Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Berry production and red-backed voles at Kluane Lake, Yukon Territory Cowcill, Kevan Anthony 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-0170.pdf [ 5.8MB ]
Metadata
JSON: 831-1.0092514.json
JSON-LD: 831-1.0092514-ld.json
RDF/XML (Pretty): 831-1.0092514-rdf.xml
RDF/JSON: 831-1.0092514-rdf.json
Turtle: 831-1.0092514-turtle.txt
N-Triples: 831-1.0092514-rdf-ntriples.txt
Original Record: 831-1.0092514-source.json
Full Text
831-1.0092514-fulltext.txt
Citation
831-1.0092514.ris

Full Text

B E R R Y PRODUCTION A N D R E D - B A C K E D V O L E S A T K L U A N E L A K E , Y U K O N TERRITORY  by  K E V A N ANTHONY COWCILL  A THESIS SUBMITTED IN P A R T I A L F U L F I L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF , M A S T E R OF SCIENCE  in  T H E F A C U L T Y OF G R A D U A T E STUDIES  (Zoology)  T H E UNIVERSITY OF BRITISH C O L U M B I A April 2006 © Kevan Anthony Cowcill, 2006  Abstract Northern red-backed vole populations peak two to three years after snowshoe hare populations peak. Snowshoe hares cycle on a 9 to 11 year cycle and produced a large quantity of fecal pellets in their peak years. The fecal nutrient enrichment hypothesis surmises nitrogen (N) that is released from these pellets is captured by shrubs whose berries are critical food sources for the red-backed voles. These shrubs use the N to produce a large crop of berries that will provide an overwintering food supply for the voles, and reduce their overwintering mortality, resulting in an increase in red-backed vole densities in the spring. To simulate N levels provided by hare pellets I added 0.5, 1.0, 2.0, 4.0 and 17.5 g N / m to 60 plots each of Empetrum nigrum, Arctostaphylos 2  rubra, Arctostaphylos uva-ursi, Shepherdia canadensis, and Geocaulon lividum in 2004 and 2005 to determine if I could produce an abundant berry crop in 2005. Only E. nigrum had a significant increase in berry production at 1.0 g N/m . Weather conditions in 2005 were probably 2  responsible for the lack of significant response in the other plants. Data from 9 years of berry production indices and northern red-backed vole numbers indicate a strong positive correlation (r=0.92) between the combined berry production of E. nigrum, S. canadensis and A. rubra and vole densities in the following years.  ii  Table of Contents Abstract Table of Contents List of Tables...! List of Figures List of Illustrations Acknowledgements Chapter 1: Berry Production in Boreal Forests - A Review .• 1.1 Introduction 1.2 Do soil nutrients limit berry production in northern environments? 1.2.1 Effects of nitrogen addition on plant growth and diversity 1.2.2 Effects of N addition on berry production 1.3 How does climatic variation affect berry production? 1.4 Do berry crops affect the northern red-backed vole (C. rutilus) community? 1.5 Summary 1.6 Literature Cited  ii iii iv vi vii 1 1 1 2 3 5 8 9 11  Chapter 2: Effects of Nitrogen Addition on Berry Production 2.1 Introduction 2.2 Methods 2.2.1 Study area and experimental sites 2.2.2 Experimental sites 2.2.3 Analyses 2.3 Results , 2.4 Discussion 2.4.1 Berry production 2.4.2 Plant growth 2.5 Conclusions 2.6 Literature Cited  17 17 18 18 18 21 21 23 23 24 25 325  :  v  Chapter 3: Do changes in berry production drive Clethrionomys rutilus population changes? ... 38 3.1 Introduction. 38 3.2 Methods 41 3.2.1 The study area 41 3.2.2 Trapping methods 41 3.2.3 Berry production indices 41 3.2.4 Statistical analysis 42 3.3 Results 43 3.3.1 Berry production 43 3.3.2 Relationship between berry production and vole numbers 43 3.4 Discussion • 43 3.5 Conclusion : 44 3.6 Literature Cited ..51 (  Chapter 4: Conclusions and Recommendations for Further Work 4.1 Literature Cited  iii  55 60  List of Tables Table 2.1: A N O V A table P-values, F-values for 2004 and 2005 for berry and vegetative growth of berry producing species. Degrees of freedom for vegetative growth = 54 except for Arctostaphylos rubra where d.f. - 53. n.a. = not available 29 Table 3.1: Mean berry densities and parametric correlations between berry numbers and spring Clethrionomys rutilus population densities from 1997-2005 (1995-2005 fox Arctostaphylos uvaursi) with a one year time lag (e.g. berries in 1997 correlated with voles in 1998). Degrees of freedom = 7, except for Arctostaphylos uva-ursi where d.f. = 9 48  iv  List of Figures Figure 2.4: Arctostaphylos rubra berries/m in 2004 and 2005, with standard error bars  30  Figure 2.5: Arctostaphylos uva-ursi berry numbers in 2004 and 2005 at different levels of nitrogen (N) treatment. Bars are standard error  31  Figure 2.6: Empetrum nigrum new growth (mm) in 2004 and 2005 with standard error bars  32  2  Figure 2.7: Empetrum nigrum berries/m for 2004 and 2005 with different treatment levels of nitrogen (N). Bars are standard error 33 2  Figure 2.8: Berry weight (g) in 2005 of 3 shrubs and one herbaceous plant {Geocaulon lividum). Bars are standard error 34 Figure 2.9: Geocaulon lividum berry weight and leaf width correlation (r=0.74). Bars are standard error  35  Figure 3.1: Clethrionomys rutilus spring density and snowshoe hare spring density (line) from 1973 to 2005 in the Kluane Lake region, Yukon. Clethrionomys rutilus data is derived from the following sources: 1973-1975 Gilbert and Krebs (1981); 1976-1986, Gilbert and Krebs (1991); 1987-2005 Krebs et al. (unpublished data). Hare data is from the following sources: 1971-1975, Keith (1990); 1976-1986, Krebs et al. (1986) and Boutin et al. (1995); 1987-1996, Krebs et al. (1995), 1997-2005, Krebs et al. (unpublished data). Citations from Boonstra et al. 2001. Figure courtesy of Kluane Ecological Monitoring Project 46 Figure 3.2: Berry production of 4 shrubs, (a) Arctostaphylos uva-ursi, (b)Arctostaphylos rubra, (c) Empetrum nigrum, (d) Shepherdia canadensis and 1 herbaceous plant (e) Geocaulon lividum from the Shakwak Trench region of south-western Yukon. Geocaulon lividum can be seen increasing and decreasing out of phase with the dwarf shrubs ..47 Figure 3.3: Total mean berry production of the shrubs/m Shepherdia canadensis (n=100/yr), Arctostaphylos rubra (n=100/yr), and Empetrum nigrum (n=100/yr) (line) plotted against mean number of spring Clethrionomys rutilus/ha (histogram) measured from 3 small mammal trapping grids between 1997-2005 at Kluane Lake, south-west Yukon Territory. Note there is a 1-year lag between high berry production and C. rutilus densities (r = 0.92, 9d.f.) (i.e. berries at time t, and voles at time t+1) , 49 2  Figure 3.4: Results of the predictive equation of Clethrionomys rutilus densities compared with actual Clethrionomys rutilus densities (r = 0.92). Densities were predicted using the following equation: Density of C. rutilus = 0.23 Empetrum nigrum -1.31 Arctostaphylos rubra + 0.06 Shepherdia canadensis with P = 0.002, P = 0.005, P = 0.007 for each of the shrubs respectively, where indices of berry production are used for each shrub. Bars are standard error 50  v  List of Illustrations Figure 2.1: General location of the study site. Image adapted from 2005 TeleAtlas, Image 2005 EarthSat, GoogleEarth 26 Figure 2.2: Arctostaphylos uva-ursi plot 60 (U-60), diameter 1.15 m. North is located at the top of the picture 27 Figure 2.3: Vinyl quarter-circle used for measuring percent cover of vegetation within the 1 m plots. R l to the apex of the quarter-circle covers 11% of the vinyl. R l to R2 covers 33% and R2 to R3 covers 56% of the quarter-circle. The R l are is subdivided into two sections with R2 and R3 regions subdivided into 4 sections to faciliate more accurate percent measurements 28 2  vi  Acknowledgements Thank you to my supervisors and committee members, Dr. C.J. Krebs, Dr. R. Boonstra, Dr. R. Turkington and Dr. J. Myers, for their support and excellent feedback during the field season and writing stages of the thesis. Also a thank you to Dr. D. Hik for his information on hare pellets, Alice Kenney for her database support, and Bill Miller for his weather data from Burwash Station. Elizabeth Hofer was invaluable in helping set up the project, training me, and allowing me to run off with her truck for two days at the last minute. Her help and encouragement during those first two weeks when I was tempted to quit made all the difference. Thank you, Liz. Elizabeth Gillis helped me establish the small mammal live-trapping grids, and Todd Heakes helped spread half a metric tonne of fertilizer over 11 hectares of boreal forest, checked small mammal traps and provided good company. Thanks also go out to Amanda Collins, Sylvie Mitford, Heather Milligan, and Svenja for their help in spreading fertilizer, checking small mammal traps, and measuring bits of plants. Your help made things so much easier for me. I owe a big thank you to those who made Kluane feel like home. Lance Goodwin, whose, friendly smile and bush knowledge made him always a pleasure to chat with; Sian Williams who talked me into the bike race where I had a great time; Andy Williams with his many entertaining stories could make mealtimes memorable. Jennie McLaren, Danielle Hodgson, Crystal Cerny, and Ben Gilbert also were wonderful people whose company I enjoyed. Thank you. Finally, a thank you to Bronwyn who taught me that playing in a sandbox refreshes the soul.  vii  Chapter 1: Berry Production in Boreal Forests - A Review 1.1 Introduction Growth and production of plants can be limited either from the top down by strong herbivory or from the bottom up by resources including climatic factors such as temperature and rainfall. In this study I am particularly concerned with berry-producing shrubs in the Kluane Lake region of the south-western Yukon Territory. Since berries have been produced in agricultural systems for centuries, a large body of literature has developed to suggest methods of maximizing berry crop production. I will assume for the moment that berry production is limited from the bottom up by resources and climate, and I will proceed to address the following hierarchy of questions: 1. Do soil nutrients affect berry crops? -If so, which soil nutrients are critical? 2. Does climatic variation affect berry crops? - If so, what are the key variables? 3. Do berry crops affect the northern red-backed vole (Clethrionomys rutilus Pallas) community? -If so, what are the effects? Addressing these questions leads to the main focus of my thesis, the fecal nutrient enrichment hypothesis, that hypothesizes small amounts of nitrogen (N) found in snowshoe hare (Lepus americanus) pellets when hares populations are at their peak can subsequently increase berry production in ground shrubs in the next year resulting in an increase in northern red-backed vole numbers in the following year.  1.2 Do soil nutrients limit berry production in northern environments? Lack of available nutrients controls the standing crop of vegetation, and the main limiting factors in northern forests are the major soil nutrients such as nitrogen (McKendrick et al. 1980, Turkington et al. 1998). Many studies have added nutrients such as nitrogen (N), potassium (K), and phosphorous (P) to soils to determine the effects on plant productivity, and on bottom-up community interactions (reviewed in Eaton 1994, Turkington et al. 1998, Turkington et al. 2002). In this study, I used N additions because earlier studies had determined that boreal forest soils, including those in the Kluane region, were N-limited and P and K additions had little  1  additional effect (Vitousek and Matson 1984, Bonan and Shugart 1989, Nams et al. 1993). Also, since my experiment was an attempt to increase berry production, just N was added because N combined with potassium (K) leads to fewer fruit being produced (Penney et al. 2003). Nitrogen is limited in the boreal forest because the main source of useable soil N comes from the small amounts of fixed N from aerial sources, estimated to be about 1 to 3.5 kg/ha (0.10.35 g/m ) per year for northern areas in both European and North American regions that are 2  unaffected by industrial sources (Tamm 1982, Aber et al. 1989). Other forms of N accumulate slowly from the breakdown of organic matter (Tamm 1982). 1.2.1 Effects of nitrogen addition on plant growth and diversity Nutrient addition to increase plant productivity was first developed in agriculture. However, agricultural studies typically deal with fertile soils while many soils in northern Canada are infertile. Information taken from agricultural studies have only limited applicability to plants on infertile soils, since wild plants on these nutrient-poor sites do not respond the same way to added nutrients as plants found growing in more fertile soils (Chapin et al. 1986). Ericaceous shrubs that grow in infertile soils have adapted to nutrient-poor conditions by having slow growth and continuous N uptake throughout the growing season and are Stress Tolerators (sensu Grime 1979). They can also conserve and store available nutrients for times when the nutrients may be scarce (Chapin and Shaver 1989). B y contrast, fast-growing deciduous, herbaceous and graminoid species quickly uptake N for new growth and have it stored in leaves and twigs by mid-summer (Shaver et al. 1986). These fast-growing species are able to take up and utilize N before the slower-growing species, and the higher levels of N in the leaves enable them to increase their photosynthetic rates (Hikosaka et al. 1998), which in turn enables them to outcompete ericaceous shrubs both above ground for light and below ground for nutrients and root space (McKendrick et al. 1980, Ohlson et al. 1995, Hikosaka et al. 1998). In the long-term, rapid N uptake is a winning strategy in nutrient-rich environments but a losing one in nutrient-poor environments, whereas the ericaceous shrubs' strategy of slow but continual uptake is a better adaptation in rather infertile soils. This means ericaceous shrubs on infertile soils are somewhat buffered against short-term nutrient fluctuations and will exhibit little change in growth in response to nutrient additions (Gerdol et al. 2000). Adding available N to infertile soils thus favours the herbaceous and graminoids species that outcompete ericaceous  2  shrubs, and reduce overall diversity within the area fertilized (Raatikainen and Niemela 1994, Nordin et al. 1998, Press et al. 1998, Turkington et al. 2002). Nitrogen additions can be harmful to berry-producing shrubs by increasing the incidence of disease and fungal colonization of shrub leaves (Graham and Turkington 2000, Strengbom et al. 2001, Nordin et al. 2005, Strengbom et al. 2006). Chapin and Shaver (1985) also reported that higher N levels (25 g/m ) may lead to increased mortality although low levels of N (> 5 g/m ) can significantly increase growth of dwarf shrubs. This is in contrast to Gerdol et al. 2  (2000) study who showed minimal effects of N additions. In my study high levels of N (17.5 g/m ) caused high rates of fungal leaf infection, as well as leaf mortality caused by fertilizer burn 2  on Arctostaphylos rubra (Rehd. & Wils.) Fern. At the same levels of nitrogen Empetrum nigrum L. experienced high whole plant mortality within a month from fertilizer burn, and subsequently showed no regrowth the following