3.5. Modelling diversity & production

3.5. Modelling climate change responses of biodiversity, primary production, and local food production in Eeyou Istchee and Nunavik

When people eat local food, they are consuming biodiversity. They are feeding themselves with what the land and water provides. Food is the primary environmental determinant of health in that it is created by a combination of environmental conditions (e.g., sunlight) and biological processes (e.g., photosynthesis, herbivory) and is essential to nutrition and well-being (Ort et al. 2015). Although some have suggested otherwise (Jones and Porup 2012), plants and plants alone can use solar energy and water to capture atmospheric carbon into their tissues. This process of photosynthesis turns gaseous carbon into organic compounds that can be eaten by people and animals. At the same time, photosynthesis generates oxygen. The gift of plants and their gift of photosynthesis provides us with the oxygen we breathe and the food we eat, whether we obtain our food directly from plants or indirectly from animals that feed on plants or other animals. In describing and modelling these ecosystem functions and trophic transfers in some detail, this section descends into the complexities, terminologies, and empirical limitations of ecosystem and food web ecology. And there is a lot more biocomplexity out there than the carbon-based trophic pyramids we consider here (e.g., Hillebrand et al. 2014; Selosse et al. 2017). Despite the jargon, the complexity, and the reductionism, the approach might be seen to be prioritizing the same relationalities as those emphasized in Indigenous worldview and its recognition of relationship and co-dependency between people and nature. A trophic pyramid and estimation of primary and secondary production are, in some ways, an ecologist’s clunky and awkward attempt to communicate the relationality expressed more directly and eloquently by the Kanien’kehá:ka verb for traditional foods - Tiohnhehkwen - ‘they provide us and sustain us with life’ (see section 3 and Decaire 2021). Ecological science focus on biodiversity considers the "they" - who they are, where they live, what they are like, and how many of them there are - and our focus on ecological production considers the "provide" and the sustainability of this provisioning.

We approach modeling response of biodiversity using two complementary approaches, the first focused on species diversity patterns and the second on primary and local food production. We relate spatial patterns of diversity and production to annual average temperature variation, then use these climate-diversity and climate-production relationships together with future climate scenarios to project future diversity and production in northern Quebec under climate change.

3.5.1. Biodiversity gradients and ecological production

Regional biodiversity varies dramatically with latitude [1, 2]. Animal and plant diversity tends to increase from the equator to the tropics, then decline precipitously from tropical to polar latitudes [2]. The well-studied biogeography of North American fishes, birds, and mammals conform to these general latitudinal diversity patterns [3, 4, 5]. Species diversity of all three groups is generally highest in Florida and/or Mexico, then gradually declines across temperate forests, the boreal, and the tundra to reach a continental minimum in the high arctic (Figure 31). For North American mammals, five environmental variables, representing seasonal extremes of temperature, annual energy and moisture, and elevation, account for 88% of this continental variation in species density [3]. Within Canada, mammal species density varies from greater than 80 in southern British Columbia to less than 10 in the arctic archipelago, and 74% of this variation is explained by annual average temperature alone [6].

3.5.4. Modelling climate change responses of northern Quebec primary productivity and local food production

An alternate approach to forecasting climate change impacts on ecological production and its support of local food systems, which is distinct from but complementary to the previous section focus on climate diversity gradients the specificities of species interactions, is to focus on the ecosystem processes of primary and secondary production.

Net primary productivity (NPP) has been estimated over the Canadian landmass at 1‐km resolution, based on an instantaneous leaf-level photosynthesis model applied over large areas, combined with remote sensing of leaf area index (10-day intervals) and land cover, and daily meteorological data, available soil water holding capacity and forest biomass [39]. Similar to the species diversity analysis described above, we relate spatial patterns of estimated primary production from Liu et al. [39] to annual average temperature variation, then use this climate-productivity relationship together with future climate scenarios to project future primary production in northern Quebec. For this analysis, we focus on the same 1,945 km south-north transect extending from 45.0^oN, 73.7^oW (south of Montreal at the US border) to 62.5^oN, 73.7^oW (northeast of Salluit, Nunavik at the edge of the Hudson Strait). At eight points along this transect (every 2.5^o of latitude), we determined estimated primary production from Liu et al. [39], and average annual temperature (1981-2010 reference period) and projected climate change (horizon, 2041-2070; emissions scenario, high (RCP 8.5); percentile, 50) from Ouranos (2021).

