Draft Salt Issue

The CCME guideline for choride ion for the protection of aquatic life is 120 mg/L for long term exposure, and 640 mg/l for short term exposure, corresponding approximately to EC values of 470 and 2410 uS/cm respectively, using the conversion formula given in AECOM 2020.

The current value for water at the surface of Sandy Lake is approaching 200 uS/cm,  well above its pristine state (about 40 uS/cm), and still well below 470 uS/cm but the trend is clearly upwards. Values in streams entering Sandy Lake that drain settled areas are frequently above 470 uS/cm. In our monthly monitoring, the highest stream value was 1744 uS/cm, still below the short term toxicity level (~2410 uS/cm), however it’s quite likely that continuous monitoring would reveal spikes  in the acute toxicity range (2410 uS.cm and above, see data for Little Sackville River).

EC values for water at the surface and near the bottom in deepest area of Sandy Lake over 50 years

The increase in hypolimnion EC/chloride compared to the surface water  in Sandy Lake is potentially a  more serious threat from road salts than posed by toxicity of chloride.

If the increase continues it could  impair normal turnover of the lake, most likely the spring turnover, if it has not already. Impairment of spring turnover  reduces oxygenation of deep water which can have highly adverse effects, including increased release of phosphate from sediments (“Internal phosphprus loading), increased mobility of metals such as cadmium, lead, chromium, and reduced biodiversity of benthic and other aquatic communities of a lake (Novotny 2012).

Currently, our limited information of values during summer/early stratification indicate deep water oxygen levels are dropping below limits for salmonids, while data presented in AECOM 2014  suggests the lake has at times dropped to levels sufficient to mobilize sediment phosphorus.

While it is well documented from the synoptic observations on 50 HRM lakes that salt levels in HRM lakes are increasing and are highest in lakes in more settled areas (SL Fig 11 above),   there has been  little or no highlighting of the possible longer term effects on  turnover of stratified lakes in provincial and HRM documents until very recently (Re: AECOM 2020).   Even on a North American scale, it’s hard to find more than a dozen or so papers documenting the phenomenon, although it is now commonly cited in reviews on impacts of road salt as a concern (e.g, Dugan et al., 2017; Hintz and Relyea, 2019).

I am aware of only one  situation in NS for which there is good   evidence  that salt stratification has impaired normal turnover.   Oathill  Lake in Dartmouth is a small lake (some stats: surface area 4.2 ha, watershed 33 ha, max depth 8.5 m, average depth 3.6 m) situated in a highly settled area. It ranks 9th amongst the 51 lakes in the synoptic surveys in regard to surface water EC  on April 7, 2011  (Sandy Lake is # 32) – see Fig SL 11 above. A slide set on Oathill Lake Restoration posted on the HRM website provides an overview of recent history of the lake, and results of monitoring including vertical profiles. It is a popular lake for swimming and fishing. In 2009 extreme coliform counts and the lake “turning brown” raised  alarm bells and volunteers they began regular monitoring.  In 2017, deep water EC was circa 1200 uS/cm March through October, compared to values in the range 250 to 550 uS/cm on the surface;  there was apparently, no spring turnover, but the lake turned over at the end of October and the deep water EC level fell abruptly. Road salts were the major sources of EC, but release of ions from anoxic sediments also contributed. Solar powered Aquago devices were installed to aerate deeper waters without disrupting the hypolimnion with some very good results. It is a complex  story, but surely makes the point that if it is possible to avoid reaching such a state in the first place, that is a far more preferable option.

I  have cited* salt stratification as related in some way to  Williams Lake (63 ha, max. depth 20 m) on the Halifax south mainland not turning over by early December in 2015, but it is not clear how they are related. There was a sharp gradient in oxygen  and EC at 7-9 m; the hypolimnion oxygen was only 1-1.3 mg/L (12.1-12.3 mg/L in the epilimnion). The difference in EC between the epilimnion (277-281 uS/cm) and bottom (307 uS/cm) was not large and  was very unlikely to be sufficient to inhibit mixing.   The higher EC values in the hypolimnion could be attributable to mobilization of ions under low oxygen. As at Sandy Lake, streams draining settled areas had  high EC values (282-427 μS/cm where they enter the lake) and streams draining natural areas very low values (34-59 μS/cm);  the lake value at the outlet, circa 290 uS/cmm reflected the significant  influence of water from settled areas.
* Water quality measurements on Williams Lake and Colpitt Lake (Halifax, N.S.) Dec 7-13, 2015 with reference to possible impacts of road salt, DG Patriquin, 2016., Report to Williams Lake Conservation Company.

