2.1 Wildlife harvest and local foods

2.1.1 Wildlife harvest and local food use in northern Quebec

The nature and importance of local food systems have been characterized in a variety of ways including diet recall questionnaires [1,2], wildlife harvest surveys [3], and semi-structured interviews [4]. How local food contributes to broader issues of nutrition, health, and well-being is often accomplished by combining information on local food use with food security questionnaires and health surveys [5,6].

2.1.2. Data Sources and Methods of Analysis

The following analyses are based on harvest surveys conducted in the 1970’s in Eeyou Istchee [7] and Nunavik [8] and food frequency questionnaires conducted in the 2000’s in Eeyou Istchee [9,16, 26] and Nunavik [10]. Wildlife harvest surveys generate counts, or estimates, of the number of animals by category taken by a specific group of harvesters during a specific time period, the results of which reflect reporting accuracy, bias and methods of extrapolation, the socio-economic circumstances of harvesters, and the abundance and accessibility of wildlife populations [11]. The harvest data from Eeyou Istchee and Nunavik presented here arose from an agreement, reached during negotiations leading up to the signing of the James Bay and Northern Quebec Agreement, to jointly fund and conduct studies intended to quantify present levels of harvesting of wildlife by the Cree and Inuit of Northern Quebec. Food frequency questionnaires are a method used to collect dietary data, usually intended to establish relationships between consumption patterns and health indicators, that focus on a context-specific food list and ask respondents how often each is eaten (e.g., time per day, days per week, days per month, etc. [12]. In some cases, food frequency questionnaires include questions about typical servings or portion size, allowing estimates of food frequency to be converted into estimates of food intake (e.g., [13]). Although food frequency questionnaires provide direct estimates of food use, and, in some cases, the amount consumed, they are also recognized to be influenced by recall and reporting inaccuracies and biases [14]. The food frequency data from Nunavik and Eeyou Istchee presented here arose from two independently conducted health surveys, both of which intended to assess the overall health of the Cree and Inuit population, including contaminant exposures through consumption of local food species.

As described in [7], Cree harvest of 32 species groups were estimated from 1972-73 to 1978-79 by means of diary/calendars and questionnaire/interviews conducted at the community level. A total of 4,525 interviews were completed during the study period. Harvests were reported for defined geographical areas and periods of time. Here we focus on annual reported harvest of the 32 species for each of the eight Cree communities included in the harvest study. The body size and consumed edible fraction of harvested species, as well as community population size were also reported by [7], permitting estimation of annual reported harvest in individuals per year, edible kg per year, and edible kg per person per year.

As described in [8], Inuit harvest of 30 species groups was estimated from January 1976 to December 1980 by the Northern Quebec Inuit Association using diary or calendars and questionnaires or interviews conducted at the community level. Here we focus on annual reported harvest of the 40 species groups for each of the 13 Inuit communities included in the harvest study. Unlike the Cree harvest study [7], [8] does not include estimates of the body size and consumed edible fraction of harvested species. Because many key species harvested by Inuit in Nunavik are rarely harvested by Cree in Eeyou Istchee, [7] did not provide adequate information about the average body masses and edible yield of species harvested in Nunavik. Accordingly, edible yield mass for species harvested in Nunavik was obtained from [15]. Combining information from these three sources permitted reporting annual reported harvest estimates for the 40 species categories in individuals per year, edible kg per year, and edible kg per person per year.