Net primary productivity declines from south to north across Canada and within Quebec [39]. Expressing this Quebec productivity gradient in relation to average annual temperature indicates minimal primary production of about 0.01 kg of carbon per square metre per year (hereafter kgC/m²/yr) at the northern limit of Nunavik, where T_{avg annual} is -7.5^oC, that gradually but exponentially increases with increasing temperature (Figure 33A). Primary productivity is characterized by the largest absolute increases within a T_{avg annual} range of -2.5^oC to +2.5^oC. Situating this climate range within Northern Quebec's bioclimatic domains (Figure 1), the largest increase in primary productivity occurs from the southern edge of the forest tundra (0.025 kgC/m²/yr at ≈ -2.5^oC, 52^oN), through the black spruce lichen forest (0.05 kgC/m²/yr) and the black spruce moss forest (0.2 kgC/m²/yr), to the northern edge of balsam fir-white birch forest (0.4 kgC/m²/yr at ≈ +2.5 oC, 47.5^oN). Primary production continues to increase from T_{avg annual} 2.5 to 7.5^oC (0.49 kgC/m²/yr, 45^oN at the southern edge of Quebec), but does so more gradually and asymptotically. Because the climate-productivity gradient assumes a logistic-like form, climate change is expected to cause the largest absolute changes in primary production at intermediate temperatures and latitudes, whereas % difference in productivity due to climate change is predicted to be highest in the far north and decline southward (Figure 33B), similar to diversity predictions (Figure 32B).

Primary production estimates are important because they reflect rates of production at the base of the trophic pyramid, which is a key functional determinant of secondary production [40] and the total number of trophic levels that can be supported [41]. Primary production estimates are also useful because they tend to be highly predictive of species diversity, including tree and plant diversity [42, 43] and the diversity of consumers [44, 45], often outperforming climate variables as predictors of total diversity and community structure. However, primary production estimates are not directly relatable to prediction of local food production now and in the future, in part because most local food consists of fish, birds, and mammals, which are not primary producers, and because primary production estimates are dominated by the tallest and most abundant vegetation type (e.g., herbs and grasses in herbaceous tundra, dwarf shrubs in shrub tundra, conifer trees in the boreal forest), which although sometimes used as food or medicine tend not to be primary food species.

Secondary production is a measure of biomass production by heterotrophic consumers in a system. Secondary production often considers only the production realized by plant-eating herbivores (with the production of carnivores consuming herbivores referred to as tertiary production) but in some cases secondary production is defined to include the production of all heterotrophs (the sum of production occurring at all trophic levels above primary production). Secondary production is more difficult to measure and less often estimated, in part because it is not as amenable to remote sensing. However, estimates of secondary production are directly relevant to prediction of local food availability, since the tissues of fish, mammals, and birds are primarily what people consume. Production rates are also directly relevant to the harvest of wildlife as food, since sustainable yields depend more on the annual production of a population or trophic level rather than on its overall biomass or density [46]. As long as annual harvest does not exceed annual production, at least in theory, a wildlife population can be harvested in perpetuity without causing a decline [47, 48]. Thus, our ability to describe and assess secondary production in Eeyou Istchee and Nunavik, and how it will be affected by climate change, represents a promising avenue to characterizing ecological production as a supporting condition in the local food value chain. As is always the case, models and prediction are only as good as the logic and information that goes into them, and the following preliminary estimation of secondary production is informed by limited data, no ground truthing and is restricted to the terrestrial (not aquatic, coastal, or marine) realm. It is intended and included here as an illustration of a possible approach not as a research finding or climate change prediction.

