Many models used to estimate nutritional carrying capacity (NCC) for ungulates differ structurally, but the implications of those differences are frequently unclear. We present a comparative analysis of NCC estimates for a large herbivore in a dynamic landscape, using models that differ in structure and scope. We compared three model structures across three estimates of winter ranges under three winter-severity scenarios for an isolated, introduced moose Alces alces population on the Copper River Delta of south-central Alaska. Model estimates of NCC ranged from 205 to 4,592 moose, demonstrating the critical influences of model structure and assumptions when applying NCC. Furthermore, population estimates during recent severe winters suggest that past models underestimated NCC on the Copper River Delta. We conducted a sensitivity analysis of a preferred model and determined that model components with the highest and lowest sensitivity were snow depth and lignin- and tannin-caused reductions in forage nutritional quality, respectively. Our low sensitivity values for lignin and tannin influences on NCC contrast with results in other NCC estimates. Overall, our results reinforce the need for, and will hopefully assist, adaptive management in response to landscape, population, behavioral, and climatic changes on the Copper River Delta, and demonstrate the importance of understanding model assumptions and structure in application of NCC estimates in the management of large herbivores in variable ecosystems.

Nutritional carrying capacity (NCC) is the largest stable population of a specified organism supported by the total available nutritional energy, assuming a stable ecosystem within a defined range and time (Caughley 1979; Monteith et al. 2014). Thus, it acts as an idealized baseline to which further factors may be added to refine population management goals. In highly variable environments, the relationship between a population of animals and its food resource may fluctuate, undermining an equilibrium (Caughley et al. 1987; McLeod 1997). Therefore, application of NCC requires close consideration of the ecology of the organism and its environment. For ungulates, NCC depends on the species' foraging behavior relative to the ecology of plant species being consumed (Jefferies et al. 1994; Hobbs et al. 2003; Windels and Hewitt 2011). Within temperate and arctic ecozones, decreases in nutritional quality and availability of browse in winter increase the necessary digestion and foraging time relative to other seasons (Schwartz and Renecker 1997), and thus historically winter was considered the limiting period for temperate ungulates. However, temperate ungulates compensate during winter by lowering metabolic rates, decreasing activity periods, and metabolizing fat reserves (McEwan and Whitehead 1970; Mautz 1978b; Regelin et al. 1985; Renecker and Hudson 1986), while using spring and summer months to restore body fat, preparing for the energetically draining behaviors of lactation, rut, and pregnancy (Schwartz and Renecker 1997). Considerable research now suggests that summer forage quality is more influential to NCC for most temperate ungulate populations (Bobek 1977; Merrill and Boyce 1991; Cook et al. 2001, 2004; Beck et al. 2006; Dungan et al. 2010), and determining which season is more significant to a specific population depends on relationships between the animal's nutritional requirements and seasonal quality and accessibility of forage resources.

Beyond consideration of seasonal factors, model structures used over time to calculate NCC have used different definitions of forage availability. Early models often assumed complete forage availability, equal selection probability, or complete forage consumption (Mautz 1978a). However, later models recognized direct (e.g., snow burial, secondary chemicals) or indirect (e.g., predation risk) constraints on forage availability (Beck et al. 2006; Windels and Hewitt 2011) and the complexity of diet selection (Renecker and Schwartz 1997), including browsing tolerance of forage (Van Gils et al. 2004). The implications of these different structures are often unclear and complicate the application of NCC estimates as tools to guide management.

Methods for calculating forage available to ungulates vary, but initial “forage supply–animal demand” structures were very simple, assuming equal selection probability of all available forage species (Mautz 1978a). Ungulates, however, preferentially select certain forages over others because of their higher nutritional quality (e.g., crude protein, gross energy) and lower amounts of secondary chemicals (e.g., tannins), and to ensure the proper nutrient ratios (i.e., “gut mixing”) in their diet (Bryant et al. 1992; Windels and Hewitt 2011). Thus, these simple models likely overestimate the forage available and NCC estimates. With increased knowledge of browsing behavior, space use, and forage tolerance to herbivory, estimates of forage biomass within NCC model structures were later constrained by ungulate choice, dietary needs, or ability of forage to recover from consumption. Such models account for ungulate nutritional needs by calculating forage availability relative to its ability to meet ungulate dietary needs or by limiting consumption to highly preferred forages (solely, or in conjunction with, other lesser preferred forages), which are also most vulnerable to overbrowsing (Hobbs and Swift 1985; Hanley and Rogers 1989; Stephenson 1995). However, these structures inform or depend on accurate dietary analyses, which can vary seasonally, regionally, and by methods used (Holechek et al. 1982; Alipayo et al. 1992; Renecker and Schwartz 1997). Furthermore, Hobbs and Hanley (1990) cautioned against assuming ungulate diets are temporally static or inflexible. Thus, more complex models can both misrepresent ungulate diets or underestimate diet flexibility under changing environmental conditions or levels of intraspecific competition. Just as ungulate diets are context-specific, forage tolerance to herbivory is species-, seasonally-, regionally-, and temporally-specific while also influenced by the intensity of herbivory, the length of browsing periods, and soil nutrient loads (Danell et al. 1994; Augustine and McNaughton 1998; Persson et al. 2007). The implications of these different structures are often unclear, and complicate the use of NCC by managers.

An isolated moose Alces alces population on the Copper River Delta (CRD; Delta) in south-central Alaska (Figure 1), established by translocations starting in 1949 in an area previously unavailable to moose, provides a simplified single-ungulate system in which to evaluate NCC models. Estimates of NCC with varying definitions of available biomass have been used to set population management goals for moose on the CRD in the face of landscape change and strong seasonal fluctuations in resource availability (MacCracken et al. 1997; Stephenson et al. 2006), while recent changes in population size and distribution complicate the applicability of past estimates. As such, this research area serves as a case study in demonstrating the need to gauge the impact of modeling decisions and assumptions on NCC estimates.

Figure 1.

Map of the west Copper River Delta (CRD) in south-central Alaska, including available roads and sampling plots used (2012–2013) to estimate winter available forage for moose Alces alces. Three alternate moose wintering ranges used in estimation of current moose nutritional carrying capacity (NCC) are illustrated. These include the whole west Delta and both the primary and secondary regions of the total past observed range as estimated by MacCracken et al. (1997) using data from collared animals during 1987–1989. Dominant stand types (color coded) were determined by satellite mapping (Satellite Pour l'Observation de la Terre, version 5; 2011. Redcastle Resources, Inc., Salt Lake City, UT) and were used to estimate stand areas with available moose forage.

Figure 1.

Map of the west Copper River Delta (CRD) in south-central Alaska, including available roads and sampling plots used (2012–2013) to estimate winter available forage for moose Alces alces. Three alternate moose wintering ranges used in estimation of current moose nutritional carrying capacity (NCC) are illustrated. These include the whole west Delta and both the primary and secondary regions of the total past observed range as estimated by MacCracken et al. (1997) using data from collared animals during 1987–1989. Dominant stand types (color coded) were determined by satellite mapping (Satellite Pour l'Observation de la Terre, version 5; 2011. Redcastle Resources, Inc., Salt Lake City, UT) and were used to estimate stand areas with available moose forage.

Close modal

Highly variable winters and changes in moose winter ranges since the establishment of the new moose population further complicate estimation of NCC on the CRD. From 1917 to 2012, winter length (total days from first snow until last) ranged from 65 to 221 d, with annual snow depths averaging 2.7–108.0 cm (WRCC 2015). Thus, effects of winter severity on forage availability (snow burial) vary by year (Visscher et al. 2006). Furthermore, past research that included NCC estimates (1987–1989; MacCracken et al. 1997) found 90% of collared moose wintered within primary and secondary ranges of 4,800–12,900 ha (Figure 1), presumably because of forage availability, snow drifting, and extreme winter winds (−40°C wind chill) from the Copper River Canyon. Although detailed space-use data are unavailable for the current population, recent (2012–2014) aerial surveys suggest the moose population has expanded its winter range beyond those described in 1987–1989 to encompass most of the west Delta. Nevertheless, these same surveys indicate the highest moose densities may still occur within the 1987–1989 primary range (C. Westing, Alaska Department of Fish and Game, personal communication). Whether this range expansion is due to exploration from the release points, population growth, or changes in vegetation resulting from climate- or earthquake-driven succession is unknown, but the increased forage availability could significantly affect NCC estimates.

Here, we evaluate different approaches to estimating winter NCC in order to guide management of a large herbivore on a dynamic landscape. We developed three model structures of energy-based winter NCC with varying levels of forage availability and plant tolerance to herbivory. To maintain historical precedent set by previous studies (MacCracken et al. 1997; Stephenson et al. 2006), we assumed that winter was the limiting season and compared model estimates of NCC across different winter scenarios (mild, moderate, and severe) and winter range sizes (12,995–54,735 ha). Our objectives were to 1) calculate the digestible energy available to moose in different stand types and develop a set of energy-based NCC models, varying in structure and assumptions, in order to represent the diverse factors that managers on the CRD may consider relevant to the moose population with their current knowledge of CRD moose ecology; 2) compare estimates from these models to demonstrate the differences resulting from NCC model structures and assumptions; 3) perform a sensitivity analysis to determine the relative influence of factors within one model, range, and winter scenario within that combination.