The Nituuchischaayihtitaau Aschii Environment-and-Health Longitudinal Study in Eeyou Istchee [9,16,26] was designed to i) investigate health effects in relation to lifestyle, environmental contaminants exposure, and diet; and ii) investigate the links between wildlife health, quality of aquatic environments and human health. Surveys were conducted in during two- to four-week period during the spring and/or summer in Mistissini in 2005, Eastmain and Wemindji in 2007, Chisasibi and Waskaganish in 2008, and Waswanipi and Whapmagoostui in 2009. A random sample of participants, with sample size scaled to community population size, was selected from each community. Total number of participants ranged from 150 to 288 individuals per community, depending on community size and participation rates, and across all communities a total of 1,405 persons from all age groups participated. The study component relevant to analysis presented here was an assessment of the intake frequencies of traditional foods. This assessment was achieved through structured interviews and specifically designed dietary questionnaires. Dietary intake was assessed using a semi-quantitative food frequency questionnaire of traditional foods, which took into account seasonal variations. The survey documented use across 49 local food categories representing different species (e.g., Category 22. Sturgeon) or species groups (e.g., Category 24. Red or white sucker) as well as different parts and cooking methods of a single species (e.g., Category 4. Moose meat dried; 5. Moose meat cooked; 6. Moose liver or kidney). Participants were asked “In the past 12 months, did you eat any of the following animals” and “If yes, how often did you eat these animals for each of the following seasons?” and, in the 2005 Mistissini survey only “What is your usual portion size?”. Results from this survey are presented as a series of community-by-community tables (Tables A7.2A-D in [9,16,26]), with seasonal responses combined into annual totals representing the % of respondents reporting consumption of the food category and, if consumed, the days/month that it was consumed. Percentages and days/month are presented separately for girls (<19 years of age), boys (<19 years of age), women (≥19 years of age), men (≥19 years of age), as well for the total population. Focusing on the total population estimate, we multiplied the percent consuming by days per month (among those that reported consumption) to obtain a total population estimate of days per month consumption, then multiplied by 12 to obtain a days per year (d/y) estimate, and then multiplied by an average daily portion of 174g to obtain a consumed kg per person per year (kg/p/y) estimate. Although portion sizes for all or some of the food categories appear to have been estimated as part of the 2005 Mistissini survey [16], these results were not included in [16] or [9,16,26] and we used the average daily portion size across all local foods reported in [13]. To facilitate species comparisons, we summed kg/p/y estimates across food categories representing different parts or cooking methods of the same species, which reduced the number of species categories to 24.

The “Qanuippitaa? How are we?” Nunavik Inuit Health Survey was a cross-sectional study conducted in the fall of 2004 (August 27 to October 1) on a representative sample of the Inuit population (889 adults aged between 18 and 74 years old) present in the 14 communities of Nunavik [10]. As part of this survey, questionnaires were administered to document socio-demographic data, lifestyle habits, and food consumption frequencies. A food frequency questionnaire documented the frequency of consumption of country foods and market foods during the year prior to the survey. The country food portion of the questionnaire documented use across 49 local food categories representing different species (e.g., lake whitefish), species groups (e.g., trout and salmon), different parts and cooking methods within a single species (e.g., beluga mattaaq, beluga blubber, beluga meat), and preparation methods inclusive of multiple species (e.g., Pitsik made from Arctic char, brook trout or lake whitefish). The questionnaire estimated consumption frequency of the different food items (days per year) as well as item-specific daily portion size (grams), allowing an estimate of annual consumption in kg per person per year (kg/p/y). Summary data from this food frequency questionnaire are published in [17], with survey results presented for three Nunavik regions (Hudson Bay, Hudson Strait, and Ungava Bay) and for Nunavik as a whole. However, results presented by [17] focus only on adult female respondents and the intake values presented in Table S4 are expressed per unit body mass of respondents (g consumed/kg body mass/year). We used reported adult female body masses from three regions (Table 1 of [15]) to re-express consumption values as kg per person per year.

2.1.3. Harvest and food use of different species

The following tables summarize harvest and food use data characterizing local Indigenous food systems in Northern Quebec, including harvest surveys conducted in the 1970’s in Eeyou Istchee ([7]; Table 2) and Nunavik ([8]; Table 3) and food frequency questionnaires conducted in the 2000’s in Eeyou Istchee ([9,16,26]; Table 4) and Nunavik ([10]; Table 5).