McNaughton et al. [40] assembled literature-reported ecosystem-level values of net primary production and estimated secondary production from many terrestrial ecosystems distributed around the globe. Production estimates at trophic levels higher than herbivores were too rare in the literature and thus were excluded. Thus, these estimates of secondary production are limited to production of plant-eating herbivores. Secondary production was shown to be positively correlated with net primary productivity, but different habitat types and consumer categories were characterized by unique intercepts, which equates to different ecotrophic efficiencies (secondary production : primary production ratios). We derive the following equation relating expected secondary production (expSP) to net primary production (NPP; both in units kJ/m²/yr) for the 20 vertebrate-dominated terrestrial ecosystem estimates included in figure 7.2 [40].

expSP = 0.00007*NPP^1.2082 (eq. 1)

The applicability of this prediction to a Quebec-focused application is enhanced by the similar range of NPP values that the 20 data points in McNaughton’s [40] figure 7.2 span (850 - 16,300 kJ/m²/yr) compared to the range of NPP observed from northernmost to southernmost Quebec (390-19,110 kJ/m²/yr; units re-expressed as kJ to facilitate comparison).

We then proceeded with four additional conversions or assumptions intended to render these estimates of secondary production more relatable to local food availability. First, we re-expressed kJ/year as kcal/day since kcal are unit directly relatable to people’s daily caloric requirements. The recommended dietary allowance (RDA) for an adult male is often assumed to be 2500 kcal and for an adult female 2000 kcal, so we averaged the two values to arrive at a daily per person food requirement of 2250 kcals.

Second, we expanded the area-based measure of m² to 550 km². Selection of this amount of land area was motivated by Eeyou Istchee’s trapline and tallyman land tenure system. Across more than 300 traplines delineated within Eeyou Istchee, trapline area varies from < 70 km² to > 14,600 km², but most traplines are > 500 km² and < 1,000 km² and, within that range, most are > 500 km² and < 600 km². Thus 550 km² is used here as reflective of a typical Eeyou Istchee trapline. We recognize that this trapline way of thinking about and communicating land area has limited relevance in Nunavik, but we hope a roughly 24 x 24 km area of land represents something that can be envisioned there too.

Finally, we assumed that 5% of the secondary production on a trapline-sized area might be available to a land user to be harvested and consumed. This is, in many ways, nothing more than a wild guess, but the following explains four subtractions that caused us to arrive at 5%. First, measures of secondary production include all tissues that animals produce, whereas harvesters consume most but not all parts of the animals they harvest. JBNQNHRC [49] did a lot of work to arrive at edible yield estimates for all the species included in the 1970’s Cree harvest survey. For most species, it was estimated that 50-70% of total body weight was consumed as food. If we use 60% as an average, then 40% of secondary production is lost to non-edible yield (100%-40%=60%). These non-edible parts may be used in other ways, including traditional uses of hides, hair, feather, and bones, but from a food and nutritional perspective they are a fraction of secondary production that is not consumed. Second, our estimate of secondary production includes every species, every population, and every individual herbivore present in the area; not only moose, beaver, porcupine, and grouse, but also insects, rodents, and songbirds.

