Experimental design and data collection

Experimental design and data collection

Problématique :

Des changements au niveau de la végétation, des températures plus élevées, et des conditions hydrologiques plus sèches sont observés par les Inuit dans l’Arctique canadien (p. ex. Thorpe et al. 2002; McDonald et al. 1997; Ford & Community of 19loolik 2006; Jolly et al. 2002; Govt. of Nunavut 2005). L’augmentation des températures globales affecte fortement les régions arctiques où des changements de végétation sont prévus (Walker et al. 2006) et déjà détectables par endroits (Myers-Smith et al. 2011 ; Elmendorf et al. 2012). Cependant, peu d’information est disponible sur la résistance des communautés végétales aux changements climatiques (Hudson & Henry 2010). De tels changements pourraient notamment influencer la composition, la structure, et la dynamique des communautés végétales (Aerts et al. 2006; Wookey et al. 2009), ainsi que les interactions entre espèces (Aerts et al. 2006; Parsons et al. 1994; Chapin et al. 1992, 1995). Présentement, une augmentation du couvert des arbustes érigés en Alaska (Tape et al. 2006), dans l’Arctique canadien (Ropart et Boudreau 2012; Lantz et al. 2010; Tremblay et al. 2012) et ailleurs dans le monde est de mieux en mieux documentée (Elmendorf et al. 2012; Myers-Smith et al. 2011). Au Nunavik:, Betula glandulosa est l’espèce responsable de plus de 90 % de l’augmentation du couvert d’arbustes observée entre 1960 et 2003 (Tremblay et al. 2012; Ropart et Boudreau 2012). Ceci pourrait avoir un impact négatif sur la distribution et la productivité des arbustes nains producteurs de petits fruits, car ceux-ci sont sensibles à la compétition pour la lumière et les éléments nutritifs (Press et al. 1998).

Pour évaluer l’hétérogénéité des changements temporels de la végétation dans les régions vastes et isolées, la télédétection est un outil utile. La carte « Circa 2000 Land Cover Map », développée à partir de l’indice de végétation de la différence normalisée (NDVI) d’images Landsat de résolution moyenne (30 m), caractérise le couvert végétal selon les formes de croissance dominantes pour 15 zones à travers l’Arctique canadien (Olthof et al. 2009). Cette carte nous a permis de choisir nos sites en fonction de la présence d’arbustes producteurs de petits fruits. Ensuite, sur le terrain, l’échantillonnage nous a fourni une caractérisation détaillée du couvert végétal.

Data analysis:

For statistical purposes, each of the four shmb species which grew over 40 cm tall (Betula glandulosa, Salix alaxensis, S. glauca, S. planifoUa) was treated as either tall (> 40 cm) or low « 40 cm), so that in any given plot, there could be both taU and low slmlbs of the same species. Consequently, the vegetation data matrix consisted of 103 taxa. Given the multivariate nature of our data, an indirect ordination analysis was performed using CANOCO 4.53 (Microcomputer Power, Ithaca, NY). We used a detrended correspondence analysis (DCA, CANOCO 4.53, ter Braak & Smilauer 2004) to detect the patterns in the relations among the species we sampled (n = 103) and passive1y with the environmental variables available (n = 30). Rare species were downweighted to lessen the influence they would have on the final result (Jongman et al. 1995). A scatter plot permitted us to visually evaluate the grouping of different plant communities.

Results :

Vegetation :
Ninety-five vascular species and four classes of cryptogams (crustose, fruticose and foliose lichens, and bryophytes) were identified during the course of our field season. Prostrate dwarf shrub vegetation clearly dominate the landscape around Baker Lake and is classified as Zone 7 on the Circa-2000 Map of Northem Canada (Olthof et al. 2009). While we tried to sample as many shrub sites as possible in the different shrub zones (Z3, Z5, Z6, Z7), 35 of the 83 sites sampled belonged to Z7 (Table 2.2). Consequently, Zones 3, 5 and 6 may be less well represented in our dataset.

Berry productivity :
There was no significant difference in the berry weight of either species across the zones, however, V vitis-idaea berries were significantly smaller in 2010 than they were in 2009 (p = 0.004). E. nigrum showed the same tendency, however the difference was marginal (p = 0.055) (see Appendix D).

Productivity values of both E. nigrum and V vitis-idaea varied significantly between the two years we sarnpled them (p < 0.001). For both species 2009 was a year of higher productivity than 2010. Productivity tended to be higher in E. nigrum than V vitis-idaea for both years (5-112 g/m2 and < 1-87 g/m2 respectively in 2009; < 1-22 g/m2 and < 1-11 g/m2 in 2010). V vitis-idaea was most productive in Zone 3 and E. nignlm in Zone 5, yet there was no significant difference in the mean productivity of either species across the different zones.