year. 1.2.2 Effects of N addition on berry production Effects of fungi and increased herbivorous damage along with a shift in plant diversity will cause a reduction in berry production per unit area if N is added to shrubs living on infertile soils (Raatikainen and Niemela 1994). However, a time lag of several years may exist before these effects are seen. Nordin et al. (1998) in their one-year study of adding N (0.05 to 5.0 g/m ) 2  within the boreal forest reported no significant treatment effects on above or below ground biomass of ericaceous shrubs, grasses, and mosses. Mols et al. (2000) also reported a lack of a clear response by any single species when alpine communities were fertilized for 3 years with 0.7, 3.5 and 7 g N/m , Turkington et al. (2002) did not detect any effects of 17.5 g N/m /yr until 2  2  the fifth year at which time two shrubs, Linnaea borealis L. ssp. americana (Forbes) Hulten var americana (Forbes) Rehd., and Arctostaphylos uva-ursi (L.) Spreng. s.l declined probably owing to shading by an increased number of graminoids. However, none of these studies measured the effects of N on berry production. Most work done on the effects of N additions on the reproductive output of plants comes from studies of agricultural plants in fertile soils (Hanson and Retamales 1992, Spayd et al. 1993). However, there are studies dealing with ericaceous agricultural crops, such as blueberries and cranberries (Vaccinium spp.), that grow in comparatively nutrient-poor conditions in acidic sandy soils and waterlogged soils respectively. These are more applicable to my study because  3  these kinds of soils are typical of those found in the Kluane region and in the boreal forest biome as a whole. The timing of N addition in agricultural systems is crucial for berry production. In acidic sandy soils nutrients leach out of the root zone fairly quickly to the deeper part of the soil profile where the shallow-rooted plants cannot access it (Retamales and Hanson 1989, Hanson and Retamales 1992). However, if N is applied during maximum uptake times such as when the roots and shoots are growing (usually in the first 4 weeks after flowering) the N can be incorporated into the plant before it leaches out of the soil (Hanson and Retamales 1992). Also small amounts of fertilizer were more effective than large amounts as it seems excessive nutrients can not be utilized by shrubs as easily (Lehmushovi 1975 in Raatikainen and Niemela 1994), and incidence of disease increases with increasing N additions (Mukula and Raatikainen 1983 in Raatikainen and Niemela 1994, Nordin et al. 1998, Strengbom et al. 2003, 2006 ). The timing of N addition, along with the low levels of N , are mirrored in the Kluane system with spring snow melt. Shallow-rooted shrubs such as A. uva-ursi which grow in acidic sand-gravel soils receive an influx of N in the spring melt-water at an optimal time. This N comes from decaying matter, animal waste products (McKendrick et al. 1980), and from the accumulation of aerially deposited N within the snowpack (Stoddard 1994 cited in Nordin et al. 2005). In boreal and grassland systems input of N from rodent feces and urine alone amounts to an estimated minimum of 0.3 to 1.3 g N / m per year, and during years of high vole density may 2  surpass that of ungulate contributions (Pastor et al. 1993, 1996, Baker et al. 2004, Clark et al. 2005). Snow melt-water percolating through these organic materials may act similarly to a slowrelease fertilizer and enhance plant growth (McKendrick  al. 1980) as well as possibly  increasing berry production in ericaceous shrubs (Hanson and Retamales 1992). This organic form of N and the time of its application may be quite important in satisfying a plant's N needs (Nordin et al. 2001). Ericaceous dwarf shrubs in particular are able to take advantage of this organic N as the relationship they have with ericoid mycorrhizae enables them to take up organic N (Kerley and Read 1997, Nordin et al. 2001, Persson et al. 2003). For example, in wild highbush blueberry shrubs (Vaccinium corymbosum) Hanson and Retamales (1992) increased berry yield by about 10% for 5 years through two applications of 3.8 g/m N at bud-break and again at petal fall. Similarly lingonberry (Vaccinium vitis-idaea) yield 2  4  in Finland was also increased over a 4 year period from yearly applications of fertilizer in the spring (Niittymaa 1983 in Raatikainen and Niemela 1994). Penney et al. (2003) applied 6 g N/m  2  to Vaccinium angustifolium in the second year prior to or shortly after spring flower bud swelling and increased ripe fruit yield by 65% in the same year compared with fruit yield on unfertilized plots. There are transient dynamics that could be difficult to follow in the short-term, but in general addition of small amounts of N in the spring should increase berry production in ericaceous shrubs for a few years before herbaceous plants begin to outcompete the shrubs resulting in a decrease in berry production (Paivinen 1976 in Raatikainen and Niemela 1994, Mukula and Raatikainene 1983 in Raatikainen and Niemela 1994).  1.3 How does climatic variation affect berry production? Berry yield is greatly affected by climatic factors such as rainfall, temperature and snowcover (Kuchko 1988, Yudina and Maksimova 2005), and therefore large interannual variations in climate, as well as the slow growth rate of northern species, can overshadow correlations between N-addition and berry yield (Vander Kloet and Cabilia 1996, Mols et al. 2000). Two favourable years (e.g. optimum temperature, rainfall, no late flower killing frosts) are necessary to produce a bumper crop with flowering primordia being set the first year, and a climatically favourable spring and summer in the second year, which allows flowers to develop (Kalela 1962 in Dyke 1971, Kuchko 1988). However, an extreme weather event in an otherwise favourable year can reduce berry output. For example, warm summer temperatures at Great Slave Lake in 1966 resulted in a berry bumper crop, that year and a rich crop of flower buds in the fall. These buds failed to produce an abundant berry crop the following year which was attributed to a delayed early spring period in which the snow cover stayed on the ground till mid-May (Dyke 1971). Kuchko (1988), in a 10year study, reported that spring frosts during flowering of cowberry (V". vitis-idaea) and bilberry (Vaccinium myrtillus) decreased berry production in four of the years by killing the flowers. In Finland V. vitis-idaea and V. myrtillus yield decreased by about 31% relative to previous average long-term production over a 5-year period despite yearly fertilization as extreme winters and spring frosts reduced the berry yield of both shrubs in the 3 and 4 year of rd  th  the study (Mukula and Raatikainen 1983 in Raatikainen and Niemela 1994). Therefore, flower  5  buds can be killed by frost damage in the late fall or early spring, or when there is inadequate winter snow cover combined with cold winter temperatures, as well as during prolonged snow cover in the spring (Kuchko 1988, Raatikainen and Niemela 1994, Selas 2000). In the summer and fall, high temperatures will also reduce the number of flower buds produced because of dehydration stress or heat stress (Selas 2000). These weather events will reduce berry numbers in the next summer regardless of how favourable the rest of the season was, and these events may confound simple N and berry yield correlations. Precipitation extremes can also affect berry production. In the summer, both low and high amounts of rainfall during berry ripening reduce berry production (Selas 2000). During periods of low precipitation, plant growth is slowed and energy is stored rather than placed into flowering primordia (Selas 2000). Low rainfall usually correlates with higher than normal temperatures so both factors then work against the plant's reproductive output and confound climate-berry correlations. For example, long-term warmer temperatures can produce a primary response of increased vegetative growth, but not increased reproductive output (Arft et al. 1999). Only later will a secondary response be seen in the form of an increase in reproductive output (Arft etal. 1999). During periods of excessive precipitation average temperatures may drop, and inhibit plant reproductive output. Cloudy rainy days are usually cooler, and temperatures of less than 10.5° C for more than 4 days significantly reduce shoot and rhizome carbohydrates in arctic and alpine plants (Bliss 1971). Lowered reserves leave fewer carbohydrates for flower set in the fall resulting in fewer berries produced the following year. Cooler temperatures in the early spring can lead to delayed plant phenologies of approximately 7 to 11 days (Dyke 1971, Belonogova 1988). This results in reduced vigour of vegetation and small fruit crops in soapberry (Shepherdia canadensis (L.) Nutt.), bearberry (A. uva-ursi), crowberry (E. nigrum), toadflax (Geocaulon lividum (Richards) Fern.), cranberry (Vaccinium vitis-idaea), and blueberry (Vaccinium spp.) among others (Dyke 1971, Phoenix et al. 2001). Therefore, even with optimal precipitation flower and berry production may be reduced if there are cooler temperatures and excessive cloud cover (Bliss 1971, Phoenix et al. 2001). Climatic factors may also indirectly affect berry production by affecting insect pollinators during plant flowering times. Kuchko (1988) reports that low temperatures decreased the activity level of insect pollinators and reduced berry yield in Vaccinium vitis-idaea and V. myrtillus.  6  J  Pelletier et al. (2001) excluded pollinators from plots, and fruit set in cloudberry (Rubus chamaemorus) declined from 63%-98% to 13%-18%. Pollinator insects, such as blackflies (Simulidae), may be reduced through summer heat and drought, or through natural population swings, leaving fewer offspring to pollinate flowers in the following spring. In these cases berry production is not only directly affected by climatic variables, but also indirectly affected from a reduction in pollinators. This would further obscure climate-berry production correlations as an optimal berry production season may coincide with a low point in the pollinator population cycle. In addition to climatic and soil nutrient factors which may mask climate-berry correlations there are intrinsic events within the plants that may also be governed by climatic factors. For example, many species of plants have mast years where every few years they produce a large quantity of seeds. During intermast years decreased seed quantity cannot sustain high levels of seed predators, so in mast years there are enough seeds to result in 'seed predator satiation' (Janzen 1971, Schnurr et al. 2002). Each seed then has a higher chance of survival, and large seed crops are more efficient than small ones as a means of propagating the species (Koenig and Knops 2005). This reproductive adaptation to mast seems to be a selected evolutionary trait that is modified by climate (Janzen 1971, Kelly and Sork 2002) Mast years and the factors that produce them are well studied in deciduous tree species (Jensen 1982, Pucek et al. 1993, Sork et al. 1993, reviewed in Kelly and Sork 2002), but relatively little work has been done looking at mast years in dwarf shrubs (Dyke 1971, West 1982). It is only recently that regular berry cycles in ericaceous shrubs (Vaccinium spp.) have been documented (Vander Kloet and Cabilia 1996). Ericaceous shrubs seem to have their own internal cycle that is dependent upon the previous year's production. For example, Selas (2000) using 50 years of berry production data in V. myrtillus concluded that the best predictor of berry production was the berry production of the previous 3 years providing climatic factors were also taken into account. Berry production is thus heavily dependent upon climatic factors that modify an existing internal cycle, and N addition to soils may only serve to amplify the effects of this cycle. In other words berry production may decrease some years despite N addition, but does not decrease as much as it would have if N had not been added. Finding the correlations between N , climate and  7  berry production requires knowing which factor, if any, was the driving force as well as a little luck. 1.4 Do berry crops affect the northern red-backed vole (C rutilus) community? Berries are a high quality food upon which C. rutilus seems to rely more than other small mammals, and are an important part of their diet (Dyke 1971, West 1979, 1982). Dyke (1971) found that the stomach contents of red-backed voles consisted almost exclusively of berries in berry mast years. West (1982) reported that berries and seeds comprised 62%-92% of the summer and fall diet in northern red-backed voles, and hypothesized that C. rutilus winter survival was dependent on a large berry crop. Boonstra et al. (2001) also felt berries were the key food for good overwintering survival, which in turn would lead to higher red-backed vole densities in the following summer. For example, in bank voles (Clethrionomys glareolus) the population growth indices were strongly correlated with the previous and current years bilberry (V. myrtillus) production index with high vole density corresponding to increased berry production (Selas et al. 2002), and other studies have reported an increase in C. glareolus due to extra high-quality food (Jensen 1982, Hansson 1984, Pucek et al. 1993, Prevot-Julliard et al. 1999). Increased vole densities due to better quality food have also been seen in a number of other species such as prairie voles (Microtus orchogaster) (Cole and Batzli 1978), M. townsendii (Taitt and Krebs 1981), as well as both C. gapperi and C. rutilus (Fuller 1977, Gilbert and Krebs 1981, Krebs and Wingate 1985, Gilbert and Krebs 1991, Schweiger and Boutin 1995). Abundant high quality food for C. rutilus may occur when berry producing dwarf shrubs experience a mast year. Since seed-predator numbers are low not all the berries will be eaten before the snowfall covers the remaining berries. Under the snow, the berries are preserved and provide a food supply for subnivean mammals throughout the winter and into early spring (West 1982, Boonstra et al. 2001). A n early snowmelt though would expose the remaining berries to other frugivores and decomposition, and vole numbers may decline as there could be too long a time lag between disappearance of the berries, and the availability of other foods (West 1982). With a delayed snow-melt, however, berries would remain as a viable nutritious food source till other plants became available, and vole survival rates should increase resulting in a higher vole densities (West 1982, Korpimaki et al. 2004).  8  1.5 Summary 1. Nitrogen is the main soil nutrient that influences production of berry crops. Berry production is also influenced by both the amount and timing of N addition to the soil. An interesting experiment would be to add a constant amount of N to shrubs, but to do it at different intervals, such as one large application and more smaller application, similar to the pulsing experiments done with water in some desert studies (Novoplansky and Goldberg 2001), or the Hanson and Retamales (1992) experiment. 