~C>SCIENCE 4 (2): 170-178 (1997)

Growth responses of Carex ramenskii to defoliation, salinity, and nitrogen availability: Implications for geese-ecosystem dynamics in western Alaska!

Roger W. RUESS, Institute of Arctic Biology, University of Alaska, Fairbanks, Alaska 99775, U.S.A., e-mail: ffrwr@aurora.alaska.edu ( Daniel D. ULIASSI, Christa P. H. MULDER & Brian T. PERSON, Department of Biology and Wildlife, University of

Alaska, Fairbanks, Alaska 99775, U.S.A.

Abstract: The Yukon-Kuskokwim River Delta in western Alaska is the principal nesting area for several species of geese, including Pacific black brant. Grazing by geese on Carex ramenskii, one of the most abundant plant species throughout much of this region, appears to have increased in recent years. The purpose of this study was (i) to evaluate the effects of early-season defoliation and fertilization on plant growth and nutrient cycling processes within field plots of C. ramenskii over a 3-year period, and (ii) to study the interactive effects of defoliation, N availability, and salinity stress on growth, and biomass and N allocation in C. ramenskii under controlled, greenhouse conditions. Relative to control plots, clipped-fertilized plots showed significant increases in aboveground net primary production (AGNPP) and leaf N concentration, resulting in significant increases in offtake biomass and offtake N during both 1991 and 1992. In the greenhouse, total production of clipped and unclipped plants did not differ, but clipped plants had significantly higher offtake biomass, and biomass and N allocation to offtake compared to unclipped plants. Both field and laboratory experiments found that rapid regrowth following defoliation was dependent on soil nutrient availability. Fertilization increased soil respiration rates each year, but tended to decrease rates of net N mineralization, indicating that the soil microbial biomass is a strong nutrient sink in this ecosystem. In addition to the direct positive effects that goose feces have on plant growth, nitrogen recycled through feces may be an important source of nitrogen contributing to salinity tolerance in C. ramenskii. Our results also suggest that the observed increase in grazing pressure on patches of C. ramenskii early in the growing season may in.crease forage quality and quantity within these swards, and have important implications for geese-ecosystem interactions at a time of rapid goose population increase. Keywords: goose grazing, nitrogen cycling, salt marsh, Alaska.

Resume: Le delta forme par les rivieres Yukon et Kuskokwim (ouest de l'Alaska) est Ie principal site de nidification de plusieurs especes d'oies, dont la bemache cravant. Le broutage par les oies de Carex ramenskii, une des especes vegetales les plus abondantes de la region, semble avoir ete plus intensif au cours des dernieres annees. Cette etude poursuivait deux objectifs. En premier lieu, nous voulions evaluer, au cours d'une periode de trois ans et au sein de parcelles contenant des individus de C. ramenskii, les effets de la defoliation hiitive des Carex et de la fertilisation sur la croissance des plantes et sur Ie processus de recyclage des elements nutritifs. En second lieu, nous voulions etudier, en conditions controlees (serre), les effets interactifs de la defoliation, de la disponibilite en azote et du stress dil 11 la salinite sur la croissance, la biomasse et I'allocation en azote de C. ramenskii. Les parcelles defoliees et fertilisees ont une production primaire nette au-dessus de la surface du sol et une concentration en azote des feuilles significativement superieures 11 celles des parcelles temoins. Cela a entraine une hausse significative de la biomasse et d' azote consomme par les oiseaux en 1991 et en 1992. En serre, la pro­duction totale des plantes defoliees et non defoliees n'a pas varie de fal<0n significative. Par ailleurs, les plantes defoliees avaient une biomasse consommee et un taux de biomasse et d'azote disponible significativement plus eleves par rapport aux plantes non defoliees. Les experiences sur Ie terrain et en serre montrent que la reprise de croissance rapide des plantes 11 la suite de la defoliation depend de la disponibilite des elements nutritifs du sol. La fertilisation augmente les taux de respira­tion du sol, mais tend 11 diminuer les taux de mineralisation de I'azote. Cela indique que la biomasse microbienne du sol represente un puits important pour les elements nutritifs dans cet ecosysteme. En plus des effets positifs directs des feces d'oie sur la croissance des plantes, I'azote recycle via les feces peut s'averer une source importante d'azote contribuant a augmenter Ie niveau de tolerance 11 la salinite de C. ramenskii. Nos resultats suggerent egalement que I' augmentation de la pression de broutage sur les communautes de C. ramenskii au debut de la saison de croissance peut augmenter la qualite et la quantite du broutage au sein de ces communautes. Cela peut avoir des consequences importantes en ce qui a trait aux interac­tions entre I'ecosysteme et les oies lors d'une periode caracterisee par une augmentation rapide du troupeau d'oies. Mots-des: broutage par les oies, cycle de I'azote, marais sale, Alaska.