Discussion :

Enviromnental variables, in particular topography, snow cover, soil moi sture and soil pH (Gould & Walker 1999) and increased shrub cover (Pajunen et al. 2011) have been correlated with plant species richness, which varied from quite low in the tall shrub community to almost three times higher in grarninoid communities. The greatest total number of species was recorded in the prostate dwarf shrub zone, where vascular plant cover was less than half of what it was in the tall shrub zone. While this may in part be a consequence of our sarnpling so many sites in this zone, the greater heterogeneity of this zone is visible by the greater spread on the DCA. Meanwhile, tighter grouping on the DCA reveals how much more homogeneous the sites within the grarninoid and tall shrub communities are.

The DCA also showed a clear moisture gradient along the first axis, suggesting that sites in the dwarf shrub zones were more poorly drained than those in the pro strate dwarf shrub zone, as evidenced by their left leaning position along the horizontal axis. This demonstrates the importance of topographic position and drainage characteristics in Arctic plants (Tedrow 1977 in Bliss & Svoboda 1984), and could possibly explain the lower mean cover and more prostrate growth habit of the shrubs in this zone.

More than half of all northern species are nonvascular (Matveyeva and Chernov 2000), and lichens are an important group in the Arctic for their use as winter forage for caribou and the ability of sorne to fix nitrogen in strongly nitrogen limited systems (Callaghan et al. 2004). Mosses dominated in the mesic regions where clay and silt levels in the soil were moderately higher while lichens became more abundant in the sandier, better drained zones. Experimental data have shown that lichen and bryophyte cover decrease with increasing temperatures, likely as a result of shading by vascular species such as shmbs, but also grasses and sedges, which tend to increase in height and cover in warrner conditions (Comelissen et al. 2001 , Walker et al. 2006). Interestingly, the EIders of the community have noted an increase in graminoid species around Baker Lake (Gérin-Lajoie & Spiech 2009, unpubl. data). Lichen and moss identification would have been a positive addition to the plant list compiled in our study, and would have also influenced the results of our DCA, possibly helping to distinguish differences amongst the species found in the prostrate dwarf shrub community from those found in the other two dwarf shrub communities.

Vaccinium uliginosum is suited to both moist and dry soil conditions (Aiken et al. 2007). Though it seemed to prefer the dwarf shrub community, where moss cover was high, it also did well in the prostrate shrub zone, where the higher sand content meant better drainage of the sites and subsequently the replacement of mosses by greater lichen coverage. This seems to have been the most distinctive factor differentiating it from the other dwarf shrub zones.

In soils of the bedrock zone where E. nigrum and V. uliginosum were common, soil was restricted to small pockets where litter and nutrient-rich mn-off from the bedrock accumulated, accounting for the significantly higher organic matter and nitrogen levels. The tall shrub zone was also significantly higher in organic matter than the other zones, likely a result poorer drainage as well as the constant abundance of litter covering the ground (Weintraub & Schimel 2003). Betula glandulosa and V. vitis-idaea cover were at their highest in this zone, reflecting their preference for higher organic content (Aiken et al. 2007).

Conclusion :

Given our sampling strategy, with a bias on sites within the prostrate dwarf shrub communities, and avoiding wetlands and non-homogeneous areas, we have not identified the full diversity of vegetation around Baker Lake. Future studies could focus on plant communities po orly represented here. More extensive travel along the Thelon River, to taller shrub and more mesic sites would help complete the botanical picture and may have yielded greater numbers of Rubus chamaemorus. Shrubs and cryptogams dominated the landscape, and it is clear that the berry producing shrubs, while ubiquitous, vary greatly in their spatial distribution. Betula glandulosa also represented significant cover across the territory, and this may be a worthy incentive to monitor changes in the vegetation. We recommend that the sites be re-evaluated every five to ten years to track changes in the vegetation structure and diversity, and more specifically to the berry producing shrubs. AIso, an annual harvest of the berries in our sites would provide invaluable long-tenn data on the productivity of the berry producing shrubs.

Table des matières

Introduction
Problématique
Objectifs
Méthodologie
Site d’étude
Espèces étudiées 
Les entrevues
La végétation
La récolte de petits fruits
Résultats
Végétation
Petits fruits l.6 Discussion et conclusions
CHAPITRE II
BERRY SHRUB DISTRIBUTION AND PRODUCTIVITY OF
EMPETRUM NIGRUM L. AND VACCINIUM VITIS-IDAEA L. IN THE
VICINITY OF BAKER LAKE, NUNA VUT, CANADA 
Résumé
Abstract
Introduction
Methodology
Study site
Study species
Experimental design and data collection
Data analysis
Results 
Vegetation
Berry productivity
Discussion
Conclusion

Télécharger le rapport complet