2. Berry production responses to N addition in the boreal region are often delayed by several years, with secondary reduction in output from grass and herb competition occurring even later. Alternatively some studies indicated that a primary response in berry production may be seen after two years of favourable weather followed by a reduction in output. This reduction in berry output may be one of the factors that contribute to the decline in C. rutilus. 3. Berry production is influenced more by unfavourable climatic conditions that either directly or indirectly limit or kill flower buds rather than favourable conditions that would allow flower buds to bloom. With time we will have enough data to compare berry production indices with weather data from the Kluane region to determine to what extent this holds true. 4. Berry numbers in some plants are also influenced by at least the past three years of production due to internal rhythms of plant growth. Further analyses of long-term berry data will clarify which, if any, plants in the Kluane region this would apply to. 5. Natural fluctuations in pollinators caused by both external (climate) and internal (population fluctuations) factors may confound climate-berry production correlations. 6. There is ample evidence that larger amounts of N addition increases graminoid and herbaceous plant growth, and in some cases may negatively affect berry production. 7. Red- backed vole winter survival increases with availability of high quality foods such as seeds from berry plants. 8. Population indices in other vole species are strongly correlated with berry production indices. Based on the information in this literature review I will test the fecal nutrient enrichment hypothesis which hypothesizes that small amounts of nitrogen (N) found in snowshoe hare  9  pellets can increase berry production in ground shrubs resulting in a subsequent increase in C. rutilus numbers. To test this hypothesis I will do the following: 1. Simulate the N from hare pellets by adding low levels of N (0.5-4.0 g/m ) to berry 2  producing shrubs to determine if these levels increase berry production. 2. Analyze both long-term berry production and C. rutilus densities to look for correlations between the two. 3. Add low levels of N (1.0 and 2.0 g/m )) to small mammal trapping grids to determine if 2  C. rutilus density increases on these grids over the next 3 to 4 years. I predict that low levels of N will cause berry production to increase in the short-term, and that increased berry production will lead to an increase in C. rutilus densities the following year. Furthermore, analysis of long-term berry production data and C. rutilus density data will show a correlation between the two with a one year time lag. Chapter 2 reports on the results from adding low levels of N to berry producing shrubs, and Chapter 3 analyzes long-term berry production and C. rutilus data. The mammal trapping grids will continue to be monitored and results presented in the future.  10  1.6 Literature Cited Aber, J.D., J.K. Nadelhoffer, P. Steudler, J.M. Melillo. 1989. Nitrogen saturation in northern forest ecosystems. BioSci. 36(6):378-386. • Arft, A. M . , M . D. Walker, J. Gurevitch, J. M . Alatalo, M . S. Bret-Harte, M . R. T. Dale, M . Diemer, F. Gugerli, G. H . R. Henry, M . H . Jones, R. D . Hollister, I. S. Jonsdottir, K . Laine, E. Levesque, G. M . Marion, U . Molau, P. Molgaard, U . Nordenhall, V. Raszhivin, C. H. Robinson, G. Starr, A. Stenstrom, M . Sentstrom, O. Totland, P. L. Turner, L. J. Walker, P. J. Webber, J. M . Welker, and P. A. Wookey. 1999. Responses of tundra plants to experimental warming: meta-analysis of the international tundra experiment. Ecol. Mono. 69:491-511. Baker, E.S., H. Olff, M . Boekhoff, J.M. Gleichman, and F. Berendse. 2004. Impact of herbivores on nitrogen cycling: contrasting effects of small and large species. Oecol. 138:91-101. Belonogova, T.V. 1988. Yield forecasting and optimization of berry harvesting in the forests of Southern Karelia. USSR. In Proceedings of the Finnish-Soviet symposium on timber forest resources in Jyvaskyla, Finland. 25-29 August 1986. Edited by I. Vanninen and M . Raatikainen. Acta. Bot. Fenni. 136:19-21. Helsinki. Bliss, L. C. 1971. Arctic and alpine plant life cycles. Annu. Rev. Ecol. Syst. 2:405-438. Bonan, G. B., and H. H. Shugart. 1989. Environmental factors and ecological processes in boreal forests. Annu. Rev. Ecol. Syst. 20:1-28. Boonstra, R., C. J. Krebs, S. Gilbert, and S. Schweiger. 2001. Chapter 10: Voles and Mice. Pages 215-239 in C. J. Krebs, S. Boutin, and R. Boonstra, editors. Ecosystem Dynamics of the Boreal Forest: The Kluane Project. Oxford University Press, New York. Chapin, III, F.S., and G. R. Shaver. 1989. Differences in growth and nutrient use among arctic plant growth forms. Func. Ecol. 3:73-80. Chapin, III, F.S., and G. R. Shaver. 1985. Individualistic growth response of tundra plant species to environmental manipulations in the field. Ecol. 66:564-576. Chapin, III, F.S., P. M . Vitousek, and K. Van Cleve. 1986. The nature of nutrient limitation in plant communities. Amer. Nat. 127:48-58. Clark, J.E., E.C. Hellgren, J.L. Parsons, E.E. Jorgensen, D . M . Engle, and D . M . Leslie. 2005. Nitrogen output from fecal and urine deposition of small mammals: implications for nitrogen cycling. Oecol. 144:447-455. Cole, F.R., and G.O. Batzli. 1978. Influence of supplemental feeding on a vole population. J. Mamm. 59:809-819.  11  Dyke, G. R. 1971. Food and cover of fluctuating populations of northern cricetid. Ph.D. Thesis, Univ. Edmonton, Alberta. Eaton, L. J. 1994. Long-term effects of herbicide and fertilizers on lowbush blueberry growth and production. Can. J. Plant Sci. 74:341-345. Fuller, W. A. 1977. Demography of subarctic population of Clethrionomys gapperi numbers and survival. Can. J. Zool. 55:42-51. Gerdol, R., L. Brancaleoni, M . Menghini, and R. Marchesini. 2000. Response of dwarf shrubs to neighbour removal and nutrient addition and their influence on community structure in a subalpine heath. J. Ecol. 88:256-266. Gilbert, B.S. and C.J. Krebs. 1991. Population dynamics of Clethrionomys and Peromyscus in southwestern Yukon 1973-1989. Hoi. Ecol. 14: 250-259. Gilbert, B. S., and C. J. Krebs. 1981. Effects of extra food of Peromyscus and Clethrionomys populations in,the southern Yukon. Oecol. 51:326-331. Graham, S. A., and R. Turkington. 2000. Population dynamics response of Lupinus arcticus to fertilization, clipping, and neighbour removal in the understory of the boreal forest. Can. J. Bot. 78:753-758. Grime, J.P. 1979. Plant strategies and vegetative processes. Wiley Press. Chinchester, New York, NY. Hanson, E. J., and J. B . Retamales. 1992. Effect of nitrogen source and timing on highbush blueberry performance. Hort. Sci. 27:1265-1267. Hansson, L. 1984. Winter reproduction of small mammals in relation to food conditions and population dynamics. Bull. Cam. Mus. Nat. Hist. 10:225-234. Hikosaka, K., Y. T. Hanba, T. Hirose, and I. Terashima. 1998. Photosynthetic nitrogen-use efficiency in leaves of woody and herbaceous species. Func. Ecol. 12:896-605. Janzen, D.H. 1971. Seed predation by animals. Annu. Rev. Ecol. Syst. 2:465-492. Jensen, T. S. 1982. Seed production and outbreaks of non-cyclic rodent populations in deciduous forests. Oecol. 54:184-192. Kalela, O. 1962. On the fluctuations in the numbers of arctic and boreal small rodents as a problem of production biology. Annal. Acad. Sci. Fenni. Series A IV 66:1-38. Kelly, D. and V . L . Sork. 2002. Mast seeding in perennial plants: why, how, where? Annu. Rev. Ecol. Syst. 33:427-447.  12  Kerley, S.J. and D.J. Read. 1997. The biology of mycorrhiza in the Ericacea. X I X . Fungal mycelium as a nitrogen source for the ericoid mycorrhizal fungus Hymenoscyphus ericae and its host plants. New Phytol. 136:691-701. Koenig, W.D. and J.M.H Knops,. 2005. The mystery of masting in trees. Amer. Sci. 93: 340347. Korpimaki, E., P. R. Brown, J. Jacob, and R. P. Pech. 2004. The puzzles of population cycles and outbreaks of small mammals solved? BioSci. 54:1071-1079. Krebs, C. J., and I. Wingate. 1985. Population fluctuations in the small mammals of the Kluane Region, Yukon Territory. Can. Nat. 99:51-61. Kuchko, A. A . 1988. Bilberry and cowberry yields and the factors controlling them in the forests of Karelia, USSR. In Proceedings of the Finnish-Soviet symposium on timber forest resources in Jyvaskyla, Finland. 25-29 August 1986. Edited by I. Vanninen, and M . Raatikainen. Acta. Bot. Fennica 136:23-25. Lehmushovi, A. 1975. Puolukka, uusi viljelykasvi. Pellervo 76:32-34. McKendrick, J. D., G. O. Batzli, K. R. Everett, and J. C. Swanson. 1980. Some effects of mammalian herbivores and fertilization on tundra soils and vegetation. Arc. Alp. Res. 12:565-578. Mols, T., J. Paal, and E. Fremstad. 2000. Response of Norwegian alpine communities to nitrogen. Nord. J. Bot. 20:705-712. Mukula, J., and M . Raatikainen. 1983. Voidaanko metsamarjojen satoa lisata. Kaytannon Maamies 32:57-59. Nams, V.O., N.F.G. Folkard, and J.N.M. Smith. 1993. Effects of nitrogen fertilization on several woody and nonwoody boreal forest species. Can. J. Bot. 71:93-97. Niittymaa, L. 1983. Puolukan iannoituskokeista. Metsantutkimuslaitoksen Tiedonantoja 90:161163. Nordin, A., T. Nasholm, and L. Ericson. 1998. Effects of simulated N deposition on understorey vegetation of a boreal coniferous forest. Func. Ecol. 12:691-699. Nordin, A. T., P. Hogberg, and T. Nasholm. 2001. Soil nitrogen form and plant nitrogen uptake along a boreal forest productivity gradient. Oecol. 129:125-132. Nordin, A. J. Strengbom, J. Witzell, T. Nasholm, and L. Ericson. 2005. Nitrogen deposition and the biodiversity of boreal forests: implications for the nitrogen critical load. R. Swed. Acad. Sci. 34(l):20-24.  13  Novoplansky, A., and Goldberg, D.E. 2001. Effects of water pulsing on individual performance and competitive hierarchies in plants. J. Veg. Sci. 12(2): 199-208. Ohlson, M . , A . Nordin, and T. Nasholm. 1995. Accumulation of amino acids in forest plants in relation to ecological amplitude and nitrogen supply. Func. Ecol. 9:956-605. Paivinen, L. 1976. Lannoittamalla lisaa puolukoita. Nuorten Sarka 31:10-30. Pastor, J . , B. Dewy, and D.P. Christian. 1996. Carbon and nutrient mineralization and fungal spore composition of fecal pellets from voles in Minnesota. Ecogr. 19:52- 61. Pastor, J. B. Dewey, R.J. Naiman, P.F. Mclnnes, and Y. Cohen. 1993. Moose browsing and soil fertility in the boreal forests of Isle Royale National Park. Ecol. 74: 467-480. Pelletier, L., A . Brown, B . Otrysko, and J.N. McNeil. 2001. Entomophily of the cloudberry (Rubus chamaemorus). Ent. Exper. Appl. 101: 219-224. Penney, B.G., K . B . McRae, and G.A. Bishop. 2003. Second-crop N fertilization improves lowbush blueberry (Vaccinium angustifolium Ait.) production. Can. J. Plant Sci. 83:149155. Persson, J., P. Hogberg, A . Ekblad, M . N . Hogberg, and A. Nordgren. 2003. Nitrogen acquisition from inorganic and organic sources by boreal forest plants in the field. Oecol. 137:252257. Phoenix, G. K., D. Gwynn-Jones, T.V. Callaghan, D. Sleep, and J. A . Lee. 2001. Effects of global change on a sub-arctic heath: Effects of enhanced U V - B radiation and increased summer precipitation. J. Ecol. 89:256-267. Press, M.C., J.A. Potter, M . J.W. Burke, T. V . Callaghan, and J.A. Lee. 1998. Responses of a subarctic dwarf shrub heath community to simulated environmental change. J. Ecol. 86:315-327. Prevot-Julliard, A.C., H . Henttonen, N.G. Yoccoz, and N.C. Stenseth. 1999. Delayed maturation in female bank voles: optimal decision or social constraint? J. Anim. Ecol. 68:684-697. Pucek, Z., W. Jedrzejewski, B. Jedrzejewska, and M . Pucek. 1993. Rodent population dynamics in a primeval deciduous forest (Bialowieza National Park) in relation to weather, seed crop, and predation. Acta. Therio. 38:199-232. Raatikainen, M . , and M . Niemela. 1994. The effect of fertilization on the yield of wild forest berries pp. 123-129. In Effects of Fertilization on Forest Ecosystems. University of Jyvaskyla. Jyvaskyla, Finland. Biological Research Reports. Retamales, J.B., and E.L. Hanson. 1989. Fate of N-labeled urea appied to mature highbush blueberries. J. Amer. Soc. Hort. Sci. 114:921-923. 15  14  Schnurr, J., R.S. Ostfeld, and C D . Canham. 2002. Direct and indirect effects of masting on rodent populations and tree seed survival. Oikos 96:402-410. Schweiger, S., and S. Boutin. 1995. The effects of winter food addition on the population dynamics of Clethrionomys rutilus. Can. J. Zool. 73:419-426. Selas, V. 2000. Seed production of a masting dwarf shrub, Vaccinium myrtillus, in relation to previous reproduction and weather. Can. J. Bot. 78:423-429. Selas, V., E. Framstad, and T.K. Spids0. 2002. Effects of seed masting of bilberry, oak, spruce on sympatric populations of bank vole (Clethrionomys glareolus) and wood mouse (Apodemus sylvaticus) in southern Norway. J. Zool. Lond. 258:459-468. Shaver, G. R., I. Chapin, F.S., and B. L . Gartner. 1986. Factors limiting seasonal growth and peak biomass accumulation in Eriophorum vaginatum in Alaskan tussock tundra. J. Ecol. 74:257-278. Sork, V . L . , J. Bramble, and O. Sexton.1993. Ecology of mast-fruiting in three species of North American deciduous oaks. Ecol. 74:528-541. Spayd, S.E., R.L. Wample, R.G. Stevens, R.G. Evans, and A . K . Kawakami. 1993. Nitrogen fertilization of White Riesling in Washington: Effects on petiole nutrient concentration, yield, yield components, and vegetative growth. Am. J. Enol. Vitic. 44:378-386. Stoddard, J.L. 1994. Long-term changes in water-shed retention of nitrogen—its causes and aquatic consequences. In Environmental Chemistry of Lakes and Reservoirs. Baker, L.L. ed. Advances in Chemistry Series 237. American Chemical Society. New York, pp. 223284. Strengbom, J., G. Englund, and L. Ericson. 2006. Experimental scale and precipitation modify effects of nitrogen addition on a plant pathogen. J. Ecol. 94:227-233. Strengbom, J. M . Walheim, T. Nasholm, and L. Ericson. 2003. Regional difference in the occurrence of understorey species reflects nitrogen deposition in Swedish forests. R. Swe. Aca. Sci. 32:91-97. Strengbom, J, A. Nordin, T. Nasholm, and L. Ericson. 2001. Slow recovery of boreal forest ecosystem following decreased nitrogen input. Func. Eco. 15: 451-457. Taitt, M . J., and C. J. Krebs. 1981. The effect of extra food on small rodent populations. 2. Voles (Microtus townsendii). J. Anim. Ecol. 50:125-137. Tamm; C O . 1982. Nitrogen cycling undisturbed and manipulated boreal forest. Phil. Trans. R. Soc. Lond. 296:419-425.  15  Turkington, R., E. John, C. J. Krebs, M . R. T. Dale, V. Nams, R. Boonstra, S. Boutin, K. Martin, A. R. E. Sinclair, and J. N . M . Smith. 1998. The effects of fertilization on the vegetation of the boreal forest. J. Veg. Sci. 9:333-346. Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of fertilization and herbivory on the herbaceous vegetation of the boreal forest in north-western Canada: a 10-year study. J. Ecol. 90:325-337. Vander Kloet, S. P., and P. Cabilia. 