Introduction

The Yukon-Kuskokwim River (Y-K) Delta encompasses a large expanse of coastal saltmarsh and inland wet meadow ecosystems in western Alaska. The coastal landscapes of this region, bounded on the north by Kokechik Bay (61 0 40' N,

1660 7' w), and on the south by Hazen Bay (61 0 0' N, 1650

15' w), are the principal nesting areas for black brant geese

lRec. 1996-04-26; acc. 1996-11-06.

(Branta bernicla nigricans), cackling Canada geese (B. canadensis minima), greater white-fronted geese (Anser albifrons frontalis) and emperor geese (A. canagicus). The region is characterized by subtle elevational changes over an area of approximately 160 000 km2, and landscape patterns in elevation, permafrost depth, and salinity are important determinants of plant community distribution, and the asso­ciated landscape selection and vegetation use by geese

(Kincheloe & Stehn, 1991; Babcock & Ely, 1994).

Carex ramenskii (Kom) is the most abundant plant species within river and slough levee meadows, and tidal flat meadow communities. Together they constitute the largest percentage of the coastal salt-marsh landscape in the Y-K Delta (Tande & Jennings, 1986; Kincheloe & Stehn, 1991; Babcock & Ely, 1994). On average, the intensity and frequency of goose grazing in these communities is substan­tially less than that in Carex subspathacea swards, which occur as monospecific grazing lawns along coastal mudflats and along some inland ponds and slough margins. These are similar to grazing lawns studied on the Hudson Bay lowlands (Cargill & Jefferies, 1984 a,b; Bazely & Jefferies, 1985; 1989; Hik & Jefferies, 1990). However, in specific areas or at certain times of the year, C. ramenskii is intensively grazed. A common example of this is at the abrupt interface between C. ramenskii and C. subspathacea-dominated communities, where C. ramenskii is intensively grazed throughout the growing season, and has a short, dwarfed growth form strikingly similar to that of C. subspathacea. Another habitat where C. ramenskii is frequently grazed throughout the growing season is along the margins of inland ponds, particularly in the vicinity of nesting brant geese. Finally, during spring ice break-up, uneven patterns of ice and snowmelt expose patches of C. ramenskii to grazing by geese prior to nest initiation. In all these instances, it appears that once grazed, repeated grazing shifts the growth form of C. ramenskii to a shorter and potentially more palat­able, rapidly-growing form that leads to repeated grazing (pers.observ.).

Our observations, and those of other researchers working on the Y-K Delta (C. Babcock, J. Sedinger, P. Flint, pers. comm.), suggest that brant geese have begun to rely more on C. ramenskii, particularly during ice break-up, and during summer in swards close to the boundary with C. subspathacea. The number of breeding pairs of black brant at the Tutakoke River colony, one of the four major colonies of the Y-K Delta, has increased recently from an estimated 1100 in 1986 to 4601 in 1993 (Sedinger et at., 1993), and distribution of C. subspathacea grazing lawns in the region may actually be limiting relative to this population increase. One of the purposes of this study was to begin to evaluate the consequences of this early-season grazing on C. ramen­skii meadows, by examining plant growth characteristics and nitrogen cycling processes in response to defoliation and fertilization within C. ramenskii swards over 3 years.

An additional purpose of this study was to determine the importance of soil nitrogen availability to salinity tolerance in C. ramenskii. Along the Bering Sea coast and inland slough levees, the average elevation above mean high tide of Carex ramenskii communities ranges from 5 to 20 cm above that of C. subspathacea swards. This results in C. ramenskii communities being flooded less frequently than C. subspathacea, which can be subjected to daily sea water inundation (Kincheloe & Stehn, 1991). However, along the upper margins of C. subspathacea swards, where C. ramenskii is subjected to more frequent flooding and intensive grazing, higher rates of nitrogen cycling by geese (Jefferies, 1988; Ruess, Hik & Jefferies, 1989) may improve salinity tolerance in C. ramenskii by increasing the avail-

ECOSCIE~E, VoL. 4 (2), 1997

ability of nitrogen for the synthesis of nitrogen-based osmoregulators. We tested this hypothesis by examining the interactive effects of defoliati.on, N availability, and salinity stress on growth, and biomass and N allocation in C. ramenskii under controlled greenhouse conditions.

Our results indicate that grazing may indeed increase the productivity and nutrient content of C. ramenskii swards, and create the potential for C. ramenskii swards to serve as alternative foraging habitats for geese.