The CRD is the largest continuous wetland (>283,000 ha) in the Pacific Northwest (Kesti et al. 2007). Presumably excluded from the Delta by topography in the past, moose were introduced (n = 23; 1949–1958) as a subsistence resource for the town of Cordova (MacCracken et al. 1997). The population has since split into two semidistinct populations (west and east; 600 and 230 individuals, respectively) separated by the Copper River (C. Westing, Alaska Department of Fish and Game, unpublished data; Figure 2; Text S1, Data S1, Supplemental Material). We address the western population, which is better understood and managed. Hunting began in 1960 and is the greatest source of mortality in adult moose, although heavily harvested wolf Canis lupus and brown bear Ursus arctos alascensis populations contribute to adult and calf mortality (MacCracken et al. 1997). The population supports a relatively high twinning rate (58–65%; C. Westing, Alaska Department of Fish and Game, unpublished data).

Figure 2.

Single (dot) and loess moving average (line) estimates of the moose Alces alces population abundance on the west Copper River Delta, Alaska, since introductions during 1949–1959 (light grey). Moose counts (1963–2014) were conducted via winter aerial surveys and annual totals were estimated using minimum counts (black) or Gasaway surveys (Gasaway et al. 1986).

Figure 2.

Single (dot) and loess moving average (line) estimates of the moose Alces alces population abundance on the west Copper River Delta, Alaska, since introductions during 1949–1959 (light grey). Moose counts (1963–2014) were conducted via winter aerial surveys and annual totals were estimated using minimum counts (black) or Gasaway surveys (Gasaway et al. 1986).

Close modal

The west Delta (∼55,000 ha, 0–300 m elevation) is encircled by the Pacific Ocean, glaciers, the Chugach Mountain Range (>2,000 m elevation), and the Copper River (Figure 1), which channels winter gusts up to 193 km/h through its canyon. Soil types include aeolian and alluvial sands, gravels, and marine glacial silts (Davidson and Harnish 1978). Mean annual (1909–2014) precipitation, winter snow depth, and temperature were 230 cm, 23 cm, and 4.0°C (monthly range −8.9–16.3°C; WRCC 2015). The CRD has 336 annual growing-degree days (base 50°F or 10°C) on average. Hydrological dynamics and channel changes determine vegetation distribution on the Delta (MacCracken 1992; Boggs 2000). Woody browse important to wintering moose include feltleaf, Barclay's, undergreen, Hooker's, and Sitka willow Salix alaxensis, S. barclayi, S. commutata, S. hookeriana, and S. sitchensis, as well as Sitka alder Alnus viridis sinuata, black cottonwood Populus trichocarpa, and sweetgale Myrica gale (MacCracken et al. 1997). These eight forage species are found in seven stand types classified via a map generated by Satellite Pour l'Observation de la Terre, version 5 (SPOT5; 2011, Redcastle Resources, Inc., Salt Lake City, UT): spruce–hemlock Picea–Tsuga, spruce–cottonwood Picea–Populus, cottonwood, alder Alnus, alder–willow AlnusSalix, willow, and sweetgale Myrica. A 1964 earthquake (9.2 magnitude) uplifted the area by 1.0–4.0 m, initiating changes in hydrology, soil salinity, and vegetation while generally increasing the proportion of stands dominated by woody species, especially Sitka alder, sweetgale, and Sitka spruce P.sitchensis (Ferrians 1966; Stover and Coffman 1993; Boggs 2000; Thilenius 2008).

Modeling nutritional carrying capacity

Calculating digestible energy and development of NCC model structures

To evaluate the effects of model type, winter severity, and winter range on NCC estimates, we developed three digestible energy-based NCC models (Figure 3) using the “forage supply–animal demand” structure (Mautz 1978b; Hobbs et al. 1982). All three models incorporated species- and stand-type-specific estimates of nutritional quality and available biomass (kg/ha) to calculate the digestible energy (kcal/ha) available within the seven stand types. Nutritional quality estimates incorporated effects of secondary chemicals (reduction of available protein due to tannins, and reduction of digestible dry matter due to lignin; Windels and Hewitt 2011), gross energy, and neutral detergent fiber. We adjusted estimates of available biomass within stand types according to the mean snow depths (m) in winter scenarios. We averaged species-specific values of available biomass and nutrition, but used a Monte Carlo simulation to bootstrap stand-type-specific digestible energy values and to provide estimates of within-stand-type variability. We then multiplied the within-stand variability of digestible energies by range-specific estimates of stand-type areas (ha) to estimate median stand-type digestible energies. We summed the total median stand digestible energies, and compared the total calculated winter scenario- and range-specific digestible energy with the total energy required by an average adult female moose (500 kg) for maintenance (Schwartz et al. 1988) across a scenario-specific winter length (days) to determine the total number of moose supported within a specified range (see Figure 3 footnotes for maintenance requirements, all NCC formulas, and associated references).

Figure 3.

General model structure used to estimate nutritional carrying capacity of the Copper River Delta, Alaska, for moose Alces alces wintering in 2012–2013. Model structure was applied across model types (open-diet 25% use, diet-restricted 100% use, and diet-restricted 25% use), winter ranges (total past observed including primary and secondary ranges, whole west Delta, adjusted west Delta), and winter scenarios (mild, moderate, severe).

Footnotes:

1 PR = protein reduction (g of protein lost/100 g protein); Spalinger et al. (2010).

2 BSA = bovine serum albumin (mg/g DM); Spalinger et al. (2010).

3 LC = lignin-cutin concentration (g/100 g NDF); Spalinger et al. (2010).

4 DM = dry matter estimated from regression equations (g/stem); MacCracken et al. (1997), Stephenson et al. (2006), and Smythe (2015).

5 NDF = neutral detergent fiber (% of DM).

6 NDS = neutral detergent solubles (g/100 g DM); Spalinger et al. (2010).

7 DDM = digestible dry matter (g/100 g DM); Spalinger et al. (2010).

8 SN = snow effects (g DM/stem).

9 biomass0.53.0m = biomass within moose browsing reach without effects of snow depth (g DM/stem).

10 %biomass = percent of total stem biomass newly reached or buried by snow.

11 AB = available biomass (g DM/stem).

12 AU = adjusted use (%) according to model type (25% of all forage species in open-diet; 0% for nonpreferred species and 25 or 100% for willow species in diet-restricted).

13 DE = digestible energy (%); Schwartz and Renecker (1997).

14 DE = digestible energy (cal/g); Schwartz and Renecker (1997).

15 GE = gross energy (cal/g).

16 DE = digestible energy (kcal/ha).

17 SD = stem density (N/ha).

18 Browse species for wintering moose on the Copper River Delta.

19 Area (ha) as determined by SPOT5 (2011) satellite vegetation map.

20 Stand types relevant to wintering moose ecology on the Copper River Delta.

21 Either full west Delta range, road-bias-adjusted west Delta range, plot inference range, or the past observed ranges (MacCracken et al. 1997).

22 Resting digestible energy metabolic need: 148.6 + 20% due to activity (kcal/kg BW^0.75/day), where BW = body weight; Schwartz et al. (1988) and Parker et al. (1984).

23 Weight (kg) of average adult female moose; Miquelle et al. (1992).

24 WL = winter length (days) adjusted according to winter scenario (mild, moderate, or severe).

Figure 3.

General model structure used to estimate nutritional carrying capacity of the Copper River Delta, Alaska, for moose Alces alces wintering in 2012–2013. Model structure was applied across model types (open-diet 25% use, diet-restricted 100% use, and diet-restricted 25% use), winter ranges (total past observed including primary and secondary ranges, whole west Delta, adjusted west Delta), and winter scenarios (mild, moderate, severe).

Footnotes:

1 PR = protein reduction (g of protein lost/100 g protein); Spalinger et al. (2010).

2 BSA = bovine serum albumin (mg/g DM); Spalinger et al. (2010).

3 LC = lignin-cutin concentration (g/100 g NDF); Spalinger et al. (2010).

4 DM = dry matter estimated from regression equations (g/stem); MacCracken et al. (1997), Stephenson et al. (2006), and Smythe (2015).

5 NDF = neutral detergent fiber (% of DM).

6 NDS = neutral detergent solubles (g/100 g DM); Spalinger et al. (2010).

7 DDM = digestible dry matter (g/100 g DM); Spalinger et al. (2010).

8 SN = snow effects (g DM/stem).

9 biomass0.53.0m = biomass within moose browsing reach without effects of snow depth (g DM/stem).

10 %biomass = percent of total stem biomass newly reached or buried by snow.

11 AB = available biomass (g DM/stem).

12 AU = adjusted use (%) according to model type (25% of all forage species in open-diet; 0% for nonpreferred species and 25 or 100% for willow species in diet-restricted).

