Acid Rain and Wildlife Modelling
Modelling is an extremely important component of the program because there is no direct means of predicting the nature and extent of aquatic ecosystem recovery from acidification. Modelling enables both scientists and policy makers to utilize existing information to make predictions about the eventual status of ecological components of aquatic ecosystems in eastern Canada under various emission scenarios. With the aid of models, we can estimate responses of waterfowl, loons and their breeding habitats to predicted changes in acidic deposition following SO2 emission reductions and report on the progress of recovery in aquatic ecosystems. With more information, improved biological models will be used to evaluate critical loads to ensure the protection and recovery of sensitive surface waters in eastern Canada.
Can we predict the eventual benefits of SO2 emission controls on the suitability of waterbird breeding habitat in eastern Canada?
Our modelling work has focused on three indices of water bird reproduction: the occurrence of breeding pairs (measures suitability of nesting habitat), the occurrence of broods (measures suitability of brood-rearing habitat), and the survival of common loon two-chick broods (measures specific suitability of brood-rearing habitat for loons, a key indicator species for acid rain) (see Models). Each of these dependent variables is influenced by the physical and chemical characteristics of lakes slightly differently.
We used logistic regression to develop models for predicting these reproductive parameters. The models developed for piscivores (loons and common mergansers) had the best fit and most variation explained. This can be attributed to the fact that pH and lake size have a strong effect on the occurrence of fish, which in turn directly influences the probability of having fish-eating birds nest on the lake. Because of this clear and direct relationship, we used piscivore models to demonstrate effects of emission reductions on aquatic biota, although models for other waterbird guilds (species) have also been developed. In general terms, the ability of habitat to support breeding piscivores depends both on area and pH. Larger and higher pH lakes are more likely to support breeding pairs (IP) and broods than smaller or more acidic lakes. For example, a 10 ha lake with pH 7 has a 50% chance of supporting a breeding pair of piscivores, a 30% chance of supporting a loon brood, and 80% chance of having 2 loon chicks survive to fledging stage (see chart).
To model the effects of SO2 emission reduction scenarios on waterbirds, we used the Waterfowl Acidification Response Modelling Systems (WARMS). WARMS is a computer software package that consists of a series of linked models: it links sulphate (SO4) deposition (secondary watershed level) to lake chemistry (pH or alkalinity), and fish and wildlife models (see chart). Input to WARMS consists of lake files (chemical and physical characteristics), characteristics of secondary watersheds (runoff coefficients, and wet, dry and background sulphur deposition), predicted changes in SO4 deposition (from atmospheric models) based on emission scenarios, and logistic or linear fish and water bird models to estimate effects of changes in acidic deposition on habitat suitability for selected biota, as well as providing eventual pH and alkalinity for those lakes.
Wet Sulphate Deposition
Revised wet SO4 deposition values (kg/ha/yr) were generated from atmospheric receptor site data provided by the Atmospheric Environment Service of Environment Canada (S1: 1982-1986; S2: 1990-1993) (see sites). Interpolated maps for northeastern North America show that levels of SO4 deposition have generally decreased and will continue to do so under the current legislated sulphur emission cutbacks, specifically after Canadian controls (S3: ca. 40% cut by 1994, achieved) and after U.S. controls (S4: ca. 40% cut by 2010) (see chart). Southern Ontario typically receives the highest deposition of SO4, but a broad region of central Ontario and Quebec receives high levels as well. Under each of the emission reduction scenarios, improvements occur in most regions, except that Atlantic Canada receives relatively little benefit until the 2010 emission cutbacks, indicating that much of the Atlantic deposition problem comes from U.S. emissions. Much of southern Ontario and Quebec will continue to receive 15-20 kg/ha/yr SO4, which is below the target load set in the mid-1980s, but which exceeds the critical loads for many sensitive waters in this region.
Predicted pH and Habitat Suitability Change in Eastern Canada
pH - For pH, WARMS modelling results showed that most chemical changes (expressed as % of affected land mass) would occur in areas where current lake pHs are between 5 to 6 (see chart); understandably, no change would be expected for lakes above pH 7, and minimal improvements are expected for lakes with pH < 5. In fact, further emission reductions of 75% from 2010 targets should result in over 80% of eastern Canada supporting lakes with pH above 6 (i.e. above critical load levels) (see chart).