1996. Ten year study of the annual variation in berry and seed production in a population of Vaccinium corymbosom L . Amer. Mid. Nat. 135.349356. Vitousek, P. M . , and P. A . Matson. 1984. Mechanisms of nitrogen retention in forest ecosystems: a field experiment. Science 225:51-52. West, S. D. 1982. Dynamics of colonization and abundance in central Alaskan populations of the northern red-backed vole, Clethrionomys rutilus. J. Mamm. 63:128-143. West, S. D. 1979. Habitat responses of microtine rodents to central Alaskan forest succession. Ph.D. Thesis. University of California, Berkeley, California. Yudina, V. F., and T. A. Maksimova. 2005. Dynamics of yielding capacity of small cranberry in southern Karelia. Rus. J. Ecol. 36:239-242.  16  Chapter 2: Effects of Nitrogen Addition on Berry Production 2.1 Introduction Ten year population fluctuations of snowshoe hares (Lepus americanus) are a prominent characteristic of the boreal forest in south-western Yukon. Population densities of other mammals in this community also fluctuate. One of these, the northern red-backed vole (Clethrionomys rutilus Pallas), has been monitored since the mid-1970s, and has undergone sporadic population increases with 3 large peaks in 1973, 1984 and 1992. These peaks coincided with declines of the snowshoe hare densities (Boutin et al. 1995, Boonstra et al. 2001). One explanation for this vole-hare relationship has been formulated as the fecal nutrient enrichment hypothesis (Boonstra et al. 2001). When hares are at their population peak they produce a large quantity of fecal pellets. Nitrogen (N) released from the decay of these pellets can reach 0.5 to 1.0 g/m with highs of 2.0 g/m being reached in areas where pellets accumulate 2  2  (Hik pers. comm.). The N released from these pellets could be absorbed by berry producing shrubs whose fruit and seeds are critical in the diet of C. rutilus. Given favourable weather conditions these shrubs could use the extra N and produce a large crop of berries, many of which will last through the winter and into the next spring. The availability of berries may increase the overwintering survival of C. rutilus which would then increase in density two to three years after L. americanus peaks (Fuller 1977, Gilbert and Krebs 1981,West 1982, Boonstra et al. 2001). Boreal forest soils tend to be nutrient poor (Bonan and Shugart 1989, Lovblad et al. 1992 in Nordin et al. 1998), so dwarf shrubs growing in these soils have adapted to low nutrient conditions by having slow steady growth rates (Chapin 1980, Chapin et al. 1990). Their growth rates are based on nutrient availability over several years unlike herbaceous plants whose growth is coupled to the current N supply (Chapin 1980, Chapin et al. 1990). When N is added to the soil faster growing herbaceous and graminoid species can outcompete the dwarf shrubs, and even eventually reduce the overall diversity within the fertilized area (Nordin et al. 1998, Press et al. 1998, Turkington et al. 2002). In this case addition of N would have a negative effect on shrub berry production as the shrubs are outcompeted by faster growing plants (several studies cited in Raatikainen and Niemela 1994). A pulse of small amounts of N (<2.0 g/m ), however, such as would occur after peak hare 2  densities, may be insufficient to influence the growth of herbs and graminoids, and changes in diversity due to N addition may take several years to occur (Mols et al. 2000, Turkington et al.  17  2002). Shrubs may then have time to use the additional N for berry production before being outcompeted. With the 9 to 10 years between N pulses, competition from faster growing plants may never be sufficiently activated to change the understory plant community. In this study I propose to test the fecal nutrient enrichment hypothesis by adding N to plots at rates similar to those found in hare pellets (0.5-2.0 g/m ). The hypothesis predicts that in 2  the short term there will be an increase in berry production in selected plants.  2.2 Methods 2.2.1 Study area and experimental sites The study area is in the boreal forest region of southern Yukon near Kluane Lake on the Alaska Highway approximately 45 km north of Haines Junction. It lies within the Shakwak Trench system (61° 01'N, 138° 24' W), and in the rain shadow of the St. Elias Mountains (Figure 2.1). Mean annual precipitation is ca. 280 mm (1945-2003) which includes an average annual snowfall of approximately 100 cm. The tree community is dominated by white spruce (Picea glauca (Moench) Voss) interspersed with trembling aspen (Populus tremuloides Michx.) and balsam poplar (Populus balsamifera L.). The upper shrub layer is composed of willow (Salix spp.), soapberry (Shepherdia canadensis (L.) Nutt.) and dwarf birch (Betula glandulosa Michx.) and the ground layers are composed of dwarf shrubs and herbaceous plants such as bearberries (Arctostaphylos rubra (Rehd. & Wils.) Fern, and Arctostaphylos uva-ursi (L.) Spreng. s.L), crowberry (Empetrum nigrum L.), blueberry/cranberry (Vaccinium spp.), toadflax (Geocaulon lividum (Richards) Fern.), arctic lupine (Lupinus arcticus S. Wats), and other forbs (Turkington et al. 2002). This study was done in an area of moderately open to closed spruce forest (55-80% canopy cover) with a well developed (>80% cover) herbaceous understorey, approximately 6 km to the north-west of Boutellier Summit on the Alaska highway. The site was probably last burned in 1872 (Francis 1996, Dale et al. 2001). Within this area shrub plots and treatment grids for trapping small mammals were established. 2.2.2 Experimental sites A total of 300 shrub plots were established using 4 species of shrubs and one berry-producing  s  herb. Three of the shrub species are prostrate: Empetrum nigrum, A. uva-ursi (bearberry) and A.  18  rubra (red bearberry); one shrub is upright: S. canadensis; and one is a herb: G. lividum. For each of the three prostrate species, and for G. lividum, sixty 1 m circular plots were established 2  in a non-random manner. These species have a patchy distribution in the forest, and thus I wanted to target my treatments to be as homogenous as possible within a species. They were selected so that plots consisted largely of the species of interest. Each plot was a minimum of 3 m distant from any other plot and all plots were trenched to a depth of 20-25 cm to sever roots that extended beyond the plot to prevent any exchange of N along root systems with plants outside the treatment plots. Within each species the plots were assigned to 6 treatment groups of 10 plots each. The fifth shrub species, S. canadensis, was treated in the same as above except the circular plot area was 3 m as 1 m was too small to contain some of the shrubs. The diameters of 2  2  two of the main stems per plot were measured at ground level and then tagged for berry counts later in the season. Again, 60 plots were established. A l l plots were randomly assigned to one of six N addition treatments (g/m ): 0, 0.5, 1.0, 2  2.0, 4.0 and 17.5. A l l plots were fertilized in early June in 2004, and in mid-May in 2005 with ammonium nitrate (NH4NO3) from The Feed Store, Whitehorse, Y T . The 0.5 g to 4.0 g N / m  2  were added to simulate the range of N released from hare pellets during years of high hare density (Ffik pers. comm.), and the 17.5 g added as a negative control to determine if N addition would have any effect on small plots. A l l plots were photographed with a Nikon CoolPix 5.0 digital camera. Photographs were arranged with true north at the top of the picture, the plot boundary outlined by plastic tubing (an expanded hula-hoop 1.15 m in diameter) with string dividing the plot into quarters, labelled flagging tape on a nail indicating the center of the plot, and a marker indicating the north side of the plot (Figure 2.2). The plot location was drawn on a map relative to the other plots to aid in (  finding the plots. A l l plot locations were also marked with labelled flagging tape attached to a nearby tree. To assess the response of the plants to the fertilizer treatments, I made 6 types of measurements on each plot. For each measurement except percent cover 10 plants were selected every 10 cm along a north-south transect. This avoided possible observer bias from unconsciously selecting larger individuals. If fewer than 10 plants occurred along a north-south transect, then an east-west transect was also used until 10 plants had been sampled. Measurements were taken at the end-of the growing season in late July, early August.  19  1. Percent cover: In mid to late summer of 2004 and 2005 percent cover of all species was measured within the 1 m plots using A circle (pie-shaped) piece of transparent vinyl divided 2  X  into sections R l to R3 where R l was 11% , R2 was 33%, and R3 56% of the quarter-circle (Figure 2.3). The vinyl was first placed in the northwest corner of the plot and plant distribution within each of the sections was measured to the nearest 5%. Proceeding in a clockwise direction the same measurements were done for all 4 quarters of the plot. The measurements were then collated on Excel to provide an overall percentage of plant cover and species within the plot. 2. New growth: New growth on S. canadensis, E. nigrum, A. uva-ursi, and G. lividum was measured at the end of growing season in late July, early August. New growth on A. rubra was not measured due to the difficulty of differentiating between new stem growth and new leaf growth. For G. lividum the new growth consisted of the entire plant from the soil to the tip of the upwards folded top leaves. 3. Number of leaves on new growth: Number of new leaves were counted on ten samples/plot each of S. canadensis, A. uva-ursi, G. lividum and A. rubra. Empetrum nigrum leaves were not sampled as they were small, numerous and readily shed from the stem when handled. 4. Length, breadth and dry weight of leaves: Leaf samples were collected from separate plants (i.e. no individual plant had more than one leaf sampled). Leaf samples of S. canadensis, A. uvaursi, G. lividum and A. rubra were placed between the pages of a loose-leaf notebook, and the top page was rubbed with graphite to produce leaf tracings. The tracings were sprayed with hair spray to prevent the graphite from smearing. The leaves were placed in a labelled coin envelope and dried in an oven at 60° C for 5 days. Length and breadth of the leaves were measured using the graphite leaf tracings. Empetrum nigrum leaves were not measured as they were too small. 5. Number of berries: Berries on the prostrate species were counted 2.5 cm on either side of a 1m north-south and a 0.95-m east-west transect for a total area measured 975 cm . The 0.95-m 2  transect avoided the overlap where the two transects intersected in the middle of the plot. Berry numbers were extrapolated to berries/m by measuring the shrubs' percent cover within the two 2  transects. A l l berry counts were conducted while the majority of berries were unripe before animal harvesting began typically in mid to late July. Shepherdia canadensis berries were counted as per the Kluane Monitoring Protocols (2004 pp.53-57) as follows: Two branches of  20  the shrub were randomly selected, the basal diameters measured, marked with permanent aluminium tags, and berries counted on each of the two branches while they were still green. 6. Wet and dry weight of berries: Where possible 10 to 30 berries per plot were collected with no two berries coming from the same plant, sealed in labelled empty film canisters, and weighed immediately upon return from the field. The berries were then placed in coin envelopes and dried for 5 days in the oven at 60° C, and reweighed. 2.2.3 Analyses A N O V A analyses (and where appropriate with a Tukey test) was used to compare the means of number of berries, leaf width and breadth, and new growth in each of the treatments. All analyses were done using both JMP 4 and SPSS 11.  2.3 Results Summer weather conditions in 2004 were hotter and drier than average and new temperature records were set. In 2005 though, summer temperatures were slightly cooler, and much wetter than average. Precipitation in 2004 was 229 mm with 146 mm falling May to August. The average summer temperature in 2004 was 12.2° C. In 2005 precipitation was 318 mm with 237 mm of that falling May to August. The average summer temperature was 11.4 with an average June temperature of 11.9° C, two full degrees lower than the 2004 average June temperature (13.9° C) (Burwash Meteorological Station). On June 5, 2005 approximately 8 cm of snow accumulated and buried the flowering shrubs. This killed many of the open flowers. Arctostaphylos rubra There was no change in percent cover from 2004 to 2005 except on plots where the negative control (17.5 g N/m ) had caused some mortality. 2  More leaves grew on the plants at 2 g N / m but the difference was not significant. Leaf 2  length and breadth did not change, but dry leaf weight was significantly heavier at the 17.5 g N / m in 2005 than in.2004 (Table 2.1). 2  Variation in berry numbers in response to N was non-significant in 2004 with fewer berries produced as fertilizer levels increased (Figure 2.4). Very few berries formed at 17.5 g N/m due to shrub mortality. Overall berry production in A. rubra increased slightly from the controls to 1.0 g N / m but then decreased until berry levels were near 0 at 17.5 g N / m in 2004. 2  2  There was no significant change in berry weight in either year.  21  Arctostaphylos uva-ursi No significant change in percent cover occurred even at high levels of N addition as little visible damage occurred to the shrub contrary to the effects on A. rubra and E. nigrum. No significant differences in new growth, length, breadth and dry weight of leaves occurred in 2004 or 2005. Berry weights were also not significant. In 2004 number of berries at 1 g N / m increased significantly, but a corresponding 2  significant difference did not occur in 2005 due to the low number of berries growing on the shrubs (Figure 2.5). Empetrum nigrum Plant mortality due to high levels of N caused a significant change in percent cover of E. nigrum. In both 2004 and 2005 (Figure 2.6) new growth increased significantly from 0 g N / m to 1 g 2  N / m before decreasing at 2 g N/m , and then increasing again. The average length of new 2  2  growth on the controls in both years were the same (107 mm vs 104 mm in 2004 and 2005 respectively). This significant change in new growth was not translated to significant differences in dry weight of the new growth at the P=0.05 level. In 2004 the number of E. nigrum berries/m showed no significant response to fertilizer 2  treatment, but a significant response occurred in 2005. The largest difference was between the 0.5 g N / m and 17 g N / m where a large decrease occurred (Figure 2.7). Likewise, in 2005 at 2  2  1.0 g N / m E. nigrum wet berry weight was significantly heavier than at 17.5 g N / m as well for 2  2  the control. The overall trend in E. nigrum was for an increase in berry weight to 1.0 g N / m  2  followed by a decrease in weight as N levels increased (Figure 2.8). Geocaulon lividum Percent cover, new growth, .width and breadth, and dry weight of leaves did not significantly change in either year. However, in the second year a visible albeit non-significant increase in leaf breadth occurred as N levels increased. In 2004 all growth measurements were similar in all treatments. Also, despite the cooler weather in 2005 all plants were larger than in 2004. No significant difference in the number of berries or berry weight occurred. In 2005 the heaviest berries were found at the 1.0 g N / m treatment although this weight was just marginally 2  non-significant at P=0.08. A strong positive correlation (r = 0.74, d.f.