Figure 18 provides a bubble plot summary of the harvest and food use data presented in Tables 2-5. For Nunavik harvest data from 1976-80 (Figure 18A), ringed seals were, averaged across all communities, the most harvested species in the region (all community average 93 kg/p/y). Ringed seal harvest was particularly high in Kangiqsujuaq (288 kg/p/y) and Killiniq (274) and lower, relative to other communities, in Kangirsuk (40) and Kuujjuaq (19). Caribou was the next most harvested species (58), with particularly high harvest in Aupaluk (167) and Kangiqsualujjuaq (156), and low harvest in Quaqtaq (18) and Salluit (4). Beluga (49), char (37), and bearded seal (30) were the three next most harvested species, with beluga harvest particularly high in Quataq (167), Kangiqsujuaq (113), Killiniq (81), and Inukjuaq (70), char harvest particularly high in Akulivik (126), Tasiujaq (87), and Kangiqsualujjuaq (86), and bearded seal harvest particularly high in Killiniq (99), Akulivik (68), and Aupaluk (54). Among the long list of additional species that were harvested species, Killiniq’s harvest of harp seals (161) was particularly high. Nunavik food frequency results from 2004 (Figure 18B) focused only on women, were reported only regionally, and aggregated some species into groups, which reduces direct comparability. However, a major distinction between the two surveys is the reduction in importance of seal in the more recent food frequency survey relative to the earlier harvest survey. Specifically ringed seals were the most harvested species in 70’s, whereas all seals combined were only the fourth most consumed category in the 2004 food frequency data, with reported seal consumption (3 kg/p/y) substantially less than caribou (15 kg/p/y) and char (9) and slightly less than geese (5) and beluga (5). Other notable trends in the 2004 food frequency data were the consistent importance of caribou and char across all three regions, Hudson Bay communities consuming more geese, Hudson Strait communities consuming more beluga, the importance of berry consumption (which was not asked about in the 1970’s harvest data) an increasing importance of ptarmigan, from the 15^th most harvested category to the 7^th most consumed category.

For Eeyou Istchee harvest data from 1972-79 (Figure 18C), moose were, averaged across all communities, the most harvested species in the region (all community average 30 kg/p/y). Moose harvest was particularly high in Nemaska (101 kg/p/y), Mistissini (61) and Wawanipi (57) and declined from south to north along the coast (Waskaganish 20; Whapmagoostui (0). Canada geese were the next most harvested species (19) and tended to be harvested the most where moose were harvested the least. Most geese were harvested in the four coastal communities of Eastmain (40), Chisasibi (38), and Whapmagoostui (30) and Wemindji (21) and fewer geese were harvested inland in the south (Nemaska, 6; Mistisini, 5; Waswanipi 1). Beaver (17) and caribou (7) were the two next most harvested species, with more beaver harvested in more southerly communities (Eastmain 27, Wemindji 26, Nemaska 25, Waswanipi 19, Mistissini 17) and more caribou harvested in the northern most community (Whapmagoostui 29) but also Mistissini ( 15). Whitefish was the fifth most harvested species (7), and was harvested more along the coast (Whapmagoostui, 13; Wemindji, 13; Eastmain, 11; Chisasibi, 9; Waskaganish, 6 ) than inland. Among the long list of additional species that were harvested species, Waskaganish’s harvest of snow geese was particularly high (19) as was Whapmagoostui’s harvest of ptarmigan (15.) Eeyou Istchee food frequency results from 2005-9 (Figure 18D) indicated, relative to 1970’s harvest data, increased importance of geese (from 2^nd to 1^st) and ptarmigan/grouse (9^th to 3^rd) reduced importance of beaver (3^rd to 10^th). Walleye also emerged as an important food fish in Mistissini and Waswanipi, and to a lesser extent Waskaganish.

Below we present summary species icon plots for reported 1970’s harvest and 2000’s consumption from the two regions (averaged across all communities) with icons scaled to reflect variation in reported harvest or consumption amounts. Figure 19 emphasizes differences in relative importance within each survey, whereas Figure 20 maintains consistent scaling across all four to highlight differences in reported harvest and consumption levels over the two time periods and survey types.

Collectively, our analysis of the 1970’s harvest and 2000’s consumption surveys indicate some important general trends about the evolution of the local food system in northern Quebec over the 30 year period after the signing of the JBNQA. First, and most generally, the surveys indicate the continued importance of a wide diversity of fish, birds, mammals, and plants in the local food systems of Eeyou Istchee and Nunavik. Results indicate a consistent and growing importance of moose and geese in Eeyou Istchee and char and caribou in Nunavik across the two survey periods, but also important differences between regions and communities in the species that are most harvested and consumed (Figure 18,19), including north to south and coastal to inland gradients. At the same time, comparison of the estimated food amounts involved in the consumption surveys in the 2000’s and harvest surveys in the 1970’s are suggestive of a pronounced decline in per capita local food use over that 30 year period (Figure 20).

