Could a residential wood ash recycling programme be part of the solution to calcium decline in lakes and forests in Muskoka (Ontario, Canada)?

Calcium (Ca) concentrations have fallen in forest soils in eastern North America following decades of atmospheric acid deposition (Likens et al. 1998; McLaughlin 2014). Falling Ca levels in soils have reduced growth (Watmough 2002; Gavin et al. 2009; Bishop et al. 2015), health and vigour (Hallett et al. 2006; Long et al. 2009), seedling survival and regeneration (Cleavitt et al. 2011; Sullivan et al. 2013), and foliar nutrition (Bailey et al. 2004; Long et al. 2009; Miller and Watmough 2009) of sugar maple (Acer saccharum Marsh, 1785). In lakes, Ca decline has damaged populations of crayfish (Edwards et al. 2009; Hadley et al. 2015) and several species of Daphnia (Jeziorski et al. 2008), with consequences for pelagic food webs (Jeziorski et al. 2012, 2015; Korosi et al. 2012).

Despite reductions in acid deposition, the natural recovery rate of Ca will be extremely slow (Likens et al. 1996; Aherne et al. 2003; Watmough and Aherne 2008); hence, adding Ca as a crushed mineral or wood ash to forested watersheds is an attractive option. Calcite additions have certainly improved chemistry and biota in acidified lakes (Yan et al. 1996), soil chemistry (Driscoll et al. 1996; Geary and Driscoll 1996; Löfgren et al. 2006), and forest health, regeneration, and productivity (Wilmot et al. 1996; Moore and Ouimet 2010, 2014; Mainwaring et al. 2014). However, limestone must be mined and shipped to areas of need. Large doses could also destabilize forest ecosystems (Lawrence et al. 2016), and limestone contains no plant nutrients other than Ca. Thus, wood ash is a promising alternative Ca source.

Industrial sources of wood ash are common. Canadian pulp and paper mill boilers alone produced up to 0.75 million tonnes of wood ash annually in 2002 (Elliott and Mahmood 2006), and it can increase soil pH, foliar Ca, tree growth, and overall forest productivity when nitrogen is not limiting (Augusto et al. 2008; Pugliese et al. 2014; Reid and Watmough 2014). Non-industrial wood ash (NIWA) (e.g., from homes and wood-fired pizza bakeries) is also an alternative and 18 000 tonnes of wood ash are generated from residences in Ontario annually (Azan 2017). However, unlike industrial wood ash, the use of NIWA to ameliorate Ca limitation in forest soils and indirectly, downstream lakes, has not been studied, and it does pose issues that differ from industrial ash sources. Its composition, including its contaminant levels, might be more variable than industrial ash sources, given the variety of tree species and tree parts burned, the lack of consistent feedstock control, and differences in non-industrial versus industrial combustion and ash processing. We do not know if NIWA would pass government guidelines for metal contaminants (Hannam et al. 2016), and the potential toxicity of NIWA has not been evaluated. Finally, the methods for safe and efficient collection, transport, and distribution of NIWA have not been developed, but they would clearly differ from industrial ash sources.

To address these hurdles in the development of a NIWA recycling programme designed to combat environmental Ca decline in Muskoka watersheds, our objectives were to: (i) quantify the total cations and metal concentrations in typical NIWA generated from the combustion of hardwood firewood in central Ontario; (ii) quantify the soluble levels of Ca and other major cations, anions, and nutrients in NIWA; (iii) compare metal concentrations in NIWA with provincial guidelines and with industrial wood ash; (iv) quantify the actual toxicity of aqueous extracts of NIWA in both water column and sediment deposition scenarios; and (v) estimate the amount of NIWA that might be available in Muskoka, Ontario to launch a residential wood ash recycling programme that could help solve the provincial Ca decline problem.

Most central Ontario residents that heat their residences with wood burn hardwood. Azan (2017) reported that 83% of survey respondents that heat their Muskoka, Ontario homes with wood, burned wood from maple, beech, oak, birch, ash, and cherry trees (in declining order). To address our first four objectives, we obtained hardwood NIWA samples generated over the course of the 2016/2017 winter heating season from nine independent, permanent residents and one wood-fired pizza bakery located in and around Baysville and Bracebridge, Ontario, Canada, that responded to our public call for wood ash. Using a stainless steel riddle (mesh size 2.38 mm), we sieved the ash to remove occasional foreign objects (e.g., galvanized nails and staples) and the coarse clinker that is normally generated by incomplete wood combustion. We removed the clinker as our primary focus was on the fly-ash-like component that we assumed would contain the majority of the mineral phases, such as Ca. All subsequent work, analyses for solubility, trace metal concentrations, and potential toxicity once dissolved in soft water medium, was done with these sieved NIWA samples.