= 54) occurred between leaf breadth and berry weight (Figure 2.9) in 2005.  22  Shepherdia canadensis Plants did not vary significantly in any measurements although a non-significant (P=0.13) spike in berry weight occurred at 1 g N/m . A similar spike also occurred at the same treatment level in 2  E. nigrum and G. lividum (Figure 2.8).  2.4 Discussion The fecal nutrient enrichment hypothesis predicts that N levels in snowshoe hare fecal pellets are sufficient to produce an increased number or heavier berries on berry producing plants. Some evidence from this experiment supports the fecal nutrient enrichment hypothesis, but many of the plants exhibited no significant changes in berry production in response to increasing N levels. This lack of significant response may be from the low number of berries present on the plots in the second year as many of the plants showed a consistent but non-significant response 2  2  in berry production/m with peak berry production at low levels of N (0.5-2.0 g/m ). The low number of berries in the plots was influenced by a combination of a late snowfall in early June that would have killed many of the open flowers, and the wetter than average summer weather. As a result of the cooler wetter weather fewer resources would have been devoted to berry growth and production (Bliss 1971, Belegonova 1988). If weather had been conducive to berry production we might have seen a stronger response in berry production. 2.4.1 Berry production Most of the significant responses regarding berries are found around the 0.5 to 1.0 g N / m in the 2  second year following application of N . This is to be expected because shrubs produce buds for the current season's flowers in the previous year (Yudina and Maksimova 2005). Nitrogen levels at these low values are sufficient to produce more and larger berries, and levels of N above this are likely to reduce berry numbers through increased herbivory, fungal attacks, competition with forbs (Nordin et al. 1998, Press et al. 1998, Turkington et al. 2002), and from fertilizer burn. At low levels of nitrogen E. nigrum had the heaviest berries (Figure 2.8) and also the most berries produced per m (Figure 2.7). None of the other plants showed any significant 2  difference in berry weight or berry numbers. Although difference were not significant, G. lividum, A. rubra and 5. canadensis did exhibit a similar trend to E. nigrum in that the heaviest berries were found at 1.0 g N/m , and that for the most part the weight declined with increasing 2  23  N levels. In addition the number of berries produced by the other plants also followed a similar trend in that production of berries/m increased at low levels of N , and then decreased. Although 2  these trends were not significant, they were consistently seen in the plants surveyed indicating that the effects of low levels of N on the plants were not completely random. If more berries had been produced on the plots then there may have been more significant results. 2.4.2 Plant growth Only one significant result occurred in growth of the shrubs and herbaceous plants, and this was the amount of new growth in E. nigrum (Figure 2.6). In 2004 and 2005 the significant difference in new growth occurred at 1.0 g N/m . With more N new growth began to decrease, but then 2  increased again at the 17.5 g N / m in 2005. In this case though most of the plot had experienced 2  high mortality from fertilizer burning in the year previous. The plants that were left were either on the periphery of the plot, or on higher regions within the plot. They probably then did not absorb the full amount of N applied, and may instead have received lower doses that enabled them to produce more new growth. A n alternative explanation is that the surviving E. nigrum had less competition for resources and space and could have utilized the extra resources for growth (Grime 1979). This lack of results in shrub growth is not surprising as the growth rate is slow and based on nutrient availability over several years (Chapin 1980, Chapin et al. 1990). However, N addition should have stimulated some growth in G. lividum in 2004 and 2005 as herbaceous plants are adapted for quicker uptake of N compared to the shrubs (Chapin 1980, Chapin et al. 1990). This growth was not seen in 2004 and all treatments had very similar measurements. In 2005 though a visible but non-significant difference occurred among the treatments with higher levels of N producing some of the bigger plants. In fact, in all treatments the plants grew more in 2005 than they did in 2004. If, as mentioned earlier, G. lividum were utilizing N added that season then G. lividum should have shown similar growth,rates in both 2004 and 2005. However, in 2004 N was not added to the plots until early June as G. lividum first had to grow tall enough so that I could know where to place the plots. Thus the plants had already completed about half their growth, and treatment effects would not be as noticeable. Also, the optimum time for addition of N in some plants seems to be earlier in the spring (Hanson and Retamales 1992, Eaton 1994, Penney et al. 2003) and perhaps this is true for G. lividum as well. Adding N to G. lividum plots this  24  coming spring may help confirm these results if a similar trend of larger growth with increasing N levels is seen. Leaf breadth and berry weight in G. lividum were strongly correlated possibly because larger leaves have more area for photosynthesis that would then produce more sugars needed for the berries.  2.5 Conclusions Two lines of evidence argue in favour of the fecal nutrient enrichment hypothesis as a mechanism for producing a bumper crop of berries 2-3 years after the snowshoe hare peak: 1. The nitrogen level having the most consistent, positive impact was between 0.5 and 1.0 g/m . This is within the range of nutrients expected to be released following a hare peak. 2  2. As expected, levels of nitrogen addition above 2 g/m either had no impact or had a 2  negative impact, echoing the conclusion that these berry producing shrubs are adapted to a low nutrient environment and can not cope with a switch to an environment with abundant nutrients. While berry production increased on the small 1 m plots at low levels of N addition this 2  does not necessarily mean the same effect would occur on a larger scale such as the 2.8 ha small mammal live-trapping grids. The small plots were not randomly selected, and the shrub within the plot occupies most of the plot. By contrast, greater shrub heterogeneity in the trapping grids and possibly more resource competition between different shrubs or other plant species may alter berry production. An important consideration not covered in this study is the landscape abundance of these plants. Empetrum nigrum may produce more and heavier berries but if the plants themselves have overall low abundance (e.g. 10% of the food plants) then they probably would not affect the C. rutilus populations on a large scale. Further study is needed to determine the landscape abundance of each of these plant species. Overall having only one plant species out of 5 exhibit significantly heavier and more .. berries at low levels of N is not strong support for the fecal nutrient enrichment hypothesis even with the trends in the other plants. The lack of a strong response in berry production may be attributed to the poor weather conditions in 2005, but until further evidence is available this experiment offers only tentative support for the fecal nutrient enrichment hypothesis.  25  Figure 2.1: General location of the study site. Image adapted from 2005 TeleAtlas, Image 2005 EarthSat, GoogleEarth.  26  Figure 2.2: Arctostaphylos uva-ursi plot 60 (U-60), diameter 1.15 m. North is located at the top of the picture.  27  Figure 2.3: Vinyl quarter-circle used for measuring percent cover of vegetation within the 1 m plots. R l to the apex of the quarter-circle covers 11% of the vinyl. R l to R2 covers 33% and R2 to R3 covers 56% of the quarter-circle. The R l area is subdivided in two sections with R2 and R3 regions subdivided into 4 sections to facilitate more accurate percent measurements. 2  28  Table 2.1: A N O V A table P-values, F-values for 2004 and 2005 for berry and vegetative growth of the berry producing species. Degrees of freedom for vegetative growth = 54 except for Arctostaphylos rubra where d.f. - 53. n.a. = not available. P-values, F-values for 2004 Berry Berries d.f. Plant Weight rn2 berries Arctostaphylos uva-ursi  0.84, 0.41  0.15, 1.72  36  Arctostaphylos rubra  0.59, 0.75  0.60, 0.75  53  Empetrum nigrum  0.24, 1.41  0.91, 0.30  54  Shepherdia canadensis  0.73, 0.57  0.29, 1.21  37  Geocaulon lividum  0.71, 0.59  0.76, 0.51  50  Berry Weight  Berries/m2  d.f. berries  Arctostaphylos uva-ursi  0.82, 0.44.  0.72, 0.58  54  Arctostaphylos rubra  0.14, 1.74  0.14, 1.74  54  Empetrum nigrum  0.01, 2.82  0.04, 2.39  53  Shepherdia canadensis  0.13, 1.81  0.41, 0.99  53  Geocaulon lividum  0.08, 2.11  0.87, 0.37  46  New Growth  Leaf Breadth  Leaf Length  0.30, 1.19  0.88, 0.33  0.27, 1.35  n.a.  0.73, 0.57  0.13, 1.78  Empetrum nigrum  0.01, 3.49  n.a.  n.a.  Shepherdia canadensis  0.08, 2.12  0.21, 1.07  0.19, 1.55  Geocaulon lividum  0.96, 0.20  0.94 0.26  0.78, 0.51  New Growth  Leaf Breadth  Leaf Length  0.25, 1.37  0.21, 1.42  0.19, 1.57  n.a.  0.25, 1.39  0.29, 1.27  P-values, F-values for 2005 Plant  VEGETATIVE GROWTH P-values, F-values for 2004 Plant Arctostaphylos uva-ursi Arctostaphylos rubra  P-values, F-values for 2005 Plant Arctostaphylos uva-ursi Arctostaphylos rubra  n.a.  Empetrum nigrum  0.01, 3.28  , n.a.  Shepherdia canadensis  0.56, 0.74  0.97, 0.19  0.81, 0.45  Geocaulon lividum  0.59, 0.75  0.07, 2.19  0.29, 1.27  29  35  i  -5 -10 Treatment g N/m  2  Figure 2.4: Arctostaphylos rubra berries/m in 2004 and 2005, with standard error bars. 2  30  18 -, eg  E  16 -  Treatment g N/m  2  Figure 2.5: Arctostaphylos uva-ursi berry numbers in 2004 and 2005 at different levels of nitrogen (N) treatment. Bars are standard error.  31  32  140 120 in  .1 £  100 80  •I'-  r'  -2004 — 2005  60 40 20 0 -20  0.0  0.5  1.0  4.0  2.0  Treatment g N/m  17.5  2  Figure 2.7: Empetrum nigrum berries/m for 2004 and 2005 with different treatment levels of nitrogen (N). Bars are standard error. 2  33  0.45 0.40  - - -A- •-  0.35 g>  Geocaulon lividum  • — — Arctostaphylos  rubra  0.30  - Empetrum nigrum  _g 0.25  • —  0.20  - Shepherdia canadensis  0.15 0.0  0.5  1.0  2.0  Treatment g N/m  4.0  17.5  2  Figure 2.8: Berry weight (g) in 2005 of 3 shrubs and one herbaceous plant (Geocaulon lividum). Bars are standard error.  34  15.5  0.30 ~  0.25  4-  0.20 CD  Is  0.15  15.0 "E E 14.5 ^ •g 14.0 S To 13.5 CD 13.0  I  2 12.5 c o 12.0 TO  c 0.10 o 3 CO  0.05 A  CD  11.5 CD  o 0.00  11.0 0.0  0.5  1.0  4.0  2.0  Treatment g N/m  17.5  2  Figure 2.9: Geocaulon lividum berry weight and leaf width correlation (r=0.74). Bars are standard error.  35  -berry weightj 2005  leaf width 2005  2.6 Literature Cited Belonogova, T.V. 1988. Yield forecasting and optimization of berry harvesting in the forests of Southern Karelia. USSR. In Proceedings of the Finnish-Soviet symposium on timber forest resources in Jyvaskyla, Finland. 25-29 August 1986. Edited by I. Vanninen and M . Raatikainen. Acta. Bot. Fenni. 136:19-21. Helsinki. Bliss, L. C. 1971. Arctic and alpine plant life cycles. Annu. Rev. Ecol. Syst. 2:405-438. Bonan, G. B., and H . H . Shugart. 1989. Environmental factors and ecological processes in boreal forests. Annu. Rev. Ecol. Syst. 20:1-28. Boonstra, R., C. J. Krebs, S. Gilbert, and S. Schweiger. 2001. Chapter 10: Voles and Mice. Pages 215-239 in C. J. Krebs, S. Boutin, and R. Boonstra, editors. Ecosystem Dynamics of the Boreal Forest: The Kluane Project. Oxford University Press, New York. Boutin, S., C.J. Krebs, R. Boonstra, M.R.T. Dale, S.J. Hannon, K.Martin, A.R.E. Sinclair, J.N.M. Smith, R. Turkington, M . Blower, A. Byrom, F.I. Doyle, C. Doyle, D. Hik, L. Hofer, A . Hubbs, T. Karels, D.L. Murray, V. Nams, M . O'Donoghue, C. Rohner, and S. Schweiger. 1995. Population changes of the vertebrate community during a snowshoe hare cycle in Canada's boreal forest. Oikos 74:69-80. Chapin, III, F.S. 1980. The mineral nutrition of wild plants. Annu. Rev. Ecol. Syst. 11:233-260. Chapin, III, F.S., E.D. Schulz, H.A. Murray. 1990. The ecology and economics of storage in plants. Annu. Rev. Ecol. Syst. 21:423-427. Dale, M.R.T., S. Francis, C.J. Krebs, and V.O. Nams. 2001. Plant dynamics: Trees. Ecosystem Dynamics of the Boreal Forest: The Kluane Project. Edited by C.J. Krebs, S. Boutin and R. Boonstra. pp. 116-137. Oxford University Press, New York. Eaton, L. J. 1994. Long-term effects of herbicide and fertilizers on lowbush blueberry growth and production. Can. J. Plant Sci. 74:341-345. Francis, S.R. 1996. Linking landscape pattern and forest disturbance: fire history of the Shakwak . Trench, southwest Yukon Territory. M.Sc. Thesis, Univ. Alberta, Edmonton, Alberta. Fuller, W.A. 1977. Demography of subarctic population of Clethrionomys gapperi numbers and survival. Can. J. Zool. 55:42-51. Gilbert, B.S. and C.J. Krebs. 1981. Effects of extra food on Peromyscus and Clethrionomys populations in the southern Yukon. Oecologia 51:326-331. Grime, J.P. 1979. Plant strategies and vegetative processes. Wiley Press. Chinchester, New York, NY.  36  Hanson, E. J., and J. B. Retamales. 1992. Effect of nitrogen source and timing on highbush blueberry performance. Hort. Sci. 27:1265-1267. Lovblad, G., B . Andersen, M . Hovmand, S. Joffre, U . Pedersen, and A. Reisell. 1992. Mapping deposition of sulphur, nitrogen and base cations in the Nordic countries. I V L Report B 1055. Swedish Environmental Research Institute, Goteborg, Sweden. Mols, T., J. Paal, and E. Fremstad. 2000. Response of Norwegian alpine communities to , nitrogen. Nord. J. Bot. 20:705-712. Nordin, A., T. Nasholm, and L. Ericson. 1998. Effects of simulated N deposition on understorey vegetation of a boreal coniferous forest. Func. Ecol. 12:691-699. Press, M . C , J. A. Potter, M . J. W. Burke, T. V . Callaghan, and J. A . Lee. 1998. Responses of a subarctic dwarf shrub heath community to simulated environmental change. J. Ecol. 86:315-327. Penney, B.G., K . B . McRae, and G.A. Bishop. 2003. Second-crop N fertilization improves lowbush blueberry (Vaccinium angustifolium Ait.) production. Can. J. Plant Sci. 83:149155. Raatikainen, M . , and M . Niemela. 1994. The effect of fertilization on the yield of wild forest berries pp. 123-129. In Effects of Fertilization on Forest Ecosystems. University of Jyvaskyla. Jyvaskyla, Finland. Biol. Res. Rep. Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of fertilization and herbivory on the herbaceous vegetation of the boreal forest in north-western Canada: a 10-year study. J. Ecol. 90:325-337. West, S. D. 1982. Dynamics of colonization and abundance in central Alaskan populations of the northern red-backed vole, Clethrionomys rutilus. J. Mamm. 63:128-143. Yudina, V. F., and T. A. Maksimova. 2005. Dynamics of yielding capacity of small cranberry in southern Karelia. Rus. J. Ecol. 36:239-242.  37  Chapter 3: Do changes in berry production drive Clethrionomys rutilus population changes? 3.1 Introduction Multi-annual fluctuations in the northern red-backed vole (Clethrionomys rutilus Pallas) and other small mammals in boreal forests have been studied since Charles Elton's (1924) classic paper highlighted the phenomenon in voles and lemmings. Since then many others have studied and reviewed small mammal cycles (Krebs and Myers 1974, Hansson and Henttonen 1985, Norrdahl 1995, Krebs 1996, Stenseth 1999, Lindstrom et al. 2001, Norrdahl and Korpimaki 2000, Korpimaki et al. 2005b), and various hypotheses have been proposed to explain small mammal irruptions including the fecal nutrient enrichment hypothesis that I will be testing. During snowshoe hare (Lepus americanus) peak abundance, large quantities of fecal pellets are deposited that slowly release nitrogen (N) into the ground (Boonstra et al. 2001). The hypothesis proposes that N released from pellets is taken up by berry producing shrubs and facilitates production of a large crop of berries the following year. These berries provide a high quality overwintering food supply for Clethrionomys spp. and increases their overwintering survival leading to an increase in Clethrionomys density two to three years after snowshoe hare peaks. Similar delayed responses in other vertebrate populations are also seen in other ecosystems. For example, in central California two years after Quercus spp. (oaks) produce a large crop of acorns due to a nutrient pulse, ground or low shrub nesting bird populations decrease because of a higher rate of rodent predation that increased in response to the acorn surplus (Koenig and Knops 2005). This type of ripple affect can effect community composition for years (Koenig and Knops 2005). In the Kluane region of south-western Yukon, C. rutilus should reach a population peak as snowshoe hare densities are declining, according to the fecal nutrient enrichment hypothesis. Snowshoe hares have an approximate 10-year cycle between periods of high density, and Gilbert et al. (1986) reported an 11-year interval between C. rutilus population peaks. In the Kluane region there were three major red-backed vole population peaks, 11 and 8 years apart in 1973, 1984 and 1992 (Gilbert and Krebs 1991, Boonstra et al. 2001, Boonstra and Krebs pers. comm.) (Figure 3.1).  38  Two smaller peaks also occurred in 1979 and 1987 (Gilbert and Krebs 1991). The major peaks in C. rutilus occurred 2 years after the snowshoe hare peak which suggests a vole-hare link with some other factor or factors producing the smaller vole peaks. In most springs C. rutilus numbers are at low density, build to high levels in the summer and fall, and then, towards the end of winter, experience a sudden decline, sometimes by up to 50% (Fuller et al. 1969, Merritt and Merritt 1978). With low density and high density reproductive rates being fairly similar (Fuller 1985, Korpimaki et al. 2004) increases in redbacked vole densities must occur following high rates of overwinter survival (Gilbert et al. 1986, Gilbert and Krebs 1991) as opposed to increases in reproductive rates. Thus it is the overwinter survival rate that is the critical determinant of the size of the population in the spring (Fuller 1977, Gilbert and Krebs 1991). If overwintering survival is important for increasing red-backed vole populations, then what factors lead to increased winter survival rates? Fuller (1977) reported that favourable winter weather was not necessary for increasing numbers of the closely related Clethrionomys gapperi (southern red-backed vole), nor were favourable winters sufficient to explain population increases. Several studies suggest that poor winter food quality may limit vole population size (Cole and Batzli 1979, Hansson 1999) so additional high quality food in the winter should reduce mortality rates thereby leading to higher spring numbers. Low availability of quality food may lead to low body weights, lack of reproduction, and higher levels of mortality (many studies reviewed in Cole and Batzli 1978, Korpimaki et al. 2004), and increasing the quantity of food available led to a greater density of red-backed voles (Gilbert and Krebs 1981, Taitt and Krebs 1981, Krebs and Wingate 1985, Gilbert and Krebs 1991, Schweiger and Boutin 1995). Any changes in the availability of quality food may alter the density in vole populations (Cole and Batzli 1978, Batzli 1986, Korn and Taitt 1987, Agrell et al. 1995, Boonstra et al. 1998). Berries, with their high caloric seeds, provide a nutritional source of food for C. rutilus, who seem to rely on berries more than other small mammals (Dyke 1971, West 1979, 1982, Hansson 1985). Dyke (1971) reported that berries were almost the only food items in the stomachs of Clethrionomys during a year when berries were very abundant. Likewise West (1982) discovered that berries and seeds comprised 62%-92% of the summer and fall diet in C. rutilus, and voles trapped from under a meter of snow had primarily berries in their stomachs. In  39  C. gapperi, Merritt and Merritt (1978) reported that seeds made up 75% of the diet in winter. Clethrionomys gapperi may specialize on different foods than C. rutilus, but seeds and fruits seem to be an important part of both their diets (Hansson 1971,1985). Berries may occur in abundance during years of favourable conditions. During this abundance all berries may not be eaten in the fall, and the snow covers the remaining berries, and preserves them thus providing a food supply for subnivean mammals throughout the winter and into early spring (West 1982). With berries providing a high quality food vole mortality should decline (Korpimaki et al. 2004) resulting in an increase in vole numbers. What, then, causes berry crops to vary? Given favourable climatic conditions (early spring, warm growing season, plentiful moisture, absence of late frosts, adequate snow cover) soil nutrients, especially N may be a limiting factor (McKendrick et al. 1980, Nams et al. 1993, Turkington et al. 1998). While higher levels of N may decrease shrub production through increased competition with herbaceous and graminoids species (Nordin et al. 1998, Press et al. 1998, Turkington et al. 2002) low levels of 1 to 2 g N/m , such as the amount introduced from 2  snowshoe hare fecal pellets every 9 to 10 years when hares are at the peak of their cycle (Hik pers. comm.), may increase berry production in these shrubs (Hanson and Retamales 1992, Niittymaa 1983 in Raatikainen and Niemela 1994, Penney et al. 2003). To test the hypothesis that snowshoe hare fecal pellets contribute enough N to increase shrub berry production and thereby increasing C. rutilus densities two to three years after snowshoe hares (the fecal nutrient enrichment hypothesis) I will use available data from the Kluane Ecological Monitoring Project as well as data from my own field experiments. This hypothesis can be broken into two sub-hypotheses: 1) Additional small amounts of N , such as those contributed by hare pellets, will increase berry production in shrubs the following summer. Chapter 2 discusses this hypothesis using field experiments designed to test for a correlation between added N and berry production. 2) Changes in berry production will be reflected by changes in northern red-backed vole densities a year later. This latter hypothesis is tested in this chapter by examining berry production indices and C. rutilus trapping data. Future work will compare vole populations on N treated grids and controls with berry production indices on those grids. Using berry and C. rutilus data collected during the Kluane Ecological Monitoring Project, I will test the hypothesis that berry production indices are correlated to C. rutilus  40  densities, with a 1-year time lag in response by the voles. I predict that increased berry production will result in greater winter survival of C. rutilus and vole populations should increase in the following spring. 3.2 Methods 3.2.1 The study area The study site was located in the southwestern Yukon Territory near Kluane Lake by the Alaska Highway within the Shakwak Trench system (61° 01 ' N , 138° 24' W), and lies within the rain shadow of the St. Elias Mountains. Mean annual precipitation is ca. 280 mm and includes an average annual snowfall of approximately 100 cm (1945-2003). The tree community is dominated by white spruce (Picea glauca (Moench) Voss) interspersed with trembling aspen (Populus tremuloides Michx.) and balsam poplar (Populus balsamifera L.). The upper shrub layer is composed of willow (Salix spp), soapberry (Shepherdia canadensis (L.) Nutt.) and dwarf birch (Betula glandulosa Michx.) while the ground layers are composed of dwarf shrubs and herbaceous plants such as bearberries (Arctostaphylos rubra (Rehd. & Wils.) Fern, and A.uvaursi (L.) Spreng. s.l)., crowberry (Empetrum nigrum L.), blueberry, (Vaccinium spp.), cranberry (Vaccinium vitis-idaea L . ssp. minus (Lodd.) Hulten.), toadflax (Geocaulon lividum (Richards) Fern.), arctic lupine (Lupinus arcticus S. Wats), and other forbs (Turkington et al. 2002). 3.2.2 Trapping methods Clethrionomys rutilus data were averaged from live trapping data taken from several studies over the years (see Gilbert and Krebs 1981, Gilbert and Krebs 1991, Boonstra et al. 2001) that used 2.8 ha grids. Each grid had 100 stations 15 m apart in a 10x10 array with 50 Longworth traps at every other station. Traps were pre-baited with seed oats for a week before trapping, and where necessary traps were placed inside mesh cages to prevent squirrels from triggering the trap. Traps were left in place all year. Trapping sessions were conducted all summer, and small mammals captured were tagged on the right ear with numbered fingerling fish tags. (See Gilbert and Krebs 1981 for methods). 3.2.3 Berry production indices Vox Arctostaphylos rubra, A. uva-ursi, E. nigrum, V. vitis-idaea and G. lividum berry data selected plots were 0.8 m x 0.4 m in size and consisted of two 0.4 x 0.4 m quadrats laid side by 2  side for a total of 100 quadrats (50 sites x 2 quadrats each). Sites were placed systematically at 100 m intervals on snowshoe hare trapping grids at grid point that had adequate plant coverage.  41  They are meant to provide an index of berry production to measure year to year changes in berry counts. A l l ground berries within the plots were counted while still green, typically in early to mid-July, to minimize the number lost to mammals and birds. Shepherdia canadensis berries were counted on the same two stems per bush every year in July while the berries were still green. Two stems were randomly selected and marked with aluminium tags. Stem diameter was measured at the base in millimetres. Berry records are only available for A. uva-ursi from 1995, and the other shrubs from 1997 onwards. Thus correlation analysis is limited to a 9-year period from 1997 to 2005 for all of the shrubs except A. uva-ursi. 3.2.4 Statistical analysis For the period of 1997-2005 average C. rutilus/ha were provided by the Kluane Ecological Monitoring Project from Schnabel estimates of mark-recapture data based on trapping grids within the Shakwak Trench. Numbers of C. rutilus in the spring were highly correlated (r = 0.94) with numbers of C. rutilus recorded in the summer and consequently only C. rutilus estimates for the spring were used in the analyses. Berry numbers for A. rubra, A. uva-ursi., E. nigrum, G. lividum, and V. vitis-idaea were standardized to determine a mean number of berries per 50% cover. Shepherdia canadensis berry numbers per branch were standardized to a 10 mm diameter branch using an average slope of 0.7105 from the combined regression of berry numbers on branch diameter from 1997-2001 data using the following equation: # berries = ((^(Observed # of berries) + ((10 x Observed diameter) x 0.7105)))  2  and then bootstrapped to obtain the mean for the soapberry counts from each area. A square root transformation was used to stabilize the variance of the regression. The resultant standardized numbers of berries are meant as an index of soapberry production and not as an absolute estimate per unit area. Pearson correlations between yearly C. rutilus and yearly berry numbers with a 1-year lag were done using SPSS 11.  42  3.3 Results 3.3.1 Berry production In 2001 berry production showed a sharp increase for A. uva-ursi, A. rubra, E. nigrum, S. canadensis and V. vitis-idaea (Figure 3.2). By contrast, G. lividum did not increase until 2003 followed by an even higher number in 2005. Berry production in all species decreased quickly after 2001 except in G. lividum which continued to increase, and in A. rubra which increased slightly for another year before declining. Geocaulon lividum had a large increase in 1998, 2003, and 2005 which are the same years that the dwarf berry crops were either decreasing or remaining stable although overall no positive or negative correlation between G. lividum and the other woody shrubs exists (r=0.03). In 2001 and 2004 G. lividum berry production declined while the dwarf berry production increased strongly. 3.3.2 Relationship between berry production and vole numbers Of the 5 shrubs and one herbaceous plant studied, only E. nigrum, V. vitis-idaea, and S. canadensis showed a marginal positive correlation (P<0.15) with C. rutilus numbers (Table 3.1). Of these three E. nigrum had the strongest correlation to C. rutilus numbers, but it was only marginally significant (r = 0.66, P = 0.07). No significant individual correlations existed for any of the other species. A stronger correlation was seen, however, using a backward elimination multiple regression analysis of C. rutilus/ha and berry data. Arctostaphylos rubra, E. nigrum, and 5. canadensis together had the strongest correlation to C. rutilus numbers in the following year (Figure 3.3). The equation of the predictive line is Y = 0.23x - 1.31y + 0.06z (r = 0.92) where Y is the number of C. rutilus/ha, x is the index of production for E. nigrum (P = 0.002), y is the same for A. rubra (P = 0.005), and z is the same for S. canadensis (P = 0.059) (Figure 3.4). Using the berry production data for 2005 in this equation predicts C. rutilus numbers in spring 2006 to be 3.4/ha (± 0.6).  