13 DE = digestible energy (%); Schwartz and Renecker (1997).

14 DE = digestible energy (cal/g); Schwartz and Renecker (1997).

15 GE = gross energy (cal/g).

16 DE = digestible energy (kcal/ha).

17 SD = stem density (N/ha).

18 Browse species for wintering moose on the Copper River Delta.

19 Area (ha) as determined by SPOT5 (2011) satellite vegetation map.

20 Stand types relevant to wintering moose ecology on the Copper River Delta.

21 Either full west Delta range, road-bias-adjusted west Delta range, plot inference range, or the past observed ranges (MacCracken et al. 1997).

22 Resting digestible energy metabolic need: 148.6 + 20% due to activity (kcal/kg BW^0.75/day), where BW = body weight; Schwartz et al. (1988) and Parker et al. (1984).

23 Weight (kg) of average adult female moose; Miquelle et al. (1992).

24 WL = winter length (days) adjusted according to winter scenario (mild, moderate, or severe).

Close modal

The three models differed in their approach to available biomass.

  1. In our “open-diet” model (a historically relevant standard, and digestible energy-based modification of the model used by MacCracken et al. 1997), we assumed equal selection probability for all forage species (i.e., did not consider forages “preferred” or “nonpreferred”). Because complete (100%) consumption of all available forage is ecologically, physiologically, and managerially unrealistic (Miquelle and Jordan 1979; Renecker and Schwartz 1997) but the tolerance level for consumption of shrubs on the CRD are unknown, we conservatively reduced the available biomass of forage species to 25% (i.e., 25% use) to simulate sustainable browsing of forage species (Singer and Zeigenfuss 2002; Dungan et al. 2010).

  2. Our “diet-restricted” model reduces available biomass to the most preferred (and at-risk for overconsumption) forage as determined by proportional occurrence in diet analyses (Bryant et al. 1992; Windels and Hewitt 2011). Moose fecal analyses conducted by MacCracken et al. (1997) determined willows were most abundant in the CRD moose diets, although their analyses could not differentiate among species. Thus, we used the total available biomass of willows as the limiting forage.

  3. Lastly, because the “diet-restricted” model assumes 100% availability of available willow biomass, we developed a third model, which conservatively limited willow biomass to 25% (“diet-restricted, 25% use”).

Development of winter scenarios

We developed winter scenarios (mild, moderate, and severe) by categorizing winters based on mean winter snow depth (cm) and length (days from first snowfall to last) collected from 1917 to 2012 by the Alaska Climate Research Center (ACRC 2014) weather station at Cordova's “Mudhole Smith” Airport on the west Delta. We determined categories by plotting winter total snow depths and noting natural breaks. We then calculated winter lengths by averaging the number of days from first snowfall to last for each winter in each category. We used analysis of variance to compare scenarios, and then adjusted scenario NCC model components of available biomass and winter length by mean snow depths and numbers of days.

Delineation of winter ranges and estimation of stand type areas

We initially defined two ranges relevant to the wintering moose population (Figure 1): 1) an aggregate of the primary and secondary ranges (hereafter, “total past observed range”) as determined using previously Very High Frequency–collared moose (n ≈ 200; 1987–1989; MacCracken et al. 1997), and 2) the entire west Delta, to incorporate evidence from recent aerial surveys showing range expansion of the population (C. Westing, Alaska Department of Fish and Game, unpublished data). Our vegetation sampling plots were subject to logistical and safety restrictions and fell ≤1.5 km from CRD roads, which function as levees; therefore, we investigated effects of plot distance from roads (m) on nutritional energy using linear regression models to account for potential hydrological effects of roads on vegetation. We determined which stand types differed significantly (regression P-value ≤ 0.05) in nutritional energy with increasing distance from roads, retested to determine at what distance the effect became nonsignificant, and then developed a third range definition (hereafter, “adjusted west delta range”), which re-estimated nutritional energy in affected stands beyond the significant distance. We then recalculated the total nutritional energy across the Delta by summing separate estimates of nutritional energy beyond the significant distance for each affected stand type with estimates of nutritional energy within the significant distance, as well as total nutritional energy regardless of road distance for unaffected stands (Figure S1, Text S2, Data S2, Supplemental Material). This statistical adjustment to our forage biomass estimates relative to the levee-like road thus provides the least-biased correction for our forage estimates created by the logistical restrictions placed on our sampling size and design. To estimate stand type areas (ha) within winter ranges, we used the SPOT5 (2011) vegetation map, which determined the vegetation stand-type categories we used, and Graphical Information System (GIS)–based tools to sum the total area of each stand type. We corrected stand area estimates of the SPOT5 map according to an error matrix developed by Red Castle Resources, Inc. (Salt Lake City, UT) via “spot-proofing” (randomly visiting points within the study area to verify the accuracy of satellite classifications), and by visually checking stand categorization accuracy at our randomized points across the Delta.

Estimation of biomass availability and nutritional quality

To estimate the biomass available to and utilized by moose, we used GIS-based programs and the SPOT5 map to generate 7–9 randomized points within each of the 7 stand types (53 total). We restricted points to ≤1.5 km of roads because of logistical and safety constraints. Upon reaching each point, we verified the SPOT5 classification of the stand type (Viereck 1992) and established three 10 × 1-m belt transects. Two belt transects ran north and the third ran west, with start points ≥5 m apart. To capture maximum forage growth, we collected autumn (September–October 2012, 2013) basal diameters, stem densities, and stem heights, and to sample browse nutrition prior to green-up, we collected late-winter (April–May 2013) samples of the eight browse species.

In the autumn, we measured basal diameters (0.01 mm) above the moss layer on the three closest stems every 0.5 m along the transects, and then estimated available biomass (g/stem) with regression equations (MacCracken and Van Ballenberghe 1993; Stephenson et al. 1998; Smythe 2015). Very-large-stem basal diameters (>60.0 mm) increase heteroscedasticity of biomass estimates (MacCracken and Van Ballenberghe 1993). Thus, for large stems we measured a branch diameter, estimated the number of equivalent branches on the stem, and summed the branch biomasses. We counted the number of stems and calculated stem density (n/30 m2) of each forage species, measured the height (m) of three stems per species (when available) per transect and, to allow for changes in available biomass with increasing snow depth, we visually estimated the biomass (%) present on those stems in 1-m increments from 0 to 6 m (the height range of browse available to moose in mild to severe winters). We calculated the species-specific mean available biomass (stem biomass × stem density, kg/ha, and % available by incremental height classes), and height (m) within each stand type.

Finally, to estimate forage nutritional quality, we collected samples of available biomass for every browse species observed in each plot. The maximum bite diameter recorded on the Delta was 8.3 mm (MacCracken et al. 1997), so we clipped all species to that diameter, placed samples in paper bags, defoliated them, and stored them fresh-frozen from the field before sending them to Washington State University's Wildlife Habitat and Nutrition Lab where they were oven-dried and ground prior to analyses of gross energy (cal/g; bomb calorimetry), crude protein (%), lignin (%; 72% sulfuric acid method), tannin levels (mg/g, Martin and Martin 1982), and neutral detergent fiber (%, Van Soest/Ankom sequential fibers). From these values, we calculated the mean species-specific nutritional values within each stand type.

Sensitivity analysis

We conducted a sensitivity analysis on one model, range, and winter scenario combination to determine the relative influence of factors within the representative combination. On the “open-diet, 25% use” model in the road-bias-adjusted range during a severe winter, we increased nine factors (protein reduction due to tannins, lignin content, gross energy, stem biomass, stem density, snow depth, stand type area, winter length, energy needed by wintering moose, and moose mass) by 5% and calculated the change (%) in NCC and the proportional sensitivity.

Although we encountered feltleaf willow near the Scott Glacier and Copper River (S.E. Smythe, personal observation, 2013), it never occurred within our randomized plots and we subsequently removed it from our analyses. The seven remaining forage species in the seven stand types varied in mean available biomass (kg/ha), gross energy (cal/g), crude protein (%), lignin (%), tannin (mg/g), and neutral detergent fiber (%) levels (Tables S1 and S2, Text S3, Data S2, Supplemental Material). Winter length in mild, moderate, and severe winter scenarios averaged 147 (±7), 163 (±8), and 178 (±22) d, while annual mean snow depths averaged 11 (±2), 26 (±2), and 64 (±17) cm, respectively. Between 1917 and 2012, mild, moderate, and severe winter scenarios occurred 49, 29, and 11 times, while 6 winters were unclassified on account of incomplete data (Text S4, Data S3, Supplemental Material). Winter ranges varied in size (12,995–54,735 ha) and stand type composition (Table 1, Text S5).

Table 1.

Summary of total and stand-type-specific areas (ha) of winter ranges used to calculate the nutritional carrying capacity of moose Alces alces on the Copper River Delta, Alaska, in 2012–2013.

Summary of total and stand-type-specific areas (ha) of winter ranges used to calculate the nutritional carrying capacity of moose Alces alces on the Copper River Delta, Alaska, in 2012–2013.
Summary of total and stand-type-specific areas (ha) of winter ranges used to calculate the nutritional carrying capacity of moose Alces alces on the Copper River Delta, Alaska, in 2012–2013.