Material and methods

FIELD EXPERIMENT

Twelve plots (1 m2 in a grid separated by 1 m walkways) were established in a large wet meadow of Carex ramenskii in early June 1991. We recognize that plots separated by such short distances do not represent true replicates; never­theless, we assume that these plots are representative vegetation and ecosystem responses of the surrounding landscape. Plots were assigned to one of 4 experimental treatments in a factorial design that included clipping (+/-) and fertilization (+/-). Clipped plots were clipped once at the beginning of the growing season each year to a height of 3 cm to simulate spring time grazing. Fertilized plots received slow release fertilizer in early spring each year that supplied 10 g m-2 each of N, P, and K. Standing biomass was sampled in all plots at the time of clipping, and again at the peak of the growing season in late July by clipping all vegetation within one 15 cm x 15 cm quadrat placed random­ly within each plot. Samples were dried in a field laboratory, then taken to Fairbanks and redried at 60°C, weighed, and a subsample analyzed for total N on a LECO CNS 2000 auto­analyzer. Net aboveground primary production (AGNPP) was calculated as the difference between biomass harvest­ed in late July minus that harvested in early June. We used the term total offtake biomass to refer to the total above­ground biomass clipped above the 3 cm height. Total off­take biomass in unclipped plants was material clipped at final harvest, while total offtake biomass in clipped plants was the sum of material clipped in the spring and that clipped at the final harvest. AGNPP was measured at inter­vals of 45, 48, and 43 days during 1991, 1992, and 1993, respectively. In 1993, depth of thaw was measured at 3 points within each plot on 4 June, 15 June, and 30 June using a 1.5 m steel rod.

One soil core (5 cm diameter x 10 cm depth) .was collected from each plot at the time of final harvest each year, stored in a polyethylene bag in a portable cooler, and sent within 24 hours to the laboratory in Fairbanks. Each core was split vertically into 2 laboratory replicates, and approximately 25 g dry weight of soil was incubated in a 497 mL mason jar in the dark at 12°C for 30 days. Soil respiration was measured every 10 days with the use of a Shimadzu 8A gas chromatograph. Net N mineralization rate was taken as the difference in mineral N (NH4+ + N03-)

extracted with 2N KCI at day 30 and day 0 of the incuba­tion. Total soil Nand C were measured on the LECO CNS 2000 autoanalyzer.

LABORATORY EXPERIMENT

Small plugs of Carex ramenskii (10 cm diameter x 15 cm

171

\.

RUESS ET AL.: GROWTH RESPONSES IN CAREX RAMENSKII

depth) were removed over a 100 m2 area from a wet meadow near the Tutakoke River in late July and sent to Fairbanks where they were kept outside in 15 cm deep tubs and allowed to senesce and winter harden. Plants were moved into the greenhouse in early winter, sub-divided into small plugs (approximately 10 tillers per plug) and placed in 20 cm pots containing equal amounts of perlite, vermiculite, and peat moss. All plants were treated with distilled water (every 2 days) and a dilute nutrient solution (113 Hoagland's every 4 days) for 2 weeks as a preconditioning treatment. Eighty pots were randomly assigned to one of 8 treatments in a full 23 factorial design that included 2 levels each of nitrogen, clipping, and salinity. Plants were grown for 12 weeks under an 18-hour photoperiod and PPFD of approxi­mately 600 ,umol m-2 s-1 provided by HPS lighting. High (14 mM) or low (1 mM) N was provided every 6 days as a balanced mixture of NH4+ and N03-; each solution had equal proportions of other macro- and micro-nutrients. Nutrient solutions were adjusted to a pH of 6.7 with either HCI or NaOH. Salinity treatments consisted of either 0 or 19.5 g NaCI per liter (0 and 19.5 %0 salinity), mixed as a component of the nutrient solution. Plants were either unclipped or clipped every 3 weeks to a height of 3 cm. All pots were flushed with distilled water every 10 days to prevent salt and/or nutrient accumulation. Plant water potential was determined on a similarly aged, fully-expanded leaf from 6 randomly selected pots from each treatment at mid-day prior to harvest using a Pressure Bomb (PMS Instrument Co., Corvallis, Oregon). It is likely that because of restricted pot volume at the end of the experiment, plants were more water stressed at this time than would be expected in the field; nevertheless, we believe treatment differences among plant water potential measurements are reflective of plant responses to actual field conditions.

At harvest, all aboveground biomass was clipped to a 3 cm height on all plants, and the total number of tillers was counted. Roots were washed free of potting mixture, and all tissues were dried at 60°C for 48 hours, and weighed to the nearest mg. Total offtake biomass in unclipped plants was material clipped at final harvest, while total offtake bio­mass in clipped plants was the sum of material clipped throughout the experiment including that at the final harvest. Aboveground biomass below 3 cm height was termed aboveground residual biomass. Root to shoot ratio was calculated as the ratio of root to aboveground biomass, the latter being the' sum of offtake and aboveground residual biomass.

After weighing, tissues were ground in a Wiley Mill and analyzed for total N as described above. Tissue N mass was the product of tissue weight and %N. Total plant N uptake was the ratio of total plant N mass to root mass (Ruess, 1988).

Field and greenhouse data were analyzed by ANOVA using a general linear models procedure, and a repeated measures ANOVA was used to evaluate interannual effects for the field experiment (SAS, 1985). Biomass data were log transformed and proportional data were arcsin square root transformed where necessary to meet ANOVA assump­tions. In the greenhouse experiment, five plants from various treatments died during the experiment and were omitted

from the analyses. Data presented from both experiments represent true means ± standard errors. Treatment effects expressed as a percent change in the mean use true means for the field experiment, but for the greenhouse experiment, use least squares means derived from PROC GLM for unbalanced designs (SAS, 1985).