3.4 Discussion Berry production of A. rubra, E. nigrum and S. canadensis together had the strongest correlation to C. rutilus densities over a 9-year period with C. rutilus showing a 1-year time lag in response to increased berry production (Figure 3.3). This supports the hypothesis that increases in berry production are positively correlated to C. rutilus densities the following year.  43  Since data for berry production are available for only the past 9 years, continued monitoring of berry production and C. rutilus densities is needed. Providing berries are the key winter food as suggested by West (1979, 1982), longer term monitoring should then indicate that C. rutilus densities are closely tied to berry production. To confirm a correlation between high hare densities and an increase in berry production would require berry and hare data through at least two complete hare cycles. This correlation would be much more difficult to tease out because peaks may also be seen in berry production independent of hare densities. More importantly berry production is closely linked to weather factors especially frosts during the flowering period, and these late frosts or cooler temperatures can reduce berry production (Belonogova 1988, Kuchko 1988) despite high hare densities. However, hare fecal pellets release N for about 3 years (Hiks pers. comm.) and the odds of unfavourable weather conditions in all three years are lower, thus large numbers of pellets could still provide enough N to stimulate a berry crop in one of those years. Furthermore, ericaceous shrubs can store N for use when N supplies are low (Chapin et al. 1990) so N taken up one year may be used for berry production in a year when conditions are more favourable. It would be important to carefully choose which plants are to be compared with C. rutilus numbers. Geocaulon lividum had a negative although insignificant correlation with C. rutilus numbers while the shrubs had positive correlations (Table 3.1). As a herbaceous plant G. lividum is tied to weather and nutrient conditions in the current summer, but the woody shrubs to conditions in the previous summer (Kuchko 1988), or over several years (Chapin 1980, Chapin et al. 1990). Therefore it is not surprising that G. lividum berry production seems to be out of synchrony with berry production from woody shrubs.  3.5 Conclusion The strong positive correlation of E. nigrum, V. vitis-idaea and S. canadensis production with C. rutilus numbers provides evidence for the second half of the fecal nutrient enrichment hypothesis, that is, that berry production has a measurable relationship to C. rutilus populations. To date though there have been only 9 to 11 years of berry monitoring and several more years will be necessary before strong conclusions are drawn about the predicted effect that berry production has on C. rutilus.  44  If increased berry production results in higher C. rutilus densities the following year, then long-term data should show a positive correlation between higher berry production numbers and increased over-winter survival rates, increased body weights, and perhaps earlier breeding and increased litter size in C. rutilus. These assumptions could also be tested in a similar manner to previous food addition experiments by collecting and freezing berries during years of high berry abundance, and redistributing them just before snowfall on small mammal live-trapping grids during years of low berry abundance. Periodic trapping, tagging and weighing of animals throughout the winter and into the spring would provide further evidence for or against the hypothesis that berry production and C. rutilus numbers are correlated.  45  3.0  1975  1980  1985  1990  1995  2000  2005  Figure 3.1: Clethrionomys rutilus spring density (histogram) and snowshoe hare spring density (line) from 1973 to 2005 in the Kluane Lake region, Yukon. Clethrionomys rutilus data is derived from the following sources: 1973-1975 Gilbert and Krebs (1981); 1976-1986, Gilbert and Krebs (1991); 1987-2005 Krebs et al. (unpublished data). Hare data is from the following sources: 1971-1975, Keith (1990); 1976-1986, Krebs et al. (1986) and Boutin et al. (1995); 1987-1996, Krebs et al. (1995), 1997-2005, Krebs et al. (unpublished data). Citations from Boonstra et al. 2001. Figure courtesy of Kluane Ecological Monitoring Project.  46  o-  25  (a) 20 15 10  1 o  (b)  8  1  6  2  li 1  4 2  k2'  •d  3  o 0  Z  1 40  (d)  1 20 1 00 80 60 40 20 0  *1  (e) 30 A  2 0  1 0  1994  1996  1998  2000  2002  I  2004  2006  Year  Figure 3.2: Berry production of 4 shrubs (a) Arctostaphylos uva-ursi, (b)Arctostaphylos rubra, (c) Empetrum nigrum, (d) Shepherdia canadensis and 1 herbaceous plant (e) Geocaulon lividum from the Shakwak Trench region of south-western Yukon. Geocaulon lividum can be seen increasing and decreasing out of phase with the dwarf shrubs.  47  Table 3.1: Mean berry densities and parametric correlations between berry numbers and spring Clethrionomys rutilus population densities from 1997-2005 (1995-2005 for Arctostaphylos uvaursi) with a one year time lag (e.g. berries in 1997 correlated with voles in 1998). Degrees of r-value spring  Pvalue  mean berries/m  S.E.  range  Arctostaphylos uva-ursi  0.21  0.57  9.8  0.7  3.8-19.4  Arctostaphylos rubra  0.14  0.74  5.8  0.5  3.1 - 10.1  Empetrum nigrum  0.67  0.07  29.0  1.1  13.5-52.8  Vaccinium vitis-idaea  0.62  0.10  19.0  1.3  0-55.9  Geocaulon lividum  -0.40  0.32  18.6  0.9  5.4-34.1  Shepherdia canadensis  0.56  0.15  60.7  2.0  19.1 - 140.0  Plant  48  2  Figure 3.3: Total mean berry production/m of the shrubs Shepherdia canadensis (n=100/yr), Arctostaphylos rubra (n=100/yr), and Empetrum nigrum (n=100/yr) (line) plotted against mean number of spring Clethrionomys rutilus/ha (histogram) measured from 3 small mammal trapping grids between 1997-2005 at Kluane Lake, south-west Yukon Territory. Note there is a 1-year lag between high berry production and C. rutilus densities (r = 0.92, 9d.f.) (i.e. berries at time t, and voles at time t+1). 2  49  14  n  Year  Figure 3.4: Results of the predictive equation of Clethrionomys rutilus densities compared with actual Clethrionomys rutilus densities (r = 0.92). Densities were predicted using the following equation: Density of C. rutilus = 0.23 Empetrum nigrum -1.31 Arctostaphylos rubra + 0.06 Shepherdia canadensis with P = 0.002, P = 0.005, P = 0.007 for each of the shrubs respectively, where indices of berry production are used for each shrub. Bars are standard error.  50  3.6 Literature Cited Agrell, J., S. Erlinge, and J.M. Nelson. 1995. Delayed density-dependence in a small-rodent population. Proc. R. Soc. Lond. B . 262:65-70. Batzli, G.O. 1986. Nutritional ecology of California vole: effects of food quality on reproduction. Ecol. 67:406-412. Belonogova, T.V. 1988. Yield forecasting and optimization of berry harvesting in the forests of Southern Karelia. USSR. In Proceedings of the Finnish-Soviet symposium on timber forest resources in Jyvaskyla, Finland. 25-29 August 1986. Edited by I. Vanninen and M . Raatikainen. Acta. Bot. Fennica 136:19-21. Helsinki. Boonstra, R., C. J. Krebs, S. Gilbert, and S. Schweiger. 2001. Chapter 10: Voles and Mice. Pages 215-239 in C. J. Krebs, S. Boutin, and R. Boonstra, editors. Ecosystem Dynamics of the Boreal Forest: The Kluane Project. Oxford University Press, New York. Boonstra, R., C. J. Krebs, and N.C. Stenseth. 1998. Population cycles in small mammals: the problem of explaining the low phase. Ecol. 79:1479-1488. Boutin, S., C. J. Krebs, R. Boonstra, M . R. T. Dale, S. J. Hannon, K. Martin, A-. R. E. Sinclair, J. N . M . Smith, R. Turkington, M . Blower, A. Byrom, F. I. Doyle, C. Doyle, D. Hik, L . Hofer, A. Hubbs, T. Karels, D. L . Murray, V. Nams, M . O'Donoghue, C. Rohner, and S. Schweiger. 1995. Population changes of the vertebrate community during a snowshoe hare cycle in Canada's boreal forest. Oikos 74:69-80. Chapin, III, F.S. 1980. The mineral nutrition of wild plants. Annu. Rev. Ecol. Syst. 11:233-260. Chapin, III, F.S., E.D. Schulz, H.A. Murray. 1990. The ecology and economics of storage in plants. Annu. Rev. Ecol. Syst. 21:423-427. Cole, F.R., and G.O. Batzli. 1979. Nutrition and population dynamics of the prairie vole, Microtus ochrogaster, In Central Illinois. J. Ani. Ecol. 48:455-470. Dyke, G. R. 1971. Food and cover of fluctuating populations of northern cricetid. Ph.D. Thesis, Univ. Edmonton, Alberta. Elton, C.S. 1924. Periodic fluctuations in the numbers of animals: their causes and effects. Br. J. Exp. Biol. 2: 119-163. Fuller, W. A . 1969. Changes in numbers of three species of small rodents near Great Slave Lake, N.W.T. Canada, 1964-1967, and their significance for general population theory. Ann. Zool. Fennici. 6:113-144. Fuller, W. A . 1985. Clethrionomys gapperi - is there a peak syndrome. Ann. Zool. Fennici. 22:243-255.  51  Fuller, 'w. A. 1977. Demography of subarctic population of Clethrionomys gapped numbers and survival. Can. J. Zoo. 55:42-51. Gilbert, B. S., D. B. Cichowski, D. Talarico, and C. J. Krebs. 1986. Summer activity patterns of three rodents in the Southwestern Yukon. Arct. 39:204-207. Gilbert, B. S., and C. J. Krebs. 1981. Effects of extra food on Peromyscus and Clethrionomys populations in the southwestern Yukon. Oecol. 51:326-331. Gilbert, B.S. and C.J. Krebs. 1991. Population dynamics of Clethrionomys and Peromyscus in southwestern Yukon 1973-1989. Hoi. Ecol. 14: 250-259. Hansson, L. 1999. Intraspecific variation in dynamics: small rodents between food and predation in changing landscapes. Oikos 86 (1): 159-169. Hansson, L. 1985. Clethrionomys food—generic, specific and regional characteristics. Ann. Zool. Fennici.22(3): 315-318. Hansson, L. 1971. Small rodent food, feeding and population dynamics—a comparison between granivorous and herbivorous species in Scandinavia. Oikos 22: 183. Hansson, L., and H. Henttonen. 1985. Regional differences in cyclicity and reproduction in Clethrionomys species—are they related. Ann. Zool. Fennici. 22:277-288. Hanson, E. J., and J. B . Retamales. 1992. Effect of nitrogen source and timing on highbush blueberry performance. Hort. Sci. 27:1265-1267. Keith, L.B. 1990. Dynamics of snowshoe hare populations. Curr. Mamra. 2:119-195. Koenig, W.D. and J.M.H Knops. 2005. The mystery of masting in trees. Am. Sci. 93: 340-347. Korn, H., and M . J. Taitt. 1987. Initiation of early breeding in a population of Microtus townsendii (Rodentia) with the secondary plant compound 6-MBOA. Oecol. 71:593-596. Korpimaki, E., P. R. Brown, J. Jacob, and R. P. Pech. 2004. The puzzles of population cycles and outbreaks of small mammals solved? BioSci. 54:1071-1079. Korpimaki, E., L. Oksanen, T. Oksanen, T. Klemola, K. Norrdahl, and P. B . Banks. 2005b. Vole cycles and predation in temperate and boreal zones of Europe. J. Ani. Ecol. 74:11501159. Krebs, C. J. 1996. Population cycles revisited. J. Mamm. 77:8-24.  52  Krebs, C.J., S. Boutin, R. Boonstra, A R.E. Sinclair, J.N.M. Smith, M.R.T. Dale, K. Martin, and R. Turkington. 1995. Impact of food and predation on the snowshoe hare cycle. Sci. 269:1112-1115. Krebs, C. J., and J. H . Myers. 1974. Population cycles in rodents. Sci. Amer. 230:38-46. Krebs, C. J., and I. Wingate. 1985. Population fluctuations in the small mammals of the Kluane , Region, Yukon Territory. Can. F. Nat. 99:51-61. Krebs, C.J., B.S. Gilbert, S. Boutin, A.R.E. Sinclair, and J.N.M. Smith. 1986. Population biology of snowshoe hares. I. demography of food-supplemented populations in the southern Yukon, 1976-1984. J. Anim. Ecol. 55: 963-982. Kuchko, A . A . 1988. Bilberry and cowberry yields and the factors controlling them in the forests of Karelia, USSR. In Proceedings of the Finnish-Soviet symposium on timber forest resources in Jyvaskyla, Finland. 25-29 August 1986. Edited.by I. Vanninen, and M . Raatikainen. Acta. Bot. Fennica 136:23-25. Lindstrom, J., E. Ranta, H . Kokko, P. Lundberg, and V . Kaitala. 2001. From arctic lemmings to adaptive dynamics: Charles Elton's legacy in population ecology. Biol. Rev. 76:129-158. McKendrick, J.D., G.O. Batzli, K. R. Everett, and J.C. Swanson. 1980. Some effects of mammalian herbivores and fertilization on tundra soils and vegetation. Arct. Alp. Res. 12:565-578. Merritt, J. F., and J. M . Merritt. 1978. Population ecology and energy relationships of Clethrionomys gapperi in a Colorado subalpine forest. J. Mamm. 59:576-598. Nams, V.O., N.F.G. Folkard, and J.N.M. Smith. 1993. Effects of nitrogen fertilization on several woody and nonwoody boreal forest species. Can. J. Bot. 71:93-97. Niittymaa, L. 1983. Puolukan iannoituskokeista. Metsantutkimuslaitoksen Tiedonantoja 90:161163. Nordin, A., T. Nasholm, and L. Ericson. 1998. Effects of simulated N deposition on understorey vegetation of a boreal coniferous forest. Func. Ecol. 12:691-699. Norrdahl, K. 1995. Population cycles in northern small mammals. Biol. Rev. Cambr. Philo. Soc. 70:621-637. Norrdahl, K., and E. Korpimaki. 2000. The impact of predation risk from small mustelids on prey populations. Mamm. Rev. 30(3-4): 147-156. Penney, B.G., K . B . McRae, and G.A. Bishop. 2003. Second-crop N fertilization improves lowbush (Vaccinium angustifolium Ait.) production. Can. J. Plant Sci. 83:149-155.  53  Press, M . C , J. A. Potter, M . J. W. Burke, T. V. Callaghan, and J. A. Lee. 1998. Responses of a subarctic dwarf shrub heath community to simulated environmental change. J. Ecol. 86:315-327. Raatikainen, M . , and M . Niemela. 1994. The effect of fertilization on the yield of wild forest berries pp. 123-129. In Effects of Fertilization on Forest Ecosystems. University of Jyvaskyla. Jyvaskyla, Finland. Biol. Res. Reports. Schweiger, S., and S. Boutin. 1995. The effects of winter food addition on the population dynamics of Clethrionomys rutilus. Can. J. Zool. 73:419-426. Stenseth, N . C. 1999. Population cycles in voles and lemmings: density dependence and phase dependence in a stochastic world. Oikos 87:427-461. Taitt, M . J., and C. J. Krebs. 1981. The effect of extra food on small rodent populations. 2. Voles (Microtus townsendii). J. Anim. Ecol. 50:125-137. Turkington, R., E. John, C.J. Krebs, M.R.T. Dale, V . Nams, R. Boonstra, S. Boutin, K. Martin, A.R.E. Sinclair, and J.N.M. Smith. 1998. The effects of N P K fertilization for nine years on boreal forest vegetation in northwestern Can. J. Veg. Sci. 9:333-346. Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of fertilization and herbivory on the herbaceous vegetation of the boreal forest in north-western Canada: a 10-year study. J. Ecol. 90:325-337. West, S. D. 1982. Dynamics of colonization and abundance in central Alaskan populations of the northern red-backed vole, Clethrionomys rutilus. J. Maram. 63:128-143. West, S. D. 1979. Habitat responses of microtine rodents to central Alaskan forest succession. Ph.D. Thesis. University of California, Berkeley, California.  