Digestible energy available to moose varied by stand type and winter scenario (Figure 4, Text S6). Of the seven stand types, only digestible energy estimates in cottonwood, alder–willow, willow, and sweetgale stands were significantly affected by plot distance (m) from the Delta roads (linear regression: P ≤ 0.05; Table S3, Text S2, Data S2, Supplemental Material). The effect disappeared in plots >100 m from roads in cottonwood stands, and >200 m from roads in alder–willow, willow, and sweetgale stands.

Figure 4.

Digestible energy (kcal/ha) of moose Alces alces winter forage within dominant stand types on the Copper River Delta, Alaska, as estimated across three winter scenarios (mild, moderate, and severe) from samples collected in 2012–2013.

Figure 4.

Digestible energy (kcal/ha) of moose Alces alces winter forage within dominant stand types on the Copper River Delta, Alaska, as estimated across three winter scenarios (mild, moderate, and severe) from samples collected in 2012–2013.

Close modal

Estimates of NCC across all models, ranges, and scenarios ranged from 205 to 4,592 moose (1.0–12.5 moose/km2; Figure 5, Text S7, Data S2, Supplemental Material). Mean estimates of NCC in the three ranges (total past observed, whole west Delta, and road-bias-adjusted west Delta) across all winter scenarios and models were 849, 2,359, and 2,407 total moose, respectively. Thus, NCC estimates using the adjusted west Delta range were 2.0% higher than the west Delta range, while estimates using the total past observed range were 64.0% lower. However, our models estimated the highest moose densities (moose/km2) within the total past observed range (Figure 5; Text S7, Data S2, Supplemental Material). Mean NCC estimates using “diet-restricted (100% use)” models across all ranges and winter scenarios were 20.1% higher, respectively, than estimates using the “open-diet” model, though estimates declined by 75.0% when we limited available biomass (“diet-restricted, 25% use”). Mean NCC estimates across all models decreased by 18.9% and 32.5% as winter severity increased to moderate and severe, respectively. Percent changes in NCC from the sensitivity analysis on the “open-diet, 25% use” model in the road-biased-adjusted winter range model during a severe winter varied from −4.76 to 11.18% (Table 2; Text S8, Data S2, Supplemental Material). Proportional sensitivity ranged from −0.95 to 2.24, with the smallest and largest proportional sensitivity attributed to protein reduction (due to tannins) and snow depth, respectively.

Figure 5.

Estimates of nutritional carrying capacity (NCC; no. of moose Alces alces predicted, with SE bars) on the Copper River Delta, Alaska, in 2012–2013 using three models (“open-diet/25% use,” “diet-restricted/100% use,” and “diet-restricted/25% use”), across three ranges (total past observed including primary and secondary ranges, whole west Delta, adjusted west Delta), and under three winter scenarios (mild, moderate, and severe). The “open-diet” model allowed equal use of all seven winter forage species, while the “diet-restricted” models limited forage biomass to preferred willows. Percent use restricted biomass availability to allow forage recovery (25% use) or complete consumption (100% use). Winter range areas (ha) used in modeling ranged in size from largest, consisting of the entire study area (whole west Delta), to smallest being a range determined by space use by moose tracked in an earlier study (total past observed; MacCracken et al. 1997), with primary and secondary subranges being used in both mild and moderate winter severities, but only the secondary subrange being used in severe winters. Intermediate in area was the west Delta with estimates of available forage biomass adjusted relative to road levee effects (adjusted west Delta). Numerical estimates of NCC (total abundance and moose/km2) are summarized in Table S4 (Supplemental Material) by model, range, and winter scenario.

Figure 5.

Estimates of nutritional carrying capacity (NCC; no. of moose Alces alces predicted, with SE bars) on the Copper River Delta, Alaska, in 2012–2013 using three models (“open-diet/25% use,” “diet-restricted/100% use,” and “diet-restricted/25% use”), across three ranges (total past observed including primary and secondary ranges, whole west Delta, adjusted west Delta), and under three winter scenarios (mild, moderate, and severe). The “open-diet” model allowed equal use of all seven winter forage species, while the “diet-restricted” models limited forage biomass to preferred willows. Percent use restricted biomass availability to allow forage recovery (25% use) or complete consumption (100% use). Winter range areas (ha) used in modeling ranged in size from largest, consisting of the entire study area (whole west Delta), to smallest being a range determined by space use by moose tracked in an earlier study (total past observed; MacCracken et al. 1997), with primary and secondary subranges being used in both mild and moderate winter severities, but only the secondary subrange being used in severe winters. Intermediate in area was the west Delta with estimates of available forage biomass adjusted relative to road levee effects (adjusted west Delta). Numerical estimates of NCC (total abundance and moose/km2) are summarized in Table S4 (Supplemental Material) by model, range, and winter scenario.

Close modal
Table 2.

Changes (%) and proportional sensitivities (Sx) of the independent components in an open-diet (25% use of all forage species) model used to estimate nutritional carrying capacity (NCC) for moose within the road-bias-adjusted west Delta winter range in a severe winter on the Copper River Delta, Alaska. Samples used within the model were collected in 2012–2013.

Changes (%) and proportional sensitivities (Sx) of the independent components in an open-diet (25% use of all forage species) model used to estimate nutritional carrying capacity (NCC) for moose within the road-bias-adjusted west Delta winter range in a severe winter on the Copper River Delta, Alaska. Samples used within the model were collected in 2012–2013.
Changes (%) and proportional sensitivities (Sx) of the independent components in an open-diet (25% use of all forage species) model used to estimate nutritional carrying capacity (NCC) for moose within the road-bias-adjusted west Delta winter range in a severe winter on the Copper River Delta, Alaska. Samples used within the model were collected in 2012–2013.

Variability in model structure strongly influenced NCC estimates generated from our three model structures, winter scenarios, and winter ranges. Thus, evaluating different winter NCC models for moose on the CRD highlighted the importance of matching model structure to species' behavioral, ecological, and topological constraints, and the difficulty of comparing NCC estimates even within the same system. The model structures that resulted in the greatest limitations of forage (“diet-restricted” with 25% use, severe winter scenarios) estimated NCC below the current population, and may explain the 15–20% population decrease following the 1970–1971 severe winter, when space use by, and thus forage availability for, moose may have been highly constrained (MacCracken et al. 1997). Observations that the current population persisted through another severe winter in 2011–2012 (Figure 2; Text S1, Data S1, Supplemental Material) suggest the restricted range used in those models no longer adequately describes the area over which moose are currently accessing resources. Aerial surveys suggest moose have expanded their winter range; therefore, increased access to forage potentially buffered the 2011–2012 population in a historically severe winter (mean snow depth = 101 cm). This suggests that models dependent on assumptions of restricted winter range size (total past observed) are not currently accurate. Assessing space use in severe and mild winters should be a prerequisite for future attempts to measure NCC here and in similar systems.

Density-dependent and seasonal factors in herbivores and forage may influence browsing behavior or diets over time and space (Vivas and Saether 1987). In this analysis, the “diet-restricted” structures constrained diet preferences to those observed in 1987–1989, but potential for seasonal and temporal shifts in diet preference since then were poorly understood. Hobbs and Hanley (1990) cautioned against assuming that herbivores are limited to the diets observed during a specific period. In Kruger National Park, 10 of 11 ungulates studied displayed significant seasonal changes in grass and shrub consumption (Codron et al. 2007b), while Codron et al. (2007a) noted regional differences may explain disparities in diet recorded for African ungulates, and a study in Japan found notable differences in diets across individual serows Capricornis crispus during winter (Jiang et al. 2008). Seasonally specific free-choice diets are unknown for the CRD moose, but the recent severe winter survival (2012) and our record of high sweetgale use (Smythe 2015) suggests the limited diet assumed by the “diet-restricted” models underestimates the amount of forage available, and that applying models with limited diets in any system would require detailed understandings of diet preferences under different conditions. Considering this unknown variability, the “open-diet” model could be seen as a physiologically generous alternative, appealing to the indefinite limits of ungulate diet selection. However, the open-diet model contradicts well-known studies of ungulate diet selection and preference, which cannot be ignored even if they are not currently ecologically complex enough. Seasonally comprehensive and long-term diet studies are needed to resolve this issue.

As in other studies of NCC, our attempts to determine ecologically reliable estimates of NCC for the west CRD moose were hampered by several factors. Lack of recent detailed data on space use and resource selection by moose obscured the appropriate definition of a current moose winter range. Although the observed range defined by MacCracken is overly restrictive, moose are likely not using the full west Delta range equally or entirely. We estimated capacity for higher moose densities within the observed range, suggesting some areas could be preferred by moose. The factors contributing to this preference are largely unknown, but could include past population occurrence and exploration, human factors, or predator influences, among others beyond the influence of snow depth and forage availability discussed. Both humans and predators are known to influence ungulate space use (Brown 1999; Theuerkauf and Rouys 2008), often restricting ungulate movement to areas with less abundant or lower quality forage associated with decreased predation risk. Ungulates may also select for areas associated with humans—elk Cervus canadensis in Banff National Park utilized refugia created by human activity to decrease risk of wolf predation (Hebblewhite and Merrill 2009). In areas with limited predator effects, energy expenditure associated with increasing range sizes may function as the only significant limit to population distribution, such as with caribou Rangifer tarandus in northern Quebec (Messier et al. 1988). Nutritional carrying-capacity models that lack behavioral, predatory, or physiological restraints to a population's range risk overestimating the forage available.

