microclimate similar to forested conditions. A single line of trees can provide about 80% shade to streams when the trees have grown tall enough to achieve canopy closure (Collier et al. 1995). Five and 30m wide riparian buffers of native forest reduced the median daily maximum air temperatures by 3.25 and 3.42ºC, respectively compared with a clearcut area downstream of the site (Meleason & Quinn 2004), indicating that narrow buffers can maintain cool riparian air temperatures. The buffer widths of Coromandel forestry sites studied by Quinn et al. (2004) ranged from 8-27 m and supported stream invertebrate communities similar to those in native or mature plantation forest.
In a review of buffer width requirements for wildlife species distribution and diversity in the U.S., Castelle et al. (1994) found that many studies showed improvement in salmon, trout and benthic invertebrate communities with buffers of >30m. However, a number of habitat suitability models in the U.S. found that buffer widths could range between 3 and 107 m depending on the particular resource needs of individual species (Castelle et al. 1994). Brosofske et al. (1997) concluded that a buffer of at least 45 m was necessary to maintain a natural riparian microclimate after harvesting of Douglas fir and western hemlock. International studies of buffer width requirements for biodiversity may therefore be of limited value to New Zealand, as individual species will most likely differ in their requirements.
Contact for Enquiries
MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND
Fax: +64 4 894 0721
Contact this person