54  Chapter 4: Conclusions and Recommendations for Further Work In this thesis I have tested parts of the fecal nutrient enrichment hypothesis which is one of the hypothesis put forth to explain why major peaks in the northern red-backed vole (Clethrionomys rutilus Pallas) populations coincide with the late decline or the low phase of the snowshoe hare (Lepus americanus) population cycle. The hypothesis argues that when L. americanus are at the peak of their 9 to 10 year cycle they produce large quantities of fecal pellets which will release nutrients, specifically nitrogen (N), when they decay. Nitrogen is captured by berry-producing plants, and providing weather conditions are suitable, the extra N enables the plants to produce enough berries to last through the winter during which time they provide a high-quality food source for C. rutilus. This overwintering food source causes an increase in spring C. rutilus numbers by reducing winter mortality. Testing the fecal nutrient enrichment hypothesis requires partitioning the hypothesis into smaller questions. What factors limit berry production? What are the effects of low N levels on berry production? Are the amounts of N found in snowshoe hare fecal pellets sufficient to promote berry reproduction in the following year? What is the relationship between berry crops this summer and red-backed vole densities the following summer? Chapters 1 to 3 addressed each of these questions. Nitrogen seems to be the main limiting resource of standing crop vegetation in boreal forests (McKendrick et al. 1980), and there are many studies in the boreal forest that have measured the effects of the addition of various levels of N (Chapin et al. 1986, Bonan and Shugart 1989, Nordin et al. 1998, Turkington et al. 2002, Penney et al 2003). Effects varied depending upon application rate with larger amounts of N increasing graminoid and herbaceous plant growth (Shaver et al. 1986, Hikosaka et al. 1998), promoting a higher degree of herbivory (Nams et al. 1996), and increasing fungal attacks (Graham and Turkington 2000). In contrast, lower application rates of N in the early spring either showed no or minimal effects on shrubs (Gerdol et al. 2000), or in some cases promoted shrub growth (Chapin and Shaver 1985) and berry production (Hansson and Retamales 1992, Penney et al. 2003), providing there were no unseasonable frosts that killed the vulnerable flower buds (Kuchko 1988, Selas 2000). Effects also varied depending on the timing of the application and the delivery system. Two applications of quick release N (3.8 g N / m each application) in the spring increased berry 2  production more than one equivalent application (7.6 g N/m ) of quick release nitrogen (Hanson 2  55  and Retamales 1992). Similarly, one application of slow-release N had equivalent results to the two applications (Hanson and Retamales 1992). This initial flush of N in the spring and then a slow release during the rest of the summer is equivalent to what occurs when melt water flushes accumulated N into the soils, and hare pellets slowly decompose and release N throughout the summer. Chapter 2 tested the effects of various levels of N on berry production, at rates above and below those levels produced by hare fecal pellets during peak hare population densities. Empetrum nigrum berry production increased at low levels of N . Empetrum nigrum along with Geocaulon lividum and Shepherdia canadensis also had heavier berries at these low levels although the latter two were not significant. Based on previous studies (Hanson and Retamales 1992, Penney et al. 2003) and on my two years of N addition experiments I tentatively conclude that low levels of N , similar to those produced by hare fecal pellets, can increase berry production. The next step is to determine if there is a relationship between berry production and C. rutilus densities. For this component of my study I used 9 years of berry production indices and C. rutilus population estimates taken from 1997 to 2005. These data show that berry production indices of three shrubs, E. nigrum, A. rubra, and S. canadensis together are accurate predictors of C. rutilus densities in the following year. Therefore, C. rutilus population densities can be predicted based on berry production indices from the previous year. Based on this evidence, I find some support for the fecal nutrient enrichment hypothesis which may explain why major peaks in the C. rutilus populations occur two to three years after major peaks of snowshoe hares. Several assumptions of this study require further testing. I have assumed that the nitrogen I added to the plots will have the same effects as N from hare pellets. I have also assumed that adding N in a uniform manner on the small mammal trapping grids will be equivalent to the overall distribution of N from hare pellets. Hare pellets tend to occur in a clumped distribution pattern reflecting hares habitat use preferences (e.g. under heavy cover, along hare trails) (Wolfe et al. 1982), and may accumulate in high densities in localized areas. In areas with dense accumulations of pellets, N release may be up to 2.0 g N / m but these 2  pellets are usually found in areas with a thick undercover story and little plant growth (Hodges 2000, Hik pers. comm.). High pellet densities can also be found around wind-felled trees and thus more likely near dwarf shrubs so these shrubs may obtain 2.0 g N / m from pellets. In all 2  56  likelihood though most areas probably receive lower levels of N around 0.4 to 1.0 g N/m (Hik 2  pers. comm.). In the N addition experiment berry production at 2.0 g N/m was not significantly 2  different than at 0.5 and l . O g N / m although there was a downward trend in berry production in some of the shrubs as N levels increased beyond 1.0 g N/m . The question then is "Will unevenly 2  distributed hare pellets have the same effect on plants as evenly distributed N ? " This question could be answered after the next hare peak, sometime around 2010. During this time pellets could be collected and frozen, and when hare populations are at their low point a few years later, the pellets could be distributed in a clumped distribution in small plots and berry production monitored. Hare pellet distribution patterns can be simulated using population monitoring studies that have mapped pellet group distribution, or distributions can be mapped during the years when hare populations are at their peak. Remapping the distribution would be preferable as other aspects such as the type of plants found in the vicinity of hare pellets can also . be recorded. For example, heterogeneity of pellets could be associated with the hares' preference for dense horizontal understory cover at 1 to 3 m above the ground (Wolff 1980, Wolfe et al. 1982, Hodges 2000), and S. canadensis shrubs would provide this type of cover especially when the branches are bent under the weight of snow. Another example is E. nigrum is usually found in moister areas with abundant white spruce {Picea glauca) whose lower branches provide shelter for the hares in deep snow. Pellets may then accumulate in these snow-covered areas and fertilize nearby patches of E. nigrum. A longer term assumption is that N-treated (1.0 and 2.0 g N/m ) small mammal plots of 2  2.8 ha are large enough to produce and detect significant differences in C. rutilus densities compared to control plots. Even if berry production on the experimental plots is increased, detectable changes in the vole population might not be observable. If detected changes in the C. rutilus population do occur, they may be due to increased overwintering survival rates, or increased immigration of C. rutilus to plots with more high quality food available. Some of these questions may be answered if we have ear-tagged, aged and sexed C. rutilus. In addition to continued live trapping of the treatment sites to determine if additional N will produce an increase in C. rutilus populations, several other items should be added to the study. Most crucial is determining the landscape abundance of the berry producing plants both outside and inside the live trapping grids. By establishing the landscape abundances of various plant species outside of the grids an estimated berry yield in kg/hectare/plant species could be  57  derived. Berry production indices do not measure actual berry abundance but instead are just a relative index that follows year to year changes in berry production. Therefore berry yield may be more useful than berry production indices in determining relationships between berries and C. rutilus in untreated regions. It is also important to estimate landscape abundance within the treatment grids. From the studies described in this thesis we know that added N may increase berry crops, and increased berry crops may increase C. rutilus densities the following year. While low levels of N can increase berry production of some plants on small 1 m plots this does not necessarily mean the 2  same effect will occur on the larger 2.8 ha live trapping grids. More plant heterogeneity in larger plots may result in different berry-production effects due to increased resource competition. As well, nutrients have the potential to flow long distances through the interconnecting root systems of plants on these plots (Gerdol et al. 2000) so the N dose per plant can vary considerably. The smaller 1 m plots were all trenched to prevent this. 2  To date berry production has not been measured within the 2.8 ha vole trapping grids themselves. If C. rutilus numbers are not significantly different on the treatment grids, this could be because berry production on the grids failed to increase despite N additions. Or it is possible that, while berry crops increased, the landscape abundance of the crucial shrubs was too low to make a difference. If trapping data from treatment grids do not support the fecal nutrient enrichment hypothesis, it is important to determine why as a basis for a decision on whether to completely reject the hypothesis or not. Therefore, plant sampling transects for landscape abundance and berry monitoring plots for berry production should be established within the grids. The berry monitoring plots may help strengthen any correlations between N addition and C. rutilus population growth, and the landscape abundance transects may help determine which, if any, of the shrubs play a more crucial role in C. rutilus fluctuations. Another item to consider in the on-going study would be some winter trapping and eartagging of C. rutilus to determine how winter survival rates of C. rutilus compare to berry production indices, as well as to determine how survival rates differ on treated vs. untreated grids. Future work may include a nutrient pulse experiment similar to the water pulsing experiments done by Novoplansky and Goldberg (2001) as the timing of N addition to the soil  58  makes a difference (Hanson and Retamales 1992). Hare pellets will add a flush of N in the spring snow melt, but also release smaller amounts as they decay. This could be simulated by adding a relatively larger amount of N in the spring, and then adding smaller amounts at selected intervals over the summer, or a combination of fast and slow-release fertilizer in the spring. Finally, with more time I would propose to duplicate the small shrub plot thesis experiment during times when weather factors from one year to the other were more similar to each other. The wet cooler weather in summer 2005 was quite different from the drought weather in 2004. In fact, many flower buds were killed in 2005 when a late snowfall buried the opened flowers under approximately 8 cm of snow. Two seasons of similar weather without high flower mortality may have increased the number of significant results seen as there would have been more berries produced in 2005, and differences between treatments may have been more evident. Overall, weather factors in the second year resulted in reduced berry production, and only tentative evidence from this experiment shows berry production increases at N levels as may be found during times of high hare densities. A stronger correlation between berry production and C. rutilus densities over a 9 year period exists. Therefore, the fecal nutrient enrichment hypothesis is mostly supported by this correlation, but still awaits more evidence linking N levels to berry production.  59  4.1 Literature Cited Bonan, G. B., and H. H . Shugart. 1989. Environmental factors and ecological processes in boreal forests. Annu. Rev. Ecol. Syst. 20:1-28. Chapin, III, F.S., and G. R. Shaver. 1985. Individualistic growth response of tundra plant species to environmental manipulations in the field. Ecol. 66:564-576. Chapin, III, F.S.,P.M. Vitousek, and K. Van Cleve. 1986. The nature of nutrient limitation in plant communities. Amer. Nat. 127:48-58. Gerdol, R., L. Brancaleoni, M . Menghini, and R. Marchesini. 2000. Response of dwarf shrubs to neighbour removal and nutrient addition and their influence on community structure in a subalpine heath. J. Ecol. 88:256-266. Graham, S. A., and R. Turkington. 2000. Population dynamics response of Lupinus arcticus to fertilization, clipping, and neighbour removal in the understory of the boreal forest. Can. J. Bot. 78:753-758. Hansson, E. J., and J. B . Retamales. 1992. Effect of nitrogen source and timing on highbush blueberry performance. Hort. Sci. 27:1265-1267. Hikosaka, K., Y. T. Hanba, T. Hirose, and I. Terashima. 1998. Photosynthetic nitrogen-use efficiency in leaves of woody and herbaceous species. Func. Ecol. 12:896-605. Hodges, K . E . 2000. The ecology of snowshoe hares in northern boreal forests. Pages 117-161 in L.F. Ruggiero, K . B . Aubry, S.W. Buskirk, G . M . Koehler, C.J. Krebs, K.S. McKelvey, J.R. Squires, editors. Ecology and conservation of lynx in the United States. University Press of Colorado and United States Department of Agriculture, Rocky Mountain Research Statioin, Denver, Colorado. Kuchko, A. A. 1988. Bilberry and cowberry yields and the factors controlling them in the forests of Karelia, USSR. In Proceedings of the Finnish-Soviet symposium on timber forest resources in Jyvaskyla, Finland. 25-29 August 1986. Edited by I. Vanninen, and M . Raatikainen. Acta. Bot. Fennica 136:23-25. McKendrick, J. D., G. O. Batzli, K. R. Everett, and J. C. Swanson. 1980. Some effects of mammalian herbivores and fertilization on tundra soils and vegetation. Arc.Alp. Res. 12:565-578. i  Nams, V.O., N.F.. Folkard, and J.N.M. Smith. 1993. Effects of nitrogen fertilization on several woody and nonwoody boreal forest species. Can. J. Bot. 71:93-97. Nordin, A., T. Nasholm, and L. Ericson. 1998. Effects of simulated N deposition on understorey vegetation of a boreal coniferous forest. Func. Ecol. 12:691-699.  60  Novoplansky, A., and Goldberg, D.E. 2001. Effects of water pulsing on individual performance and competitive hierarchies in plants. J. Veg. Sci. 12(2): 199-208. Penney, B.G., K . B . McRae, and G.A. Bishop. 2003. Second-crop N fertilization improves lowbush blueberry (Vaccinium angustifolium Ait.) production. Can. J. Plant Sci. 83:149155. Selas, V. 2000. Seed production of a masting dwarf shrub, Vaccinium myrtillus, in relation to previous reproduction and weather. Can. J. Bot. 78:423-429. ..Shaver, G. R., I. Chapin, F.S., and B. L . Gartner. 1986. Factors limiting seasonal growth and peak biomass accumulation in Eriophorum vaginatum in Alaskan tussock tundra. J. Ecol.74:257-278. Turkington, R., E. John, S. Watson, and P. Seccombe-Hett. 2002. The effects of fertilization and herbivory on the herbaceous vegetation of the boreal forest in north-western Canada: a 10-year study. J. Ecol. 90:325-337. Wolfe, M . L . N . V . Debyle, C.S. Winchell, and T.R. McCabe. 1982. Snowshoe hare cover relationships in northern Utah. J. Wildl. Manage. 46:662-670. Wolff, J.O. 1980. The role of habitat patchiness in the population dynamics of snowshoe hares. Ecol. Monogr. 50: 111-130.  /  61  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092514/manifest

Comment

Related Items