Our study was designed to determine if non-industrial, mainly residential wood ash (NIWA) has the potential to be used as a forest soil amendment in central Ontario and be part of the solution to the widespread environmental Ca decline problem in central Ontario lakes and forests (Watmough and Dillon 2003a, 2003b; Jeziorski et al. 2008). To date, this NIWA source has neither been considered as a large-scale forest soil additive, nor as an aid to reverse widespread Ca decline in lakes, likely because of its unknown chemistry and distributed production. We do not think the latter reason should impede its consideration. After all, distributed production has not stopped any other residential recycling initiative.

To examine the question of whether NIWA could be part of the solution to Muskoka’s environmental Ca decline, we must: (i) document the concentration and solubility of Ca in the ash as part of the site-specific calculation of required application rates; (ii) compare levels of potentially toxic metals with regulated targets; (iii) if possible, quantify the actual toxicity of the wood ash to terrestrial and aquatic model species; (iv) quantify levels of any non-metallic ash components (organics and P) that might also pose environmental issues; (v) document quantities of NIWA that are produced annually in the region, and would likely be made available for the recycling programme; (vi) test the dosage calculations and ecotoxicological implications of NIWA issues in pilot studies, perhaps using an adaptive management philosophy; and (vii) resolve any major policy, logistical, economic, and stakeholder questions and impediments related to ash collection and post-collection treatment, e.g., pelletization, storage, delivery, distribution, and application. We have largely focused on the first five of these requirements in this paper. We recognize that the last two issues, particularly the last one, are major, so we do consider them briefly in the terminus of our Discussion.

However, to eliminate any confusion in the Discussion, we should briefly consider the goals of such a residential wood ash recycling programme. We suggest its central goals should be to: (i) increase citizen participation in environmental stewardship; (ii) increase the diversion of waste from landfill sites; (iii) reverse decades of Ca loss from watershed soils, thus increasing carbon capture and health of forests in eastern Canada; and (iv) indirectly, solve the widespread Ca loss problem in lakes by restoring soil base cation pools.

Our NIWA samples averaged 30% Ca and could replace missing Ca levels in forest soils. Such high levels of Ca are in agreement with those reported for wood-derived boiler ash: 3.6%–96.6% (as reviewed by Vance 1996), 13.2%–27% (Naylor and Schmidt 1986), 11.9%–28.4% (Steenari and Lindqvist 1997), and 9.2%–24.79% (AshNet 2018). Our results confirm Ca levels observed in ash from a home wood stove: 28.1%–32.1% (Naylor and Schmidt 1986). Although our focus was on hardwood NIWA, we had an additional source of ash from a resident that burned exclusively wood pellets made from northern pine and spruce trees. This sample had a higher Ca concentration of 37% (Azan 2018), but we would need additional samples to confirm that wood pellet stoves routinely produce ash with such high Ca levels.

Although our NIWA samples contained high total levels of Ca, it was not particularly soluble. Ca was released into aqueous media slowly from our hardwood NIWA, especially in the first 9 d of exposure (Fig. 3). However, by day 30 of the solubility experiment, 0.4% of the total Ca was in solution, reaching an average of ~11 mg/L of Ca, still lower than the 15 mg/L we might have expected from the known solubility of CaCO₃, the primary Ca compound found in wood ash (Steenari and Lindqvist 1997; Serafimova et al. 2011). Such gradual solubility suggests that hardwood NIWA could be useful in the long term as a slow-release source of Ca for Muskoka forests, and potentially for low Ca lakes should wood ash be leached to them.