Results

FIELD EXPERIMENT

In the first year of treatments (1991), addition of nutrients significantly increased peak season biomass (+ 64%), off­take biomass (+ 46%), and AGNPP (+ 96%) compared with corresponding values for untreated plants (Figures 1 and 2). Although clipping did not significantly affect peak season biomass (P = 0.10), both offtake biomass (+ 31 %, P < 0.05) and AGNPP (+ 57%, P < 0.05) were higher in clipped versus unclipped plots. The highest values of offtake biomass (891.1 ± 36.4 g m-2) and AGNPP (488.4 ± 36.4 g m-2 48 d-1)

were measured in plots which were both clipped and fertilized.

In 1992, treatment differences in standing biomass early in the season paralleled patterns of AGNPP measured in 1991. The addition of nutrients had similar effects on plant growth in 1992 as in the previous year, increasing peak season biomass (+ 66%), offtake biomass (+ 81 %), and AGNPP (+ 73%, p = 0.11). Clipping significantly decreased peak season biomass, although less so when plants received high amounts of nutrients (P = 0.08), but it had no significant effect on offtake biomass or AGNPP. Overall, 1992 was a more productive year, particularly for unclipped plots. Peak season biomass averaged 58% greater, while AGNPP (when expressed per day) was more than double values for unclipped plots in 1991. However, increases in offtake biomass (+ 27%) and AGNPP (+ 18%) in 1992, relative to values for 1991, were less in plots that were both clipped and fertilized, and clipped plots showed a 22% decline in offtake biomass and a 2% decline in AGNPP in 1992 compared to values for 1991. Unfortunately, several plots were grazed by geese in the early spring of 1993, and all plots were grazed heavily by geese prior to final harvest in 1993, preventing reliable estimates of AGNPP.

Fertilization increased the N concentration in peak season biomass each year (Figure 3), but there were no significant carry-over effects on early season plant N content the following year. Clipping had no overall effect on N content of early or peak season biomass in any year. However, during 1991 and 1993, clipping tended to increase plant N content on unfertilized plots, and decrease the plant N content on fertilized plots.

Treatments had no effect on total soil C or N content, and no differences were detected in these soil variables among years. Averaged across years and treatments, these values (%DWT) were: C =4.95 ± 0.24, and N = 0.266 ± 0.012. In 1991, clipping significantly reduced soil water content when measured in late July (- 7%, P < 0.05), while fertilizer additions tended to increase soil water content (+ 5%, P = 0.12). No treatment effects on soil water were found in 1992, but in 1993 fertilization increased soil water

172

---

\ "~ ECOSCIENCE, VOL. 4 (2), 1997

1991 1200 ....

D

1000 i1-NUT·" _NUT·"CLP"

800

600

400

200

0

19921400

] DNUT·" 1200 _ NUP" ··CLP

_NUp·· 1000

~ 800 '-'

600~ e Q 400

= 200

0

FIGURE I. The effects of clipping (± CLP) and fertilization (± NUT) on aboveground biomass at the beginning of the growing season (open bars), aboveground peak season biomass in late July (solid bars), and offtake biomass (gray bars) during 1991-1993. Data represent means ± SE (n =3). Legend bars indicate primary (NUT or CLP) and interactive (CLPNUT = clipping x fertilization interaction) treatment effects, where asterisks following abbreviations indicate +NUT > -NUT, or +CLP > -CLP at p < 0.10·, 0.05", 0.01 ..•. Asterisks preceding abbreviations indicate the opposite effect.

in unclipped plots (+ 15%, p < 0.05), and tended to decrease soil water in clipped plots (- 11%, p =0.11). During 1991, 1992, and 1993, mineral N (NH4+ + N03-) was significantly higher in fertilized plots (27.5, 54.10, and 23.82 Jig g-I dry weight of soil) compared with unfertilized plots (9.72, 9.65, and 3.04 Jig g-l dry weight of soil, respectively), as was

FIGURE 2. The effects of clipping (± CLP) and fertilization (± NUT) on aboveground net primary production (AGNPP) during 1991 and 1992. Abbreviations follow Figure 1.

extractable phosphorus (0.08 versus 0.02 Jig g-l dry weight of soil, measured in 1993 only).

Fertilization stimulated soil respiration rates every year (Figure 4). Clipping had no significant effect on soil respi­ration rates overall. Soil respiration in all years was posi­tively correlated with soil water content (Figure 5). Soil respiration rates did not vary among years, and there were no differences in treatment effects on respiration rates between years.

Rates of net N mineralization varied significantly among years (P < 0.0001). Net N mineralization was unaffected by treatments in 1991, and averaged 0.34 ± 0.10 Jig N g-l dry weight of soil dol across all plots. Fertilization significantly reduced net N mineralization in 1992 (p < 0.0001), and caused a slight decline in net N mineralization in 1993 (Figure 6).