The magnitude of the change provides strong evidence of a decline, but the entirely different methodologies and points of survey focus – the 1970’s arising from year-round harvest diaries, a 5-year period, and on harvest, the 2000’s data arising from a single interview conducted on one day, recall of consumption prior to that interview, and on consumption frequency (Eeyou Iscthee) or consumption frequency and amount consumed (Nunavik) – means this evidence is indirect and almost certainly reflects differences in methods in addition to differences in the local food system.

The pronounced decline in the apparent importance of beaver in Eeyou Istchee and especially seal in Nunavik provides an example of how local subsistence economies and local food systems have recently changed. Because beaver and ringed seals are both harvested for fur as well as food, consumption surveys could be expected to indicate lower importance of these species than harvest surveys if not all animals harvested for fur are consumed by people, or if the edible yield fraction consumed of these furbearers is not always maximized, especially when amounts harvested are very high. On the other hand, there is good reason to expect that the overall per capita harvest and consumption of both species has declined over the 30 period, not unrelated to the collapse of the fur economy since the 1970’s and 2000’s. Reduced demand and price of fur over the latter part of the 20^th century has had a major impact on land-based livelihoods, greatly diminishing a subsistence economy and an opportunity to make money or trade fur for goods that has existed in the region for nearly 300 hundred years [18,19]. Increased price of fuel over the same period exacerbates the decline in the price of fur [20,21]. [22,23] has written extensively on how the collapse of the seal fur economy has impacted Inuit harvesters and the mixed economy in Nunavut. [24] make mention of related impacts in northern Quebec and Ontario. [25] quantifies the economics of subsistence harvest in Nunavik in the 1990’s. But, in general, how the collapse of the fur economy has impacted and altered the local food systems of Eeyou Istchee and Nunavik remains under-considered.

2.1.4. Situating communities in species space: a Northern Quebec analysis

​Here we complete a quantitative analysis of community differences in reported harvest and consumption, similar to that presented in section 1.7 and [2] but in this case focus on community-to-community and region-to-region differences within Eeyou Istchee and Nunavik. Figure 17 highlights the pronounced differences in reported wildlife harvest between the Cree community of Whapmagoostui, Eeyou Istchee and the neighboring Inuit community of Kuujjuarapik, Nunavik.

​Extending this comparison across more than two communities, while including all used wildlife species in the analysis, requires a multivariate statistical approach that positions communities in species space. We present this species space, which positions communities in relation to each other and the species they use the most and the least, via a principal coordinates analysis (PCoA) plot. PCoA is a multivariate approach that we used to visualize community differences in local food use. In the following PCoA plots, a community’s food use (their proportional use of various species) is represented by a single point per community positioned in the two-dimensional space created by two multivariate axes (Dimension 1 and Dimension 2). Two communities that are positioned close to each other in this PCoA space have highly similar harvest or consumption patterns, using many of the same species in similar proportions. Conversely, two communities that are positioned far from each other in this PCoA space have low reported harvest or dietary similarity involving use of different species or some of the same species but in very different proportions. Adding species vectors to a PCoA plot helps to visualize how different species contribute to the two PCoA axes and helps to visualize where communities are situated in species space. Using Figure 17 as an example, in a PCoA plot, we would expect points representing the communities of Whapmagoostui and Kuujjuarapik to be distant from each other (because they harvest very different species in very different proportions), and for the Whapmagoostui point to be generally closer to the position of Canada geese, (because they harvest geese more than any other species) and the Kuujjuarapik point to be generally closer to the position of ringed seal (because they harvest ringed seals more than any other species).

Our first analysis focuses on the reported harvest data from Eeyou Istchee and Nunavik presented in Table 2 and 3 (above). A strength of this multivariate comparison of harvest data is the direct comparability of data sources from the two regions, including similar time periods (1972-79 in Eeyou Istchee; 1976-1980 in Nunavik) and methodologies (as described in JBNQNHRC 1982, 1988). Important limitations of the harvest analysis are that the data are from 50 years ago and edible yield from harvest rates are not necessarily equivalent to consumption rates.