Our results suggest that heavy metal concentrations in NIWA should not prevent its use as a forest soil additive. None of the samples exceeded the CM2 standard, which would prevent use. Indeed, the majority of trace metal concentrations in the NIWA fell below the regulated levels for CM1, the target for unrestricted use (Table 1). The exceptions were Cu and Zn, whose averages were just above the CM1 targets but well below the CM2 standard. The peak value we might expect to see in only 2.5% of NIWA samples also exceeded the CM1 target for Cu and Zn, as expected, given their means, but also for Cd. Such elevated levels of Cu and Zn, and on occasion Cd, likely reflect the unusually high metal deposition rates that occurred over the past 50 years or so (Pratte et al. 2013) that are now reflected in forest soils (Landre et al. 2010; Watmough 2010; Doucet et al. 2012) and lake sediments (Couillard et al. 2008; Landre et al. 2010). Any wood being burned now likely came from the same time period. As atmospheric metal deposition rates peaked several decades ago and are in decline, metal concentrations in NIWA will likely be lower in the future. Although it is unlikely that these rare individual samples would render bulk NIWA unusable, further metal analyses would certainly be required on bulk ash mixtures, as they are recycled for compost.

It is noteworthy that mean metal concentrations in our NIWA tended to be much lower than the mean and minimum values for industrial sources of wood ash (Table 2), with the exception of Cu and Zn. These results suggest that NIWA would be no more of a toxic threat, if any, than many industrial sources of ash as an additive to forest soils in central Ontario. Hence, we do not think that the use of hardwood NIWA would result in metal toxicity in the receiving soil.

We provide the first direct assessment of the potential short- and longer-term toxicity of aqueous extracts of NIWA to Daphnia. Although we did detect acute toxicity in concentrates, this completely disappeared with modest dilution, and we detected no toxicity of sedimented NIWA on Daphnia survival, maturation, or brood production. Still, the high mortality in the concentrate and DF10 treatments must be explained. Beyond the strength of the concentrate itself (a 10% ash solution), we hypothesized this mortality had one of two logical sources: (i) the high pH of the solutions, or (ii) the concentration of elements present in the NIWA, including both metals and cations, other than Ca. The pH of the concentrates ranged from 10.65 to 12.49. To determine if the high pH of the concentrate and DF10 treatments explained the observed mortality rates, we repeated the 48 h experiment after first lowering the pH to 8, a pH typical of hard water lakes. As in the initial 48 h experiment without the pH adjustments, we detected 100% mortality rates in the concentrate and DF10 treatments, but excellent survival in the DF100 and DF1000 treatments (Table S1). We concluded that high pH was not the cause of the mortality observed in Daphnia neonates.

Our alternate hypothesis was that some other element (or elements) present in the NIWA extract was responsible for the observed mortality in the concentrate and DF10 treatments. We do not think that Cu or Zn were the responsible toxicants, because at the high pH of the medium, very little, if any, Cu or Zn would have been soluble (Dortwegt and Maughan 2001; Jarvis et al. 2012). The most logical possibility was K, which was very soluble in our NIWA (Fig. 3). K levels in the concentrate and DF10 treatments ranged, respectively, from 2699 to 7871 mg/L, and 296 to 449 mg/L, much higher than the acute LC50 of 93 mg/L measured by Biesinger and Christensen (1972). To check K toxicity to our daphniid clone in our soft water medium, we ran a 48 h bioassay using potassium chloride (KCl) at K concentrations of 50, 100, 150, 200, and 250 mg/L. We observed 100% survival in the first three concentrations (50, 100, and 150 mg/L), then 80% and 0% survival in the final two concentrations, i.e., at 200 and 250 mg/L of K, respectively (Table S2). Interpolation of the survival (%) versus K concentration data provided an acute LC50 for K for our daphniid clone of between 200 and 250 mg/L, slightly higher than Biesinger and Christensen’s (1972) value. This amount of variation in LC50s for metals is well known in inter-laboratory studies using Daphnia (Baird et al. 1990). These results strongly support the hypothesis that enough K was solubilized from the NIWA to account for the mortalities observed in the concentrate and DF10 treatments in the acute bioassays. Daphnia neonates exposed to high levels of K concentrations (ranging from 195.5 to 1560 mg/L) would experience a depressed heart rate, and eventually death (Baylor 1942). However, we do not expect such K toxicity from NIWA applications in nature, given ash dilution over time and distance. All required elements can exist at sub-optimal, optimal, and toxic levels, and it was our extreme concentrate (a 10% ash solution) that produced toxic levels in our assays. Still, as a precaution, K levels in ground and surface waters should be tracked shortly after NIWA additions, especially following precipitation events. We do know that the NIWA contains a lot of soluble K (Fig. 3). Further, if spread, the NIWA should be dispersed widely, not left in concentrated piles, which might produce localized toxic levels of K after rain storms.