In 1993, depth of thaw was similar across all plots in early June, averaging 28.7 ± 0.8 cm. By 15 June, depth of thaw of fertilized plots (38.7 ± 1.4 cm) was significantly less than that in unfertilized plots (43.0 ± 0.9 cm, p < 0.01), and clipped plots (42.2 ± 1.0 cm) showed a slight increase in depth of thaw compared with unclipped plots (39.3 ± 1.7 cm, p = 0.06). These patterns continued until 30 June, and although statistical differences among treatments disap­peared, fertilized only plots still showed the least depth of thaw (58.3 ± 2.6 cm) and clipped/fertilized plots the great­est depth of thaw (76.3 ± 17.3 cm).

LABORATORY EXPERIMENT

PLANT GROWTH AND BIOMASS ALLOCATiON

Among the three treatments, salinity had the largest effect on plant growth, decreasing total plant mass an aver­age of 40% (p < 0.0001, Tables I and II) at high salinity compared to growth in the absence of sodium chloride. Salinity reduced root mass (- 39%), aboveground residual mass (-18%), offtake mass (- 60%), biomass allocation to offtake (- 21%) (allp < 0.0001), and tiller production (- 22%, p < 0.05). The water potential of high salt plants (-4.2 ± 0.2 MPa, n =30) was significantly less than that of low salt plants (-2.3 ± 0.2 MPa, n =30). Salinity had no overall

173

RUESS ET AL.: GROWTH RESPONSES IN CAREX RAMENSKII

~

TABLE I. The effects of clipping, salinity, and nitrogen (N) availability on biomass (g), biomass allocation (%), tissue N content (%), total plant N uptake (mgN g roor l) and leaf K+ and Na+ content (mg g-l) of Carex ramenskii. Data represent means ± 1 SE (n = 10)

Unclipped Clipped

-SALT +SALT -SALT +SALT -N +N -N +N -N +N -N +N

GROWTH AND BIOMASS ALLOCATION Offtake 1.21 ±0.11 6.60 ± 0.28 0.78 ± 0.11 2.49 ± 0.19 1.53 ± 0.09 8.82 ± 0.24 1.28 ± 0.15 2.66 ± 0.20 Aboveground residual 2.10 ± 0.17 4.46 ± 0.24 2.07 ± 0.21 3.56 ± 0.30 2.08 ± 0.16 4.90 ± 0.42 2.30 ± 0.10 3.14 ± 0.15 Root 12.19 ± 1.37 10.78 ± 0.61 7.39 ± 1.11 7.32 ± 1.21 10.37 ± 1.30 10.89 ± 1.07 7.29 ± 1.65 4.72 ± 0.80 Total 15.50 ± 1.32 21.84 ± 0.92 10.24 ± 1.06 13.38 ± 1.38 13.99 ± 1.40 24.61 ± 1.31 10.87 ± 1.57 10.53 ± 0.95 Allocation to offtake 8.0 ± 0.6 30.3 ± 0.9 8.2 ± 1.1 19.5 ± 1.5 11.4±0.7 36.6 ± 1.9 14.0 ± 2.7 26.4 ± 2.2 Root: shoot ratio 3.8 ± 0.5 1.0 ± 0.1 2.8 ± 0.5 1.2 ± 0.2 2.9 ± 0.3 0.8 ± 0.1 2.2 ± 0.6 0.8 ± 0.1 Number of tillers 121 ± 10 209 ± 27 94± 8 172±19 134 ± 10 230 ± 25 105±9 159 ± 15

NUTRIENT RELATIONS Plant N uptake 8.9 ± 0.5 31.5 ± 1.3 14.1 ± 1.4 37.0 ± 2.9 10.0 ± 0.4 43.5 ± 4.6 18.7 ± 2.6 50.1 ± 6.2 Offtake N 1.31 ± 0.03 2.08 ± 0.05 1.60 ± 0.05 2.26 ± 0.06 1.66 ± 0.05 2.80 ± 0.07 2.02 ± 0.09 2.94 ± 0.05 Aboveground residual N 1.04 ± 0.01 1.87 ± 0.07 1.30 ± 0.03 1.95 ± 0.04 1.06 ± 0.04 1.73 ± 0.05 1.21 ± 0.07 1.79 ± 0.06 RootN 0.54 ± 0.01 1.07 ± 0.03 0.75 ± 0.03 1.73 ± 0.11 0.54 ± 0.01 1.32 ± 0.15 0.90 ± 0.07 1.63 0.13 LeafK+ 33.74 ± 0.43 33-.58 ± 0.91 21.91 ± 0.52 27.73 ± 0.90 34.44 ± 1.18 39.19 ± 0.62 37.16 ± 1.16 37.35 0.60 LeafNa+ 1.05 ± 0.12 0.59 ± 0.04 43.79 ± 5.88 20.70 ± 2.03 0.05 ± 0.02 0.32 ± 0.02 1.48 ± 0.42 1.68 0.26

effect on root to shoot ratio. High N increased total plant mass by 39% (p < 0.0001).