As described in section 2.2.1, most species are small-bodied and most species are rare, meaning that much of the secondary production may occur in non-harvested species. Restricting our calculations to systems that McNaughton et al. [40] classify as large-vertebrate-dominated has already helped to reduce the amount of secondary production occurring in unlikely to be harvested body sizes and taxonomic groups, but it seems to us reasonable to assume that, even in large-vertebrate-dominated systems, about one third of secondary production will be produced by non-harvested species (60% * 2/3 = 40%). Again, this does not mean these species do not have value, but from a food and nutritional perspective they are a fraction of secondary production that is not consumed. We then assumed that 50% of the annual production of harvested species would be harvested by people (40% * 0.5 = 20%), whereas the the other 50% would be consumed by other predators on the landscape. A final subtraction was informed by thinking about the size and accessibility of traplines (or 24 km x 24 km areas) and the recognition that even well-used and intensively harvested areas have places that cannot be reached and times of year that cannot be harvested. We assumed 75% of harvestable secondary production is lost to inaccessibility (20% * 0.25 = 5%). This is not equivalent to assuming 25% of the area is visited on 100% of the days of the year or 100% of the area is visited on 25% of days. Rather this 25% value assumes that over the course of a calendar year, 25% of the area’s productive habitats and productive populations are accessible and harvested when the weather and time of year is right. A key consideration in interpreting the results presented here is that analyses are restricted to observation and estimation of terrestrial production arising from forest and tundra habitats as well as wetlands. Thus, these production estimates exclude aquatic and coastal production, as well as migratory species, all of which are important contributors to local food systems in Eeyou Istchee and Nunavik.

With these conversions and assumptions, we can now re-express estimated secondary production as local food production, expressed as the number of people whose nutritional needs could be supported by a 550 km² area (shortened as people fed/550 km²). Expressing this Quebec local food production gradient in relation to average annual temperature (Figure 34) indicates local food production is lowest at the northern limit of Nunavik, providing for <2 person per 550 km² area, where T_{avg annual} is -7.5^oC, then gradually but exponentially increases with increasing temperature. Local food production is characterized by the largest absolute increases within a T_{avg annual} range of -4^oC to +2.5^oC, providing for 5 people per 550 km² at -4.0^oC, 55^oN increasing to 67 people per 550 km² at +2.5^oC, 47.5^oN. Local food production continues to increase from T_{avg annual} 2.5 to 7.5^oC (83 people per 550 km²; 45^oN at the southern edge of Quebec), but does so more gradually and asymptotically. Because the climate-local food production gradient assumes a logistic-like form, climate change is expected to cause the largest absolute changes in local food production at intermediate temperatures and latitudes, whereas percentage difference in food production due to climate change is predicted to be highest in the far north and decline southward, similar to diversity and primary production predictions (Figure 34B).

Here we have modeled the supporting condition of ecological production and its contributions to the local food value chain, now and potentially in the future, using first a species diversity approach and second an ecosystem function, trophic transfer approach. As stated previously, models and predictions are only as good as the logic and information that goes into them. There are too many assumptions and approximations involved in these analyses and projections to consider them as observations or expectations. More generally, all climate envelope approaches that use contemporary spatial climate gradients to infer future responses to climate change are susceptible to correlation ≠ causation.

Just because contemporary conditions can be expressed in relation to and appear well predicted by climate variation, does not necessarily mean that responses to future climate change will reflect current climate relationships. Ecological models parameterized from modern observations are unlikely to accurately predict ecological responses to novel climates; there will be surprises [50]. These limitations notwithstanding, the modelling approaches and outcomes presented in this section emphasize an important general property of (at least contemporary) ecological systems. Globally, nationally, within Quebec, and within Eeyou Istchee and Nunavik, warmer climates generally support higher species diversity and more primary and secondary production. As climate change warms northern Quebec, part of the system responses to climate change are likely to involve increased rather than decreased diversity and increased rather than decreased ecological production. To be clear, climate change will lead to species extinctions and declines [51], these losses and impacts are already occurring and will continue to accelerate as the pace and extent of warming accelerates. But, in addition to these losses, the impacts of climate change on local food systems, particularly in cold climate regions like northern Quebec, are also likely to involve increased productivity of boreal and tundra ecosystems and the appearance of more southern species in these northern systems.