The browsing tolerance of forage species is unknown on the CRD and in many other systems. Studies in Denali and interior Alaska found winter use of forage by moose varied from 0.9–31.2% to 1–46%, respectively, and reported some short-term changes in secondary chemicals and nutrition. Neither study evaluated the long-term effects of high utilization (MacCracken and Viereck 1990; Miquelle et al. 1992). A review by Augustine and McNaughton (1998) found that intermittent herbivory (e.g., due to migration or nonherding) and nutrient-rich ecosystems aided the regrowth capacity of palatable forage species. Persson et al. (2007) confirmed that regrowth ability varied along a habitat productivity gradient and found birch Betula pubescens and B. pendula could tolerate 25–40% use in productive sites. Considering the influx of nutrients through tidal, glacial, and actinorhizal sources on the Delta, forage may be able to support greater levels (>25%) of consumption, and NCC models that restrict forage use may underestimate the forage consumable by populations without causing ecological damage. Alternatively, moose, as large herbivores, are a relatively recent pressure on forage species; therefore, it also possible the browsing tolerance of shrubs on the CRD may be <25% use, in which case sustainable NCC estimates must be reduced.

The dynamic nature of the forces governing the CRD vegetation make all estimates of NCC temporary. The hydrology, soil quality, land area, form, and successional patterns of the vegetation of the Delta are dictated by several landscape-level forces including glacial, stream, and tidal flooding, sedimentation, erosion, and salinization, along with periodic earthquake-initiated cycles of uplift and subsidence (Boggs 2000). These geological factors may interact with zoological and human influences (Maurer and Ray 1992; Danell et al. 1994; MacCracken et al. 1997; Christensen et al. 2000; Wright et al. 2002; Stephenson et al. 2006; Cooper 2007; Smythe et al. 2016) to influence forage availability in opposing ways (potentially increasing or decreasing forage available to moose; reviewed in Smythe 2015). Furthermore, as a result of its levee-like action of trapping water north of the road, and reducing both the rate and dynamics of flow south of the road, the Copper River highway acts as a constant manmade influence on the hydrology and, in turn, the vegetation growing within its influence. Our approach to estimating the spatial reach of road effects on vegetation biomass would have benefitted from additional sample size designed specifically to address that effect. Full understanding of the complex, longer term hydrological and ecological influences of the road on the Delta vegetation warrants a separate, more detailed investigation.

Climate change will likely influence many factors within NCC models and is expected to have disproportionately strong influences in other high-latitude ecosystems (IPCC 2007). We found potential evidence of increased willow stem biomass relative to 1988–1989 data (Smythe 2015), possibly due to increased mean temperatures across Alaska (ACRC 2014), while severe winters on the CRD have diminished from four in the 20-y period from 1917 to 1936 to one thus far in 1997–2015. Rates and patterns of succession may increase with further increases in temperature, nutritional quality of forage may benefit from earlier or increased growing seasons, and snow effects on forage availability may decrease as winter severities continue to lessen. Alternatively, forage biomass and nutrition may experience earlier senescence or endure harsher summer conditions with warmer temperatures, possibly creating mismatches in peak forage nutrition and peak nutritional need for lactating moose. Furthermore, as in some parts of the lower 48 states, the temperatures in Alaska may eventually exceed those physiologically manageable by moose. The CRD managers would benefit from monitoring and research in these areas.

The CRD landscape is highly dynamic, and McLeod (1997) questioned the validity of calculating carrying capacity within any nonequilibrium ecosystem. Our current winter NCC estimates will, however, inform management goals in the near future and serve as an updated benchmark as geological, hydrological, successional, climatic, zoological, and cultural influences continue to drive vegetation change. Our work serves as a reminder of the change inherent to most ecosystems and its continuous effect on animal populations, even within relatively stable landscapes. Estimates of NCC on the Delta will need to be recalibrated periodically to guide management as habitat changes, but future efforts will be informed by the lessons from this comparative study. Similarly, NCC estimates globally may require periodic reevaluation, depending on the stability of the system and factors affecting model components.

Our NCC models with larger winter ranges generally suggest that a larger moose population could be sustained on the CRD, depending on the winter scenario. Past authors used severe winter scenarios for developing conservative management goals, and our sensitivity analysis indicates snow depth is the most influential model factor, and winter length contributed to overall declines in NCC with increasing winter severity. However, the severe winter scenario may decrease in relevance if climate change reduces the frequency of those conditions, and moderate winter scenarios may be more appropriate for management goals. Delta-wide data on daily snow depths, drift formation, melt rates, and compaction are not currently available but would aid managers in weighting this factor. Given the importance of moose hunting for subsistence in the region, we further recommend managers consider setting population size targets at some margin above the level at which maximum sustained yield would be obtained (e.g., half of NCC; Caughley 1976) but lower than total NCC, depending on the model deemed most appropriate with current knowledge (which should likely be a summer-based NCC model; see below). That margin would protect against issues associated with inaccurate population estimates or within a highly variable environment that can lead to delayed recognition of over- or underharvest, population instability, and ecosystem damage (Larkin 1977; Lande et al. 1997; McCullough 2001). Given that moose are an introduced megaherbivore in this Critical Habitat Area (Alaska Department of Natural Resources 2007), managers also must carefully consider the effect of a higher moose population on vulnerable native species. More reliable and continuous space use and vitality data would greatly aid management of this population. Our sensitivity analyses suggest that managing for higher stem densities or larger areas with forage would most efficiently improve forage availability, supporting current mechanical treatment (e.g., hydraulic-axing) efforts on the CRD (Stephenson et al. 1998; Smythe et al. 2016).

Despite the low proportional sensitivity of tannins and lignin in our analysis, inclusion of nutritional quality and use of energy-based models can dramatically influence NCC estimates (Windels and Hewitt 2011; Smythe 2015). We recommend models of NCC incorporate species-specific nutritional values, and constrain available biomass (<100% use) within NCC models to allow forage recovery (Singer and Zeigenfuss 2002). We also strongly recommend estimating summer-based NCC on the CRD in the future because 1) considerable research finds ungulate reproductive success (milk yields, calf growth and viability, etc.) depends on fat reserves gained during spring and summer rather than reserves lost during winter; and 2) the forage available to the CRD moose is no longer limited by a restricting winter range; therefore, consideration of full-year forage availability may be important to understand the population's current limiting factors (Bobek 1977; Mautz 1978a; Merrill and Boyce 1991; Cook et al. 2001, 2004; Beck et al. 2006; Dungan et al. 2010). Alternatively or in complementary fashion, we recommend estimates of carrying capacity conducted via “animal indicators” (Franzmann and Schwartz 1985; Stephenson 1995; Cook et al. 2013; Monteith et al. 2013) as a supplementary method for estimating the sustainability of this population within a nuanced environment.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Text S1. Program R code to create population graph (Figure 2) in conjunction with Data S1 (Supplemental Material) Excel data to estimate moose population abundance on the West Copper River Delta, Alaska.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S1 (2 KB R).

Text S2. Program R code to test the relationship between nutritional energy for moose (Copper River Delta, Alaska) in vegetation plots relative to their distance from the roads (Figure S1, Table S3, Supplemental Material) as a baseline for the road buffer defined the adjusted west Delta range; in conjunction with Data S2 (Supplemental Material) Excel data.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S2 (13 KB R).

Text S3. Program R code to derive summarized nutritional values for moose forage species on western Copper River Delta of Alaska (Tables S1 and S2, Supplemental Material) in conjunction with Table S2 Excel data.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S3 (15 KB R).

Text S4. Program R code to summarize snow depths specific to winter scenarios (see Results) on the western Copper River Delta, Alaska, in conjunction with Data S3 (Supplemental Material) Excel data.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S4 (103 KB R).

Text S5. Winter-range total sizes (ha) and range-specific stand-type components (ha) for moose on the western Copper River Delta, Alaska, were determined using NCCRCode and Geographic Information System–based calculations of landcover rasters (SPOT 5 2011; U.S. Forest Service).

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S3 (15 KB R).

Text S6. Program R code to summarize digestible energy values of moose forage species present by stand type on the western Copper River Delta, Alaska (Figure 4) in conjuction with Data S2 (Supplemental Material) Excel data.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S3 (15 KB R).

Text S7. Program R code to summarize the NCC of moose specific to each model type, winter scenario, and winter range on the western Copper River Delta, Alaska (Figure 4, Data S3, Supplemental Material) in conjunction with Data S2 (Supplemental Material) Excel data.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S3 (15 KB R); https://doi.org/10.3996/122017-JFWM-104.S5 (103 KB R).