This was primarily a function of significant increases in aboveground tissues, resulting in an increase (+ 173%) in biomass allocation to offtake (all p < 0.0001). While N had no significant effect on root mass, high N plants had signifi­cantly lower root: shoot ratios compared with low N plants (p < 0.0001). High N also increased total number of tillers (+ 68%, p < 0.001). Nitrogen had no effect on plant water potential.

Clipping had no overall effect on total plant biomass, although clipping significantly affected patterns of biomass allocation. For example, clipping increased offtake mass (+ 27%, P < 0.0001) and biomass allocation to offtake (+ 32%, P < 0.0001). Clipping had a minor effect on root mass (- 10%, p = 0.18), but reduced root:shoot ratio (- 22%, p < 0.05) due to an allocation shift that favoured above­ground growth. Clipping had no effect on tiller production (+ 3%, P =0.50).

The largest number of significant treatment interactions occurred between salinity and N, due to the fact that high levels of salinity reduced plant responses to N. For example, N increased total plant mass by 57% under low salt condi­tions (p < 0.0001), but by only 14% (p = 0.27, n.s.) under high salt. Similarly, N-induced increases in offtake biomass were over twice as high at low salt compared with values under the high salt treatment. Nevertheless, when grown under high salt concentrations, N availability had a signifi­cant positive effect on plant growth. Of particular relevance to herbivory was the response of offtake biomass to added N, which more than doubled in high salt plants (p < 0.05). High N caused a slight improvement in plant water potential in clipped plants grown on high salt (+ 10%, n.s.).

Clipping increased offtake mass more when plants were grown at low salinity (+ 33%, P < 0.0001) compared to when plants were grown at high salinity (+ 23%, P = 0.07). No other significant interactions between salinity and clipping were found, although there were some interesting

TABLE II. ANOVA results listing percentage of significant variance explained by primary and interactive effects of clipping, salinity, and N availability on components of growth allocation, and nutrient relations in Carex ramenskii (P < 0.05*,0.01 **,0.001 ***)

Salt Nitrogen Clip Salt x N Salt x clip N x clip Salt x N x clip

GROWTH AND BIOMASS ALLOCATION Offtake Aboveground residual Root Total Allocation to offtake Root shoot ratio Number of tillers

24.4***a 5.9***a

28.2***a 39.1 ***a

4.5***a

9.8**a

49.6***b 54.7***b

15.9***b 65.4***b 45.0*** 36.4***b

2.2**c

6.4***c 2.9*

18.0*** 7.5***

7.8** 7.3*** 2.7*

0.7** 0.5** 1.0**

2.5*

NUTRIENT RELATIONS Total plant N uptake Offtake N Aboveground residual N RootN LeafK+ LeafNa+

3.4** 4.6*** 3.5***

14.5*** 15.9***a 28.9***

63.5***b 62.1 ***b

62.8***b 8.2***b 5.4***

5.0**c 23.1 ***c

1.1 *c

41.6***c 19.7***

0.6* 1.1*

1.3* 5.9***

14.5*** 19.4***

2.0* 0.9*

2.7**

1.4* 5.0*** 2.8***

a Low salt> high salt. b HighN > low N. c Clipped> unclipped.

174

\

\

ECOSCIENCE, VOL. 4 (2),1997

19912.5 D • NUT·..·CLp..

2.0

1.5

1.0

0.5

o _'L...----I_

o I

-CLP +CLP -CLP +CLP

-NUT +NUT

I

19932.5 D CLP'

• NUT···· 2.0

1.5

1.0

0.5

o I

-CLP +CLP -CLP +CLP

·NUT +NUT

FIGURE 3. The effects of clipping (± CLP) and fertilization (± NUT) on nitrogen content of aboveground shoots sampled peak season 1991-1993. Abbreviations follow Figure 1.

trends. For example, clipping reduced root growth and root:shoot ratios more when plants were grown on high salinity (- 18% and - 26%, respectively) compared to when plants were grown at low salinity (- 7%, and - 23%, respec­tively), The result of these responses was that biomass allo­cation to offtake biomass was less sensitive to salinity in

120 1991 0 NUT' CLPNUT*

110 1992 • NUT'

~ 100 QI ....

~-c 90 ...... =.:.. 80 .g ~ 70

60,~~Q.. 50'" ....QlO 40.:U's 300.0 VJ~

'-' 20 10 o

1993 • NUT'

-CLP +CLP -CLP +CLP

-NUT +NUT

FIGURE 4. The effects of clipping (± CLP) and fertilization (± NUT) on soil respiration rates during 1991-1993. Abbreviations follow Figure I.

clipped plants (- 16%, p < 0.05) than in unclipped plants (- 28%, P < 0.001). These interactive responses to clipping and salinity were, in general, independent of N treatment.