In addition to Ca and K, concentrations of other nutrients varied across the 10 hardwood NIWA samples. This result was expected as the chemical composition of wood ash can vary among tree species used as fuel, with the parts of tree in the fuel (e.g., bark, twigs, wood, and leaves), with the chemistry of the soil that supported the trees, with combustion temperature and type, and with storage of fuel wood (as reviewed by Vance 1996; Demeyer et al. 2001; Pitman 2006). In central Ontario, local residents and the wood-fired pizza bakery burn the main stem or trunk of trees, bark, or branches (in descending order) of hardwood tree species (e.g., maple, oak, and beech) as a heating source (Azan 2017), which might account for the variation in nutrients observed across the 10 hardwood samples collected. This is consistent with observations by Pitman (2006), who also highlighted the variability in nutrient concentrations in ash that resulted from burning different tree species and tree parts. Sample 9 came from a wood-fired pizza bakery, and its higher levels for K, Mg, Na, and P could be due to the dosing of pizza bases with flour as they were placed in the oven. It is worth noting that flour has high concentrations of K (4%), P (13%), and Mg (7%) compared with the bole wood of trees.

Wood ash can have high P concentrations, and we could not recommend the large-scale addition of NIWA to the forest as a treatment for Ca decline if it would lead to the eutrophication of downstream waters. However, we feel this risk is low. P was the least soluble of the elements, at only 0.02% after 30 d for ash obtained from residences, and 0.05% for ash from the pizza bakery (Sample 9). From the dry ash analysis, we noted that Sample 9 contained 1.2% of P, about twice the 0.6% P average from the other nine sources of NIWA, all of which were residential (Fig. 2). To explore the possibility that P in NIWA might pose a risk of eutrophication, we estimated the amount of readily available P that would be supplied to a watershed receiving 4 tonnes of NIWA/ha of watershed—a suggested minimum application rate that could be used to help solve the Ca decline problem in Muskoka (C. Reid, personal communication, 2017). At 4 tonnes/ha, ~800 g (0.8 kg) of P/ha (Appendix 2: Azan 2018)—would be readily available to watershed soils. This value is roughly equivalent to the natural load of P from rain and snow of 750 g (0.75 kg)/ha, recorded annually in Muskoka (Nicholls and Cox 1978), and is 16 times and three times more, respectively, than is normally released annually from forested areas in Muskoka (50 g/ha: Dillon and Rigler 1975), and from the average stream exports of six soft water lakes in south-central Ontario (231 g/ha; Eimers et al. 2018). As a precaution, changes in TP concentrations in ground and surface waters should be tracked post-addition of NIWA in watershed soils.

We do not think enough NIWA is generated in Muskoka to solve the Ca decline problem in local forests and lakes, given its widespread nature, but enough ash is generated in Ontario. Muskoka residents generate about 235 000 kg (235 tonnes) of NIWA per heating season. If the estimated net loss of Ca from the soil exchangeable pool is 7.14 kg/ha per year (Watmough and Dillon 2003a), then the amount of Ca lost over the past 50 years of acid deposition is roughly 360 kg/ha or 0.36 tonnes/ha. If all the Ca in wood ash was soluble, this would require about a tonne of wood ash/ha. Assuming 4 tonnes/ha would be appropriate, Muskoka’s 235 tonnes of annual ash would treat only 58 ha, enough for a large plot-scale study, but not enough to solve the problem given its widespread nature (Reid and Watmough 2016). However, about 18 000 tonnes of residential wood ash are generated in Ontario annually (Azan 2017). In principle, this is enough to treat 4500 ha of Muskoka forests per year that certainly begins to address the scope of the problem over a decade or two. This suggests a province-wide NIWA recycling programme is worth considering as an option to solve the Ca decline problem in central Ontario. Of course, the inclusion of industrial sources of wood ash could shorten the time required to solve the problem.

Approved wood ash dosage rates have not been set for forest soils across Canada, and most provinces determine recommended doses based on the liming requirement of the soil and (or) on the trace metal levels in the ash and receiving soil (as reviewed by Hannam et al. 2016). In Europe, approved doses range from 0.5 to 7.5 tonnes/ha, some of which are based on the quantity of Cd or P in the soil, the site (e.g., mixed peatland or upland forests), and soil type (e.g., nutrient poor sandy soils) (as reviewed by Hannam et al. 2016). Field trials currently underway by participants in Canada’s national wood ash research network, AshNet (nrcan.gc.ca/forests/researchcentres/glfc/ashnet/20279)—whose primary focus is forest productivity—are using similar dosage rates for industrial ash as used in Europe (ranging from 0.5 to 20 tonnes/ha; Hannam et al. 2018). However, there is currently no understanding of what wood ash application rates to forested watersheds are needed to raise Ca concentrations in downstream waters to >1.5 mg Ca/L—a biological threshold above which Daphnia reproduction and survival increases (Ashforth and Yan 2008). As we expect Ca concentrations in many central Ontario lakes to fall to 1 mg/L or less (Reid and Watmough 2016), there is clearly work to be done.