Text S8. Program R code to summarize percent changes and proportional sensitivities of model components in moose NCC models for the western Copper River Delta, Alaska (Table 2); to be used in conjunction with Data S2 (Supplemental Material) Excel data.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S3 (15 KB R).

Data S1. MoosePop Excel; used to plot changes in the moose Alces alces population on the west Copper River Delta in south-central Alaska from 1949 to 2014 using different survey methods. Accompanied by a ReadMe file.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S6 (10 KB XLSX); https://doi.org/10.3996/122017-JFWM-104.S7 (12 KB DOCX).

Data S2. MooseRData Excel; used to support generation of nutritional values of forage species collected in 2012–2013 and consumed by moose Alces alces on the Copper River Delta, Alaska. Data were analyzed by species and stand type, used in subsequent estimates of nutritional carrying capacity across three winter scenarios (Text S2 and Text S4, Supplemental Material), and also provides data for sensitivity analyses (Text S2, Supplemental Material) for the nutritional carrying capacity models. Accompanied by a ReadMe file.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S8 (1.45 MB XLSX); https://doi.org/10.3996/122017-JFWM-104.S9 (314 KB RTF).

Data S3. SnowScenarioData Excel; Used for calculation of total snow depths and winter lengths from 1914 to 2014 on the Copper River Delta, Alaska. Accompanied by a ReadMe file.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S10 (5 KB CSV); (https://doi.org/10.3996/122017-JFWM-104.S11 (13 KB DOCX).

Figure S1. Changes in plot nutritional energy (a product of forage biomass and factors contributing to nutritional quality; see manuscript Methods; kcal/ha) relative to plot distance (m) from road structure in different stand types (black = spruce–hemlock, red = spruce–cottonwood, blue = cottonwood, green = alder, purple = alder–willow, orange = willow, and yellow = sweetgale). Winter forage species for moose found in the seven stand types were Barclay's willow, undergreen willow, Hooker's willow, and Sitka willow Salix barclayi, S. commutata, S. hookeriana, and S. sitchensis, Sitka alder Alnus viridis sinuata, black cottonwood Populus trichocarpa, and sweetgale Myrica gale. These were used to estimate moose Alces alces nutritional carrying capacity on the Copper River Delta, Alaska, using forage samples from 2012 to 2013. We tested for effects of plot distance from roads on nutritional energy via linear regression. Nutritional energy in cottonwood, alder–willow, willow, and sweetgale plots was significantly (P < 0.05) affected. To define the modeling (winter) range accounting for this effect, we retested for significance after successively removing points for stands located at increasing distances from the roads (e.g., <100 m, <200 m). Nutritional energy in cottonwood plots >100 m from roads, and from alder–willow, willow, and sweetgale plots >200 m from roads, were no longer significantly affected. Using these results, we then re-estimated the nutritional energy available (see manuscript Methods and Figure 3) in the affected stands using plots sampled nearer and beyond their respective significant distances (from roads) and calculated total nutritional energy available for each stand type across the modeling range we refer to as the adjusted west Delta range.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S12 (65 KB DOCX).

Table S1. Stand-type- and species-specific values of mean gross energy (GE; cal/g), crude protein (CP; %), lignin (L; %), tannin (T; mg/g BSA), and neutral detergent fiber (NDF; %), digestible energy (DE: cal/g) and 95% confidence intervals (±) for the seven winter forage species Barclay's willow, undergreen willow, Hooker's willow, and Sitka willow Salix barclayi, S. commutata, S. hookeriana, and S. sitchensis, Sitka alder Alnus viridis sinuata, black cottonwood Populus trichocarpa, and sweetgale Myrica gale found in the seven stand types (spruce–hemlock, spruce–cottonwood, cottonwood, alder, alder–willow, willow, and sweetgale) important to wintering moose Alces alces ecology, sampled late winter (April–May) of 2012–2013 on the west Copper River Delta, Alaska.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S13 (41 KB DOCX).

Table S2. Stand-type- and species-specific estimates of mean available biomass (AB, kg/ha; no snow burial or height limit) and consumption (Use, %) by moose Alces alces for the seven winter forage species Barclay's willow, undergreen willow, Hooker's willow, and Sitka willow Salix barclayi, S. commutata, S. hookeriana, and S. sitchensis, Sitka alder Alnus viridis sinuata, black cottonwood Populus trichocarpa, and sweetgale Myrica gale found in the seven stand types (spruce–hemlock, spruce–cottonwood, cottonwood, alder, alder–willow, willow, and sweetgale) relevant to moose winter ecology and sampled August–September 2012–2013 (biomass) and April–May 2013 (use) on the west Copper River Delta, Alaska.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S14 (32 KB DOCX).

Table S3. Regression results for stand-specific digestible energy (kcal/ha, y variable) available to moose Alces alces relative to the distance (m, x variable) of sampling plots from the Copper River Highway sampled in 2012–2013 on the Copper River Delta, Alaska.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S15 (22 KB DOCX).

Table S4. Nutritional carrying capacity estimates (no. of moose [density of moose/km2]), Alces alces using 2012–2013 forage data and estimated according to different model structures (“open-diet or diet-restricted” with 25% or 100% use of forage species), winter ranges (west Delta, road-bias-adjusted west Delta, and total past observed), and winter severities (mild, moderate, and severe) on the Copper River Delta, Alaska.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S16 (23 KB DOCX).

Reference S1. Boggs K. 2000. Classification of community types, successional sequences, and landscapes of the Copper River Delta, Alaska. Portland, Oregon: U.S. Department of Agriculture Forest Service, Pacific Northwest Research Station. General Technical Report PNW-GTR-469.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S17 (3.11 MB PDF).

Reference S2. Christensen HH, Mastrantonio L, Gordon J, Bormann B. 2000. Alaska's Copper River: humankind in a changing world. Portland, Oregon: U.S. Department of Agriculture Forest Service, Pacific Northwest Research Station. General Technical Report PNW-GTR-480.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S18 (207 KB PDF).

Reference S3. Davidson D, Harnish C. 1978. Soil and water resource inventory of the Copper River Delta. Anchorage: U.S. Department of Agriculture Forest Service Alaska Region, Chugach National Forest.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S19 (55.63 MB PDF).

Reference S4. Ferrians OJ Jr. 1966. Effects of the earthquake of March 27, 1964 in the Copper River Basin area, Alaska. Geological Survey Professional Paper 543 - E.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S20 (17.15 MB PDF).

Reference S5. Hanley TA, Rogers JJ. 1989. Estimating carrying capacity with simultaneous nutritional constraints. Portland, Oregon: U.S. Department of Agriculture Forest Service, Pacific Northwest Research Station. Resource Note PNW-RN-485.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S21 (958 KB PDF).

Reference S6. Kesti S, Burcham M, Campbell B, Davidson D, DeVelice R, Huber C, Joyce T, Lang D, MacFarlane B, Sherman D, Yarborough L. 2007. West Copper River Delta landscape assessment Cordova Ranger District Chugach National Forest.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S22 (6.48 MB PDF).

Reference S7. Maurer MA, Ray SR. 1992. Copper River highway environmental impact studies: water quality of surface waters. State of Alaska, Department of Natural Resources, Division of Geological & Geophysical Surveys, Public Data File 92-23.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S23 (1.77 MB PDF).

Reference S8. Miquelle DG, Jordan PA. 1970. The importance of diversity in the diet of moose. Proceedings of the North American Moose Conference and Workshop 15:54–79.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S24 (1.2 MB PDF).

Reference S9. Singer FJ, Zeigenfuss LC. 2002. Ecological evaluation of the abundance and effects of elk herbivory in Rocky Mountain National Park, Colorado, 1994-1999. Fort Collins, Colorado: U.S. Geological Survey, Fort Collins Science Center. U.S. Geological Survey Open File Report 02-208.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S25 (15.64 MB PDF).

Reference S10. Smythe SE. 2015. Evaluating tools used to estimate and manage browse available to wintering moose on the Copper River Delta, Alaska. Master's thesis. Corvallis: Oregon State University.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S26 (2.64 MB PDF).

Reference S11. Stover CW, Coffman JL. 1993. Seismicity of the United States, 1568–1989 (revised). U.S. Geological Survey Professional Paper 1527.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S27 (118 KB PDF).

Reference S12. Viereck LA. 1992. The Alaska vegetation classification. Portland, Oregon: U.S. Department of Agriculture Forest Service, Pacific Northwest Research Station. General Technical Report PNW GTR-286.

Found at DOI: https://doi.org/10.3996/122017-JFWM-104.S28 (3.95 MB PDF).

Special thanks to T. Joyce, M. Burcham, C. Westing, E. Cooper, D. Crowley, and seasonal employees who assisted with project development and data collection. R. Mata-Gonzalez, D. Hibbs, and J. Van Sickle assisted with this manuscript. G. Bracher developed Figure 5. S. Smythe was funded by (11-JV-11261953-069), and received logistical support from, the Cordova Ranger District of the U.S. Forest Service and the Alaska Department of Fish and Game, along with partial support from Oregon State University. We are grateful for the constructive and insightful comments and feedback provided by our reviewers and the Associate Editor. We have no conflicts of interest to report for this manuscript.