Other than these two examples, I am not aware of any other documentation of possible or probable impairment of turnover of NS lakes related to road salts, but it must be occurring.

“Chloride enrichment” is highlighted as one of four Key Water Quality Concerns recommended to be investigated/monitored as part of a renewed Water Quality Monitoring Program for HRM, as cited in the recently released report Halifax Regional Municipality Water Quality Monitoring Policy and Program Development (AECOM 2020). The others are Eutrophication, Bacteria Contamination and Invasion of Non-native Aquatic Species. The focus for chloride is on toxicity but it is acknowledged that “In extreme cases, high chloride concentrations can increase the density of bottom water to such an extent that it prevents lakes from mixing (called meromixis). Meromixis can cause anoxia that leads to impacts on habitat for aquatic biota and the release of nutrients and other chemicals from sediments (i.e., internal loading) that can contribute to eutrophication.”

Sandy Lake is one of 25 Priority Lakes recommended in the AECOM 2020 Report for regular monitoring to represent one or more of three WQ vulnerabilities – chloride, eutrophication and  bacteria, the criteria being as follows:

Priority Eutrophication Lakes – lakes with elevated surface water TP concentrations during the ice-free period that are indicative of eutrophic conditions (i.e., >20 μg/L) based on a high-level review of monitoring data from select sources, and lakes with documented past issues with algal blooms or nuisance aquatic plant growth.

Priority Chloride Enrichment Lakes — lakes with elevated spring surface water chloride concentrations (i.e., >100 mg/L) that are approaching the Canadian Water Quality Guideline (CWQG) of 120 mg/L for long-term exposure for freshwater aquatic life, based on a high-level review of monitoring data from select sources.

Priority Bacteria Contamination — lakes with municipal beaches where HRM provides supervision due to human health risks from recreational body contact with water.

By these criteria, Two vulnerabilities – Eutrophication and Bacteria – but not Chloride apply to Sandy Lake.

A key question:

How great a difference in EC/salt content between the hypolimnion and epilimnion is required to impair turnover (i.e. between water at the surface and water in deeper parts of the lake )?

This will  depend the on same  site-specific factors that affect thermal stratification, and on other factors.

Individuality is a prominent characteristic of the observed patterns of thermal structure and is governed strongly by climatic variations, volume of inflow and outflow in relation to the volume of the basin, basin configuration, surface area of the lake, position of the basin in relation to wind action, and other factors… In most lakes, salinity gradients are insufficient to increase stability to a point where where wind energy does not cause holomixis. A large number of lakes, however do exhibit temporary or permanent meromixis as a result of salinity gradients. In such cases, the term concentration stability, in contrast to thermal stability of lakes is…more appropriate (R. Wetzel 2001, Limnology 3rd Ed.,  p 80-83)

The effect of differences in temperature on water density (and hence on stratification) are much greater at higher temperatures (circa 20 degrees) than at lower (circa 5 degrees), but temperature has a relatively small effect on how solutes affect water density (view example). For typical temperate, dimictic  lakes,  the larger impact on density at low temperatures, together with high salt input in winter months results in  concentration stability being more likely to impair spring turnover than fall turnover (Novotny et al., 2008,  2012); or for spring turnover to be the first affected as concentration stability increases in a lake.  Impairment of spring turnover but not fall turnover appears to the case for most instances in which impaired mixing has been reported,  i.e. with increasing salt loading, the mixing regime progresses from dimictic (spring and fall turnover) to monomictic (fall turnover only);  with further increase in concentration stability a lake could become meromictic (permanent incomplete mixing).

For 9 lakes affected by salts, Novotny 2008 noted that 5 lakes appeared to have stronger chemical stratification than the other four and that these five had the smallest surface area to depth ratios (0.3-3.3, compared to 3.5 to 7.6 for the 4 lakes with weakest salt stratification. (The Surface Area to volume ratio Sandy Lake is 3.6). Novotny commented that ” a very similar parameter, the lake geometry ratio defined as the ratio of maximum lake depth to the fourth root of the lake surface area was introduced (Gorham and Boyce, 1989) as an indicator of the strength of temperature stratification of lakes.”