Accordingly, we also conduct multivariate analysis of reported food use from the two regions, as reported in Table 4 and 5 (above), despite several weaknesses inherent to this comparison, including different survey years (Eeyou Istchee, 2005-09; Nunavik, 2004), populations (Eeyou Istchee, women, men, girls, and boys; Nunavik, women only), community or regional specificity (Eeyou Istchee, community-level; Nunavik, regional), and methodology (Eeyou Istchee, food frequency; Nunavik, estimated consumption).

Multivariate analysis of the 1970’s harvest data indicates the distinctiveness of species harvested by Inuit communities in Nunavik relative to Cree communities in Eeyou Istchee (Figure 21). This distinction is indicated by the clear left-to-right separation of Nunavik and Cree communities along Dimension 1 axis, including substantial distance between Whapmagoostui and Kuujjuarapik as expected from Figure 17 (above). In general Whapmagoostui clusters with other Eeyou Istchee communities and Kuujjuarapik clusters with other Nunavik communities, particularly along the dimension 1 axis. Although the major pattern is the separation of Nunavik and Eeyou Istchee communities, interestingly and more subtly, dimension 2 axis situates the two most northerly Cree communities of Whapmagoostui and Chisasibi and the two most southern Inuit communities of Kuujjuarapik and Kuujuaq closer to each other than to more southerly Cree communities and more northerly Inuit communities. Thus, within Nunavik, the largest difference in species harvested occurs between the most southerly communities of Kujjuarapik and Kuujjuaq and the northern Hudson Strait communities of Quataq and Kangiqsujuaq. Similarly, within Eeyou Istchee, the largest difference in species harvested occurs between the most northerly, coastal communities of Whapmagoostui and Chisasibi and the most southerly, inland communities of Waswanipi and Mistissini. For Eeyou Istchee in particular, Dimension 2 axis perfectly situates the five coastal Cree communities along a north-to-south gradient, with Whapmagoostui most similar to Chisasibi, Chisasibi intermediate to Whapmagoostui and Wemindji, Wemindji intermediate to Chisasibi and Eastmain, Eastmain intermediate to Wemindji and Waskaganish, and Waskaganish most similar to Eastmain. The three inland Eeyou Istchee communities - Mistissini, Nemaska, and Waswanipi - form a cluster in species space, again indicating that these three communities harvest similar species in similar proportions.

When interpreting the outcomes of these multivariate analyses, it is important to realize there is no geographical or ecological information entered into the analysis. Communities are sorted and positioned solely according to their reported harvest patterns. Thus, any emerging patterns are based entirely on similarities and dissimilarities in species harvested, not where the communities are physically located. Comparing community locations on these plots to the significant species vectors defining this species space indicates southern, inland Eeyou Istchee communities (e.g., Mistissini, Nemaska, Waswanipi) harvest more moose, walleye, sturgeon, pike, otter, suckers, beaver, black bear, lynx, and grouse than northerly coastal communities (e.g., Whapmagoostui, Chisasibi), which harvest more brook trout, geese, whitefish, porcupine, burbot, and hare. In Nunavik, the most northerly communities situated along the Hudson Strait harvest more beluga, seal, and walrus than more southern communities, which harvest more char, caribou, sculpin, and snowy owls.

Multivariate analysis of food use reports also indicates the distinctiveness of species consumed by Inuit communities in Nunavik relative to Cree communities in Eeyou Istchee. This distinction is again indicated by the clear left-to-right separation of Nunavik and Cree communities along Dimension 1 axis (Figure 22). In the case of food use data, dimension 2 axis situates the two most northerly Cree communities of Whapmagoostui and Chisasibi closer to Hudson Bay communities and more distant from Hudson Strait and Ungava Bay communities. Within Eeyou Istchee, Dimension 2 suggests a large difference in food use between the southerly, inland community of Waswanipi and the northern, coastal communities of Chisasibi and Whapmagoostui, but beyond these extremes, the south to north sorting correspondence is less clear than it was for the harvest data.

References on this page