NITROGEN UPTAKE AND ALLOCATION

Nitrogen availability accounted for the largest increases in tissue N concentrations and total plant N uptake (TPNUP) (Tables I and II). Clipping had substantial effects also, significantly increasing N concentrations in offtake biomass and TPNUP compared with corresponding values for unclipped plants. These clipping effects were greater when plants were grown on high N, and N-induced increases in both offtake N concentration and TPNUP were greater in clipped plants compared with unclipped plants (both nitro­gen x clipping interactions p < 0.001) (Tables I and 11), Salinity also increased aboveground and belowground N concentrations, and total plant N uptake (all p < 0.001). These concentration increases were offset by reduced plant size, resulting in a 58% reduction in offtake N mass in high salt versus low salt plants. However, increases in plant N uptake and tissue N concentrations were not simply a func­tion of reduced plant size, since plant size, when used as a covariate, did not significantly influence the positive effects of salt on plant N accumulation.

200 ---.-1991 r2 =0.43,p<0.001 -A-1992r2 =0.41,p<0.0011 • --+- -1993 r2 = 0.33, p < 0.01

180

~;;- 160 ~"CI

~'Z. 140

~~ 120 • .~~ 100 .. •• ._..-------­Q..'" ...1:8 80 .... ...... ~~-..~ ..~~'i.! •~~ 60

40 ~-----.\ .. 20 1"""'T'"~""'-"".-- iii

40 45 50 55 60 65 Soil water content (g H20 gdwt"l)*l00

FIGURE 5. Relationship between soil respiration rate and soil water content for 1991-1993.

2.5

~

~ 2.0 '-'-= ~ g 1.5 C.l

= ~ g 1.0

'2 .... ~ 0.5

..:l

-CLP +CLP -CLP +CLP

-NUT +NUT

1992

[~ NUT··..

I

70

175

\

RUESS ET AL.: GROWTH RESPONSES IN CAREX RAMENSKII

,­... -0

~ "'Cl O!l 1Z ­

~ 1991.5 I 1 0

l'-2l 1992 .. *NUT 1993 .. *NUT

-3 ...1'''''""-- _

-CLP +CLP -CLP +CLP -NUT +NUT

FIGURE 6. The effects of clipping (± CLP) and fertilization (± NUT) on net nitrogen mineralization rates during 1991-1993. Abbreviations follow Figure 1.

Na+ AND K+ CONTENT

Salinity markedly increased leaf Na+ content (p < 0.0001), and decreased, but to a lesser degree, leaf K+ content (p < 0.0001; Tables I and II). Clipping decreased leaf Na+ content (- 95%, p < 0.0001), and increased leaf K+ content (+ 27%, P < 0.0001). High N also decreased leaf Na+ content (- 58%, p < 0.0001) and increased leaf K+ content (+ 11 %, P < 0.0001). When grown on high levels of salt, clipped plants maintained lower Na+ concentrations and higher K+ concentrations compared with corresponding values for unclipped plants (both interactions p < 0.0001). High N significantly reduced leaf Na+ content (- 61 %, p < 0.0001) and increased leaf K+ content (+ 17%, P < 0.0001) when plants were grown on high salt. Among all high salt high N plants, clipped plants had significantly lower leaf Na+ concentrations (1.68 ± 0.26 mg g-I) and higher K+/Na+ ratios (26.14 ± 3.36 mg mg-I g-I) compared with unclipped plants (20.70 ± 2.03 mg g-I and 1.42 ± 0.14 mg mg-I g-I, respectively).

Discussion

Our field experiment suggests that intensive, early­season grazing by geese may significantly increase the aboveground production and nitrogen content of C. ramen­skii swards. Relative to control plots, clipped-fertilized plots showed significant increases in AGNPP and leaf N concen­tration, which resulted in significant increases in offtake biomass and offtake N during both 1991 and 1992. When grown in the greenhouse, total production of C. ramenskii clipped every three weeks did not exceed that of unclipped plants, but clipped plants had significantly higher offtake biomass, and biomass and N allocation to offtake compared to unclipped plants. Similar growth responses following goose grazing in the field would suggest that forage quantity could be increased without any net change in primary production, by geese modifying patterns of plant biomass allocation.

Both field and laboratory experiments demonstrate that rapid regrowth of C. ramenskii following defoliation is

dependent on soil nutrient availability. These results are consistent with a large number of other greenhouse (McNaughton & Chapin, 1985; Georgiadis et aI., 1989) and field studies (Bazely & Jefferies, 1985; Maschinski & Whitham, 1989; Hik & Jefferies, 1990). We recognize that our fertilization treatment (10 g N m-2 year-I) exceeded the expected N deposition by geese during a single grazing event. For example, if 65% of the N removed (309.8 g m-2

of early season biomass, containing 1.04% N) was returned in a plant available form (1. Sedinger, unpubl. data), we would expect approximately 2 g N m-2 deposited. Nevertheless, clipping without fertilization substantially increased AGNPP (+ 42%), foliage N content (+ 14%), offtake biomass (+ 36%), and offtake nitrogen (+ 53%) relative to corresponding values for untreated plots during 1991. Although this response was not observed in 1992, AGNPP, offtake biomass, foliage N content, and offtake N were all significantly higher in clipped/fertilized plots compared with plots only fertilized during both 1991 and 1992. These results support the notion that canopy removal alone stimu­lates aboveground growth, and that grazed plants are more capable of responding to added nutrients than ungrazed plants. Mechanisms responsible for these responses have been reviewed by Hik & Jefferies (1990), who found that total aboveground N uptake by grazed swards of Puccinellia phryganodes exceeded the sum of the amount of N returned in feces plus that taken up by ungrazed plants.