In addition to the factors summarized above,  Novotny (2012) describes how patterns of ice formation, snow melt etc. can determine whether a particular  concentration gradient will impair turnover.

There is  at least one report of  spring turnover being impaired at chloride levels below the CCME guideline for chloride ion for the protection of aquatic life  (120 mg/L, or approx.  470 uS/cm by the AECOM 2020 formula) for long term exposure:

 A reduction in spring mixing due to road salt runoff entering Mirror Lake (Lake Placid, NY) by B. Wiltse et al., 2019, Lake and Reservoir Management 36:109–121.

Recent  EC values for Sandy Lake during the summer stratification are similar to those cited for Mirror Lake in summer which is of similar size and depth; spring sampling in Sandy lake could well reveal impairment of of normal spring turnover as at Mirror Lake; indeed that seems very likely to be case in some years (view Mirror Lake – Sandy Lake comparison.

As well,  EC values of incoming streams at Sandy Lake are well above those values, and the seasonal sampling of streams suggests the ground water could have high values.  Thus we need to be concerned about the possibility of salt stratification at Sandy Lake impairing spring turnover  even before chloride values exceed CCME guidelines, indeed it seems very likely to have already have been  impacted in some years. Climate warming could make the impacts worse by strengthening and/or prolonging summer stratification (e.g., view Hadley et al., 2014, Niedrist et al., 2018).

While Best Management Practices can reduce salt loading and the impacts on lakes (e.g., view Bubeck and Burton, 1989), it’s  clear that the major anthropogenic determinant of salt loading is the percentage of a watershed that is settled/hard surfaces.  From a recent, comprehensive review  (Dugan et al., 2017):

Results … revealed that impervious land cover and road density surrounding each lake were the primary classification splits and the most important predictors for lake chloride trends and cluster grouping

In North America, specifically in the Midwest and Northeast, local salt application leaves freshwater lakes vulnerable to salinization. Of the 284 lakes in the NALR, 26 already have a chloride concentration above 100 mg L−1 at their last sampling date. The median impervious land cover within a 500-m buffer surrounding these 26 lakes is 24.8%, compared with the US mean 0.31%. If a linear relationship between time and chloride concentration is extrapolated, 47 lakes are on track to reach 100 mg L−1 by the year 2050, and 14 are expected to surpass the EPA’s aquatic life criterion concentration of 230 mg L−1 by 2050 (Fig. 4B). This is also the concentration at which a deterioration in drinking water taste is perceptible.

Hammonds Plains Road is an increasingly busy major road with low spots that drain into Sandy Lake, and new development has been approved within Bedford West sub-area 12 which  also drain into Sandy Lake. Those developments alone will increase salt loading on Sandy Lake. Major new development north of Hammonds Plains Road/west and southwest of Sandy  Lake would lie much closer to  the lake and major inflowing streams and could result in quite rapid and severe firther deterioration of  Sandy Lake

The rising salt levels in urban lakes and rural lakes impacted by large highways  all over North America is proving to be a very difficult trend to reverse, and while the salt itself is an issue, salt is also a proxy for a host of other materials associated with urbanization and highways that negatively impact lakes.

Sandy Lake represents a situation where we still have the option of choosing the most effective way to prevent the otherwise predictable deterioration of a treasured lake: to strongly limit as yet unapproved development in the watershed, and especially within 500 m of the lake.

It is a choice we may want to apply to other lakes. Currently, the principal way we attempt to predict potential  impacts of development on lake water quality is by modelling the impacts on lake total P  but without consideration of impacts of climate warming or increased salt loading on lake total P and associated processes.

In HRM, we have the historic database, the coverage of a wide range of lakes, a strong academic tradition and ongoing research  related to water resources, community support and involvement in water monitoring and a renewed commitment to monitoring that could enable us to  produce far more predictive models to aid in decision-making. Sandy Lake, already chosen as one of the Priority Lakes for monitoring (if that goes ahead) would make a fine case study of a lake in transition – and hopefully of one for which we can reverse the current trends.