In examining regional chemical change, we arbitrarily defined "levels" of pH change as follows: no change - a change of 0, or a slight decline in pH (<0.1 units); minor increase - a pH improvement of 0.01 - 0.10 units; small increase - 0.1 - 0.2 pH units; large increase - > 0.2 pH units (see chart). Under the 1994 emission scenario (S3), minor increases would occur in 44% of eastern Canada (throughout central Ontario and most of Quebec), with small pockets exhibiting small increases (2%), but there would be no large increases in pH, and no change through most of the area examined (54%), particularly the Atlantic region. With the 2010 (S4) scenario, large increases (3%) would be observed in affected regions in central Ontario and western Quebec, with small increases (12%) occurring throughout much of the remaining acid-sensitive regions of Ontario and Quebec. As well, Atlantic Canada, particularly Nova Scotia, would experience minor increases in pH (39% of total land mass). With the recommended further 75% reduction in emissions from 2010 levels, most of central Ontario and vast regions of Quebec would experience large pH improvements (29%), as would portions of southern New Brunswick. Collectively, lakes in 66% of the modelled area would be expected to show some improvement in pH.
Piscivores - Estimates of changes in the suitability of lakes to support nesting piscivores (common loons and common mergansers) were derived across eastern Canada (see chart). Categories of change were set as: no change - <0.5% increase in suitability between scenarios; small improvement - 0.5-1.5% increase between scenarios; large improvement - >1.5% increase between scenarios (maximum recorded was 9.2%). Comparing predicted changes in habitat suitability between 1982 levels (steady state) and full implementation of U.S. controls (2010), no change was projected through most of eastern Canada (73%), particularly Atlantic Canada, with small improvements predicted to occur in 23% of the region, notably through much of Quebec and central Ontario. Isolated pockets of large improvements (4%) are projected to occur, especially near Sudbury and Rouyn-Noranda (two areas of historically high, point source emissions). With a further 75% emission reduction, however, large improvements would dominate changes (29%), most noticeably through central Ontario and most of Quebec, with small improvements occurring through much of Atlantic Canada (20%).
Collectively, the projected changes in pH and changes in habitat suitability (expressed as the % of affected land mass) confirm that each emission scenario results in some improvements to the chemical conditions and nesting habitat suitability of water bodies in eastern Canada (see WARMS Predictions and Habitat Predictions). A large proportion of the land mass will not change, either because lakes in certain areas are well-buffered, or do not receive enough deposition to cause chemical deterioration. As well, for nesting piscivores, many lakes will not change in suitability simply because they have characteristics that make them unsuitable for nesting (e.g. too small, too shallow), irrespective of lake chemistry.
What is clear, however, is that even with the 2010 scenario, relatively few, large scale improvements can be expected across much of eastern Canada. Instead, it will require further reductions to bring about broad scale, significant chemical and biological improvements.
Observed vs. predicted waterbird responses
Our models predict that piscivores should prefer higher pH lakes, and diving ducks, notably the common goldeneye, should prefer low pH lakes. In both cases, the birds probably respond to the presence of fish: piscivores like high pH lakes because they support healthy fish populations, and goldeneyes like low pH lakes because they often lack fish but support abundant, acid-tolerant invertebrate prey. Thus, with increasing lake pH, we expect fish to return to lakes (see chart), and correspondingly we should see increases in the breeding density of piscivores and decreases in the insectivorous goldeneye.
In Ontario, CWS has been conducting population surveys of waterbirds in the historically acid-damaged Sudbury region since the mid 1980s (see chart). Chemical and biological improvements are expected to occur more rapidly at Sudbury. In support of the model predictions, waterbird monitoring near Sudbury has demonstrated significant increases in numbers of breeding pairs of common loons, hooded mergansers and all piscivores combined (see chart), and significant decreases in numbers of common goldeneyes. Thus, patterns in numbers of local breeding water birds are in the direction expected if chemical and biological improvements are occurring in the Sudbury area.
Conclusions
Acid rain poses a serious threat to wildlife (notably waterbirds) in eastern
Canada through a variety of processes that occur at lower trophic
levels, but which ultimately cause reproductive impairment and/or
shifts in habitat selection or diet. The CWS (Ontario Region) Acid
Rain Biomonitoring Program monitors ecological responses of waterfowl,
loons and their foods to a changing acid deposition environment
to verify spatial and temporal aspects of the biological recovery
(status) of acidified, damaged or susceptible aquatic systems. Modelling
is an extremely important component of the program because there
is no direct means of predicting the nature and extent of aquatic
ecosystem recovery from acidification. Results from monitoring and
modelling efforts together demonstrate that certain waterbirds (especially
piscivores) are effective indicators of acidification, and that
little progress has been realized to date in chemical and biological
recovery of lakes. Our predictions demonstrate that little improvement
in habitat suitability for waterbirds will occur in eastern Canada
even with the strongest legislated emission reductions (post 2010),
but that further recommended reductions should lead to substantial
improvements in some populations, as currently being observed in
the heavily damaged (but recovering) Sudbury area. Further development,
validation, and expansion of regional waterbird models (and other
biological models), together with ongoing population and production
monitoring, is necessary to improve confidence and accuracy in habitat
suitability model predictions and observed progress.
NEXT PAGE
ACID RAIN HOME
|