The analyses summarized in Figures 32-34 also help to communicate the magnitude of climate change expected during this century. The horizontal arrows in the bottom panels of these figures reflect projected climate change (for the 2040-2071 time horizon and a high emissions scenario) relative to the scale of climate difference from Quebec’s most southerly and northerly points. Quebec is a large province, and southern Quebec and northern Quebec are very different - climatically, ecologically, and culturally. The climate change projection arrows do not extend across the entirety of that climate space, but neither do they appear as only minor shift within that large climate space. The magnitude of expected change this century is equivalent to differences present now across about 5^o of latitude in southern Quebec increasing to about 8^o in the northern Nunavik. In other words, climate projections are suggesting that by the middle of this century, Salluit becomes more like Kujjuarapik, Kujjuarapik and Whapmagoostui more like Waskaganish, Waskaganish more like Waswanipi, and Waswanipi more like Val D’Or.

3.5.5. Modelling climate change responses: a new initiative

A substantial and dedicated emerging project, focused on the Nunavik portion of northern Quebec, led by Tiff-Annie Kenney, Frederic Maps, and Melanie Lemire from Laval University and involving our research team as collaborators has secured funding support from Sentinel Nord and Institut Nordique du Quebec to advance this area of research. The funded project is titled Sustainable and resilient country food systems for future generations of Nunavimmiut - promoting food security while adapting to changing northern environments and seeks to address the following question: “With accelerating climate change impacts, and increasing community demographics in Nunavik, will the quantity and nutritious quality of country food remain sufficient to support Nunavimmiut food security and health?” The project will weave together ecological and public health approaches to address the complex issue of food security of Nunavimmiut in this era of rapid environmental and socio-economic changes. Impacts of environmental change on northern ecosystems will be assessed by developing dynamic ecosystem models for each coastal region of Nunavik: Hudson Bay, Hudson Strait and Ungava Bay. This will establish a baseline against which the synergy of changing food webs and country food harvest sustainability under shifting environmental and societal conditions will be assessed. This coastal focus will help to broaden the terrestrial production focus on the preliminary analysis presented in this report. Leveraging existing data, predicted changes in biomass and nutritional quality of focal species will feed models about Nunavimmiut diet and food security to study how ecological responses may affect nutrition, food security and multiple dimensions of health of current and future Inuit generations. The coupling of both approaches will facilitate assessment of the impacts of societal changes (demography, food preferences, etc.) on country food supply and demand and thus the sustainability of projected harvests. The intent of the research is to i) enhance our understanding of the impacts of climatic and anthropogenic disturbances on Nunavik food webs in relation to climatic and societal changes and ii) improve our ability to explain and act on the links between the northern environment (via country foods) and health of Nunavimmiut, by co-developing adaptation strategies to foster food security and healthy eating in this Arctic region.

3.5.6. Summary: modelling biodiversity & food production responses to climate

Total diversity of fishes, birds, and mammals forms a strong climate diversity gradient across Quebec, declining from 244 species at the southern edge of Quebec, where T_{avg annual} is 6.9^oC, to 70 species at the northern edge of Nunavik, where T_{avg annual} is -7.5^oC. Rates of primary and secondary production also decline across this climate gradient. If these climate-diversity and climate-production relationships persist in a climate changed future, species diversity and ecological production is likely to increase across much of Quebec in response to warming. The absolute and the percentage change in diversity an production is expected to be much greater in northern Quebec than in southern Quebec, because i) climate change exposure (Δ^oC) will be greater in the north than the south, and ii) the slope of the relationship between diversity and climate (sensitivity; # species / ^oC) is steeper in the north than in the south. Poleward range expansions of temperate-zone species represents the most likely and most immediate mechanism by which boreal and arctic diversity might increase. As a result, the environmental impacts of climate change in temperate and polar regions are likely to be shaped by the appearance of new species and increased productivity in addition to the disappearance of current species. Arctic and boreal species most likely to go extinct or experience population declines are those that will be most negatively impacted by species expanding from the south and whose capacity to expand poleward is geographically constrained. However, climate change predictions derived from contemporary climate-diversity and climate-productivity patterns must be interpreted with caution, because they presume contemporary relationships will persist in a climate changed future and do not account for differential response times.

References on this page