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.

[ACRC] Alaska Climate Research Center
.
2014
.
Applied climate information system daily data browser of the University of Alaska Fairbanks
.
Available: http://climate.gi.alaska.edu/acis_data (November 2017)
.
Alaska Department of Natural Resources
.
2007
.
Copper River Delta: Management Unit 28 in Prince William Sound Area plan for state lands
.
SC-88-004A03. Available: http://dnr.alaska.gov/mlw/planning/areaplans/prince/ (June 2017)
.
Alipayo
D,
Valdez
R,
Holechek
JL,
Cardenas
M.
1992
.
Evaluation of microhistological analysis for determining ruminant diet botanical composition
.
Journal of Range Management
45
:
148
152
.
Augustine
DJ,
McNaughton
SJ.
1998
.
Ungulate effects on the functional species composition of plant communities: herbivore selectivity and plant tolerance
.
Journal of Wildlife Management
62
:
1165
1183
.
Beck
JL,
Peek
JM,
Strand
EK.
2006
.
Estimates of elk summer range nutritional carrying capacity constrained by probabilities of habitat selection
.
Journal of Wildlife Management
70
:
283
294
.
Bobek
B.
1977
.
Summer food as the factor limiting roe deer population size
.
Nature
266
:
47
49
.
Boggs
K.
2000
.
Classification of community types, successional sequences, and landscapes of the Copper River Delta, Alaska
.
Portland, Oregon
:
U.S
.
Department of Agriculture Forest Service, Pacific Northwest Research Station. General Technical Report PNW-GTR-469 (see Supplementary Material, Reference S1)
.
Brown
JS.
1999
.
Vigilance, patch use and habitat selection: foraging under predation risk
.
Evolutionary Ecology Research
1
:
49
71
.
Bryant
JP,
Reichardt
PB,
Clausen
T.
1992
.
Chemically mediated interactions between woody plants and browsing mammals
.
Journal of Range Management
45
:
18
24
.
Caughley
G.
1976
.
Wildlife management and the dynamics of ungulate populations
.
Applied Biology
1
:
183
246
.
Caughley
G.
1979
.
What is this thing called carrying capacity?
Laramie
:
University of Wyoming Press
.
Caughley
G,
Shepherd
N,
Short
J.
1987
.
Kangaroos: their ecology and management in the sheep rangelands of Australia
.
Cambridge, UK
:
Cambridge University Press
.
Christensen
HH,
Mastrantonio
L,
Gordon
J,
Bormann
B.
2000
.
Alaska's Copper River: humankind in a changing world
.
Portland, Oregon
:
U.S
.
Department of Agriculture Forest Service, Pacific Northwest Research Station. General Technical Report PNW-GTR-480 (see Supplementary Material, Reference S2)
.
Codron
D,
Codron
J,
Lee-Thorp
J,
Sponheimer
M,
De Ruiter
D,
Sealy
J,
Grant
R,
Fourie
N.
2007
a
.
Diets of savanna ungulates from stable carbon isotope composition of faeces
.
Journal of Zoology
273
:
21
29
.
Codron
D,
Lee-Thorp
JA,
Sponheimer
M,
Codron
J.
2007
b
.
Stable carbon isotope reconstruction of ungulate diet changes through the seasonal cycle
.
South African Journal of Wildlife Research
37
:
117
125
.
Cook
JG,
Johnson
BK,
Cook
RC,
Riggs
RA,
Delcurto
T,
Bryant
LD,
Irwin
LL
2004
.
Effects of summer-autumn nutrition and parturition date on reproduction and survival of elk
.
Wildlife Monographs
155
:
1
61
.
Cook
RC,
Murray
DL,
Cook
JG,
Zager
P,
Monfort
SL.
2001
.
Nutritional influences on breeding dynamics in elk
.
Canadian Journal of Zoology
79
:
845
853
.
Cook
RC,
Cook
JG,
Vales
DJ,
Johnson
BK,
Mccorquodale
SM,
Shipley
LA,
Riggs
RA,
Irwin
LL,
Murphie
SL,
Murphie
BL,
Schoenecker
KA,
Geyer
F,
Briggs Hall
P,
Spencer
RD,
Immell
DA,
Jackson
DH,
Tiller
BL,
Miller
PJ,
Schmitz
L
.
2013
.
Regional and seasonal patterns of nutritional condition and reproduction in elk
.
Wildlife Monographs
184
:
1
45
.
Cooper
E.
2007
.
Beaver ecology on the Copper River Delta. Master's thesis
.
Corvallis
:
Oregon State University
. .
Danell
K,
Bergström
R,
Edenius
L.
1994
.
Effects of large mammalian browsers on architecture, biomass, and nutrients of woody plants
.
Journal of Mammalogy
75
:
833
844
.
Davidson
D,
Harnish
C.
1978
.
Soil and water resource inventory of the Copper River Delta
.
Anchorage
:
U.S. Department of Agriculture Forest Service Alaska Region, Chugach National Forest (see Supplementary Material, Reference S3)
.
Dungan
JD,
Shipley
LA,
Wright
RG.
2010
.
Activity patterns, foraging ecology, and summer range carrying capacity of moose (Alces alces shirasi) in Rocky Mountain National Park, Colorado
.
Alces
46
:
71
87
.
Ferrians
OJ
Jr.
1966
.
Effects of the earthquake of March 27, 1964 in the Copper River Basin area, Alaska
.
Geological Survey Professional Paper 543 - E
(
see Supplementary Material, Reference S4
).
Franzmann
AW,
Schwartz
CC.
1985
.
Moose twinning rates: a possible population condition assessment
.
Journal of Wildlife Management
49
:
394
396
.
Gasaway
WC,
DuBois
SD,
Reed
DJ,
Harbo
SJ.
1986
.
Estimating moose population parameters from aerial surveys
.
Fairbanks
:
University of Alaska, Institute of Arctic Biology
.
Report 0568-8604. Available: https://scholarworks.alaska.edu/handle/11122/1498 (January 2019). Archived by WebCite: http://www.webcitation.org/75hGCfsWF
.
Hanley
TA,
Rogers
JJ.
1989
.
Estimating carrying capacity with simultaneous nutritional constraints
.
Portland, Oregon
:
U.S
.
Department of Agriculture Forest Service, Pacific Northwest Research Station. Resource Note PNW-RN-485. (see Supplementary Material, S5)
.
Hebblewhite
M,
Merrill
EH.
2009
.
Trade-offs between predation risk and forage differ between migrant strategies in a migratory ungulate
.
Ecology
90
:
3445
3454
.
Hobbs
NT,
Baker
DL,
Ellis
JE,
Swift
DM,
Green
RA.
1982
.
Energy-and nitrogen-based estimates of elk winter-range carrying capacity
.
Journal of Wildlife Management
46
:
12
21
.
Hobbs
NT,
Gross
JE,
Shipley
LA,
Spalinger
DE,
Wunder
BA.
2003
.
Herbivore functional response in heterogeneous environments: a contest among models
.
Ecology
84
:
666
681
.
Hobbs
NT,
Hanley
TA.
1990
.
Habitat evaluation: do use/availability data reflect carrying capacity?
Journal of Wildlife Management
54
:
515
522
.
Hobbs
NT,
Swift
DM.
1985
.
Estimates of habitat carrying capacity incorporating explicit nutritional constraints
.
Journal of Wildlife Management
49
:
814
822
.
Holechek
JL,
Vavra
M,
Pieper
RD.
1982
.
Botanical composition determination of range herbivore diets: a review
.
Journal of Range Management
35
:
309
315
.
[IPCC] Intergovernmental Panel on Climate Change
.
2007
.
Climate change 2007: the physical science basis
.
In
Solomon
S,
Qin
D,
Manning
M,
Chen
Z,
Marquis
M,
Averyt
KB,
Tignor
M,
Miller
HL,
editors
.
Contribution of working group I to the fourth assessment report of the Intergovernmental Panel on Climate Change
.
Cambridge, UK; and New York
:
Cambridge University Press
.
Jefferies
R,
Klein
D,
Shaver
G.
1994
.
Vertebrate herbivores and northern plant communities: reciprocal influences and responses
.
Oikos
71
:
193
206
.
Jiang
Z,
Torii
H,
Takatsuki
S,
Ohba
T.
2008
.
Local variation in diet composition of the Japanese serow during winter
.
Zoological Science
25
:
1220
1226
.
Kesti
S,
Burcham
M,
Campbell
B,
Davidson
D,
DeVelice
R,
Huber
C,
Joyce
T,
Lang
D,
MacFarlane
B,
Sherman
D,
Yarborough
L.
2007
.
West Copper River Delta landscape assessment Cordova Ranger District Chugach National Forest
(
See Supplementary Material, Reference S6
).
Lande
R,
Sæther
BE,
Engen
S.
1997
.
Threshold harvesting for sustainability of fluctuating resources
.
Ecology
78
:
1341
1350
.
Larkin
PA.
1977
.
An epitaph for the concept of maximum sustained yield
.
Transactions of the American Fisheries Society
106
:
1
11
.
MacCracken
JG,
Van Ballenberghe
V.
1993
.
Mass-diameter regressions for moose browse on the Copper River Delta, Alaska