Although we have shown that, like industrial wood ash, NIWA can be used as a forest soil additive and, in consequence, benefit Muskoka forests and lakes, there are several barriers that prevent its wide-scale use as a forest soil additive in Canada. In their recent review, Hannam et al. (2018) indicated that regulatory approval, the variability of wood ash chemistry, and a lack of knowledge of the effects of wood ash on forested sites are just some of the issues to overcome. The variability of NIWA is unlikely to be an issue, as metal concentrations are often lower than industrial sources and ash from treated or salt-laden wood are unlikely to be used. Current AshNet field trials are likely to provide useful knowledge on the effects of industrial wood ash on forested sites, which might be of use for choosing dosage rates that would provide surplus Ca that could increase levels in ground and thus surface waters.

Implementation of any wood ash recycling programme to help solve the Ca decline problem would have to address issues such as ash collection, ash distribution to forested areas (e.g., manual vs. mechanical, or aerial spreaders), dosage rates, ash storage, frequency, and methods of application. Resolving these issues can help to save money by increasing waste diversion from to municipal landfills (Hope et al. 2017) and benefit the environment. With the exception of dosage rates, these are largely not ecotoxicological issues, but economic, logistical and engineering issues, and quite daunting ones at that. Once the ecotoxicological issues are resolved, they will require lots of enlightened discussion with waste management authorities and contractors if we are to develop a residential wood ash recycling programme to help solve the environmental problem of Ca decline.

The continuous decline of Ca concentrations in surface water is disrupting the ecology of lakes of central Ontario and low Ca concentrations in soils are limiting forest growth of some regional forests. One possible solution to combatting this issue is to supplement watershed soils with Ca. Our study has demonstrated that aside from crushed limestone and industrial wood ash, NIWA may be a viable option. Large quantities are readily available during the heating season. Although NIWA contains several trace metals of concern to regulatory authorities, the majority of these levels were below the CM1 target, the threshold for unrestricted land use application in Ontario. Thus, NIWA application is unlikely to increase metal concentrations in watershed soils to levels that might negatively impact Muskoka’s forests and lakes. Although there is evidence of short-term toxic effects from high levels of K, we think that this toxicity would dissipate rapidly, and the residual ash that sank to lake bottoms would not be toxic. However, given the great solubility of K minerals in the ash, and the presence of P, which can pose an eutrophication risk, initial monitoring for metals and other elements in the landscapes receiving ash, and a large plot-scale study is recommended.

The implementation of a wood ash recycling programme may help to combat Ca decline and limitation. Such a programme would encourage residents of Muskoka and the province of Ontario to become “gardeners” and not “miners” of the forest. From our estimates, enough NIWA is available in Ontario to make a difference. In the initial stages, the wood ash recycling programme could focus mainly on residential ash, which is unlikely to negatively impact Muskoka’s natural environment. Small businesses that use wood as a heating source could also be included in the programme once their wood ash has been analysed for metals, K, and P concentrations. As there is 28% as much NIWA as industrial wood ash in Canada, researching it appears worthwhile. With the first step of ecotoxicological work completed, it is now time to collectively tackle the diversity of other considerations (e.g., application rates and frequency, proof of benefits, site specific dose requirements, logistics, distribution (e.g., wood ash pelletization), policy implications, and economics of broad-scale wood ash recycling) required to combat Ca decline.

We thank the nine local, permanent residents and one wood-fired pizza bakery that provided the wood ash for our study. We also thank Dean Haydon, our laboratory assistant, who assisted with all the work, and the Ontario Ministry of Environment and Climate Change (now the Ontario Ministry of Environment, Conservation and Parks), Dorset Environmental Science Centre and its staff and interns for logistical support, assistance, water chemistry analyses, and advice. This study was supported by a 2017 Ontario Trillium Foundation Seed Grant to Friends of the Muskoka Watershed and by NSERC funding to both Norman D. Yan and Shelley E. Arnott.