.
Journal of Range Management
46
:
302
308
.
MacCracken
JG,
Van Ballenberghe
V,
Peek
JM.
1997
.
Habitat relationships of moose on the Copper River Delta in coastal south-central Alaska
.
Wildlife Monographs
136
:
3
52
.
MacCracken
JG,
Viereck
LA.
1990
.
Browse regrowth and use by moose after fire in interior Alaska
.
Northwest Science
64
:
11
18
.
Martin,
JS,
Martin
MM.
1982
.
Tannin assays in ecological studies: lack of correlation between phenolics, proanthocyanidins and protein-precipitating constituents in mature foliage of six oak species
.
Oecologia
54
:
205
211
.
Maurer
MA,
Ray
SR.
1992
.
Copper River highway environmental impact studies: water quality of surface waters
.
State of Alaska, Department of Natural Resources, Division of Geological & Geophysical Surveys, Public Data File 92-23
(
see Supplementary Material, Reference S7
).
Mautz
WW.
1978
a
.
Nutrition and carrying capacity
.
Chapter 22 in Schmidt JL,
Gilbert
DL,
editors
.
Big game of North America: ecology and management
.
Harrisburg, Pennsylvania
:
Stackpole Books
.
Mautz
WW.
1978
b
.
Sledding on a bushy hillside: the fat cycle in deer
.
Wildlife Society Bulletin
6
:
88
90
.
McCullough
DR.
2001
.
Population manipulations of North American deer Odocoileus spp.: balancing high yield with sustainability
.
Wildlife Biology
7
:
161
170
.
McEwan
E,
Whitehead
P.
1970
.
Seasonal changes in the energy and nitrogen intake in reindeer and caribou
.
Canadian Journal of Zoology
48
:
905
913
.
McLeod
SR.
1997
.
Is the concept of carrying capacity useful in variable environments?
Oikos
79
:
529
542
.
Merrill
EH,
Boyce
MS.
1991
.
Summer range and elk population dynamics in Yellowstone National Park
.
Pages
263
273
in
Keiter
RB,
Boyce
MS,
editors
.
The Greater Yellowstone Ecosystem: redefining America's wilderness heritage
.
New Haven, Connecticut
:
Yale University Press
.
Messier
F,
Huot
J,
Le Henaff
D,
Luttich
S.
1988
.
Demography of the George River caribou herd: evidence of population regulation by forage exploitation and range expansion
.
Arctic
41
(
4
):
279
287
.
Miquelle
DG,
Jordan
PA.
1970
.
The importance of diversity in the diet of moose
.
Proceedings of the North American Moose Conference and Workshop
15
:
54
79
(
see Supplementary Material, Reference S8
).
Miquelle
DG,
Peek
JM,
Van Ballenberghe
V.
1992
.
Sexual segregation in Alaskan moose
.
Wildlife Monographs
122
:
3
57
.
Monteith
KL,
Bleich
VC,
Stephenson
TR,
Pierce
BM,
Conner
MM,
Kie
JG,
Bowyer
RT.
2014
.
Life-history characteristics of mule deer: effects of nutrition in a variable environment
.
Wildlife Monographs
186
:
1
62
.
Monteith
KL,
Stephenson
TR,
Bleich
VC,
Conner
MM,
Pierce
BM,
Bowyer
RT.
2013
.
Risk-sensitive allocation in seasonal dynamics of fat and protein reserves in a long-lived mammal
.
Journal of Animal Ecology
,
82
:
377
388
.
Parker
KL,
Robbins
CT,
Hanley
TA.
1984
.
Energy expenditures for locomotion by mule deer and elk
.
Journal of Wildlife Management
48
:
474
488
.
Persson
IL,
Bergström
R,
Danell
K.
2007
.
Browse biomass production and regrowth capacity after biomass loss in deciduous and coniferous trees: responses to moose browsing along a productivity gradient
.
Oikos
116
:
1639
1650
.
Regelin
WL,
Schwartz
CC,
Franzmann
AW.
1985
.
Seasonal energy metabolism of adult moose
.
Journal of Wildlife Management
49
:
388
393
.
Renecker
LA,
Hudson
RJ.
1986
.
Seasonal foraging rates of free-ranging moose
.
Journal of Wildlife Management
50
:
143
147
.
Renecker
LA,
Schwartz
CC.
1997
.
Food habits and feeding behavior
.
Pages
403
440
in
Franzmann
AW,
Schwartz
CC,
editors
.
Ecology and management of the North American moose
.
Washington, D.C
.:
Smithsonian Institution Press
.
Schwartz
CC,
Hubbert
ME,
Franzmann
AW.
1988
.
Energy requirements of adult moose for winter maintenance
.
Journal of Wildlife Management
52
:
26
33
.
Schwartz
CC,
Renecker
LA.
1997
.
Nutrition and energetics
.
Pages
441
478
in
Franzmann
AW,
Schwartz
CC,
editors
.
Ecology and management of the North American moose
.
Washington, D.C
.:
Smithsonian Institution Press
.
Singer
FJ,
Zeigenfuss
LC.
2002
.
Ecological evaluation of the abundance and effects of elk herbivory in Rocky Mountain National Park, Colorado, 1994–1999
.
Fort Collins, Colorado
:
U.S. Geological
Survey, Fort Collins Science Center. U.S. Geological Survey Open File Report 02-208 (see Supplementary Material, Reference S9)
.
Smythe
SE.
2015
.
Evaluating tools used to estimate and manage browse available to wintering moose on the Copper River Delta, Alaska. Master's thesis
.
Corvallis
:
Oregon State University (see Supplementary Material, Reference S10)
.
Smythe
SE,
Sanchez
DM,
Mata-Gonzalez
R.
2016
.
Temporal effects of mechanical treatment on winter moose browse in south-central Alaska
.
Alces
51
:
135
147
.
Spalinger
DE,
Collins
WB,
Hanley
TA,
Cassara
NE,
Carnahan
AM.
2010
.
The impacts of tannins on protein, dry matter, and energy digestion in moose (Alces alces)
.
Canadian Journal of Zoology
88
:
977
987
.
Stephenson
TR,
Van Ballenberghe
V,
Peek
JM.
1998
.
Response of moose forages to mechanical cutting on the Copper River Delta, Alaska
.
Alces
34
:
479
494
.
Stephenson
TR,
Van Ballenberghe
V,
Peek
JM,
MacCracken
JG.
2006
.
Spatio-temporal constraints on moose habitat and carrying capacity in coastal Alaska: vegetation succession and climate
.
Rangeland Ecology & Management
59
:
359
372
.
Stover
CW,
Coffman
JL.
1993
.
Seismicity of the United States, 1568–1989 (revised)
.
U.S. Geological Survey Professional Paper 1527
(
see Supplementary Material, Reference S11
).
Theuerkauf
J,
Rouys
S.
2008
.
Habitat selection by ungulates in relation to predation risk by wolves and humans in the Białowieża Forest, Poland
.
Forest Ecology and Management
256
:
1325
1332
.
Thilenius
JF.
2008
.
Phytosociology and succession on earthquake-uplifted coastal wetlands, Copper River Delta, Alaska
.
Darby, Pennsylvania
:
Diane Publishing
.
Van Gils
JA,
Edelaar
P,
Escudero
G,
Piersma
T.
2004
.
Carrying capacity models should not use fixed prey density thresholds: a plea for using more tools of behavioural ecology
.
Oikos
104
:
197
204
.
Viereck
LA.
1992
.
The Alaska vegetation classification
.
Portland, Oregon
:
U.S
.
Department of Agriculture Forest Service, Pacific Northwest Research Station. General Technical Report PNW GTR-286 (see Supplementary Material, Reference S12)
.
Visscher
DR,
Merrill
EH,
Fortin
D,
Frair
JL.
2006
.
Estimating woody browse availability for ungulates at increasing snow depths
.
Forest Ecology and Management
222
:
348
354
.
Vivas
HJ,
Saether
BE.
1987
.
Interactions between a generalist herbivore, the moose Alces alces, and its food resources: an experimental study of winter foraging behaviour in relation to browse availability
.
Journal of Animal Ecology
56
:
509
520
.
[WRCC] Western Regional Climate Center
.
2015
.
Alaska climate summaries
. .
Windels
SK,
Hewitt
DG.
2011
.
Effects of plant secondary compounds on nutritional carrying capacity estimates of a browsing ungulate
.
Rangeland Ecology & Management
64
:
264
275
.
Wright
JP,
Jones
CG,
Flecker
AS.
2002
.
An ecosystem engineer, the beaver, increases species richness at the landscape scale
.
Oecologia
132
:
96
101
.

Author notes

Citation: Smythe SE, Sanchez DM, Epps CW. 2019. Contrasting winter moose nutritional carrying capacity models on a dynamic landscape. Journal of Fish and Wildlife Management 10(1):163–179; e1944-687X. https://doi.org/10.3996/122017-JFWM-104

The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.

Supplemental Material