In addition to removing older, slower growing tissues, canopy removal in Carex ramenskii swards likely increases light, and in particular, heat flux to emerging shoots. The tendency for greater depth of thaw on grazed plots, and reduced depth of thaw on fertilized plots supports this latter idea. Increased soil heat flux may be particularly important in arctic and subarctic ecosystems, where net N mineraliza­tion rates have been shown to be highly temperature sensi­tive above a given threshold (Nadelhoffer et al., 1991). Although higher soil temperatures may reduce soil moisture content, and thus soil microbial processes later in the season (Figure 5), we suspect soil moisture does not limit microbial processes in early spring, and that increased soil heat flux following canopy removal by geese may notably inqease soil nutrient turnover.

Fertilization increased soil respiration rates each year, but tended to decrease rates of net N mineralization, indi­cating that the soil microbial biomass is a strong nutrient sink in this ecosystem. Relative to arctic tundra north of the Brooks Range, Carex ramenskii meadows have an order of magnitude less percentage soil C and lower C:N ratios, but similar respiration rates per gram C. For example, Kielland (1990) found that soils from Carex aquatilis/Eriophorum vaginatum wet meadows (39.4% soil C, C:N ratio = 34.6) had respiration rates averaging 65.8 pg CO2-C g soil C-I hour1. In 1993, soil respiration rates from our unfertilized and fertilized plots averaged 49.4 and 76.8 pg CO2.:Cg soil C-I hour-I, respectively. Thus, while Carex ramenskii meadows overlay younger surfaces with less organic matter accumulation, the physical, chemical, and biological constraints controlling soil C turnover are similar to those in arctic tundra. The high ratio of C respiration to net N miner­alization we found, particularly on fertilized plots, suggests

176

that nutrient availability may be more important than substrate quality in regulating soil microbial processes. Using a buried bag method, Giblin et al. (1991) also found negative net N mineralization during the growing season (July-August) in several arctic ecosystems, including wet sedge tundra. Primary production is strongly limited by nutrient availability in Carex ramenskii meadows, as indi­cated by the dramatic response to fertilization. For example, in 1991, 83% of the N added to unclipped plots was recovered in aboveground biomass at the end of the season. If our N mineralization rates are representative of mineral N production in the field, then clearly other sources of N would be required to meet plant demand. It is likely that direct uptake of amino acids by Carex ramenskii plays an important role in N cycling in these ecosystems, as has been shown for Carex aquatilis and Carex bigelowii growing in arctic wet meadow and tussock tundra (Kielland, 1990; 1994).

We found that higher N concentrations in both shoots and roots, and higher total plant N uptake of high salt plants were due to factors other than simply reduced growth relative to N uptake, suggesting nitrogenous osmotic solutes may be important for osmoregulation and/or enzyme osmo­protection in this species (Blits & Gallagher, 1991; Hasegawa et aI., 1994). Because these compounds often constitute a significant portion of the total nitrogen budget (Jefferies & Rudmik, 1991), increased plant nitrogen uptake or changes in nitrogen allocation patterns are often required. This may partially explain why clipping, which increased plant N uptake and tissue N concentrations, also increased plant water potential. Of particular relevance to herbivory was the response of offtake biomass to added N, which more than doubled in high salt plants (p < 0.05). Throughout western and southwestern Alaska, C. ramenskii occupies moderately saline marshes and wet meadows (Snow & Vince, 1984; Vince & Snow, 1984; Kincheloe & Stehn, 1991; Babcock & Ely, 1994). Given the high nitrogen demands associated with N-based osmoregulation, the nitrogen recycled through goose feces may be an important source of nitrogen con­tributing to salinity tolerance in C. ramenskii throughout these coastal saltmarsh regions. We suspect that such an indirect effect of geese on plant salt tolerance may also contribute to the observed growth stimulation of the more salt-tolerant and more heavily grazed species Carex sub­spathacea within dense grazing lawns along the Bering Sea coast (Person, unpubl. data) and in eastern Canadian salt marshes (Hik & Jefferies, 1990).

In addition to the potential importance of organic solute accumulation, Carex ramenskii may mediate osmotic adjustment through changes in ion accumulation. Elevated leaf K+/Na+ is one mechanism associated with reducing ion toxicity in salt-tolerant species, although it is not entirely clear whether K+/Na+ selectivity occurs within the root or at the level of the leaf plasmalemma, or both. Nor is it under­stood if such selectivity is mediated by increased K+ and/or reduced Na+ influx, or how the process affects ion compart­mentation within the leaf (Hasegawa et al., 1994). We found that defoliation and fertilization independently contributed to elevated leaf K+/Na+ concentrations, and that when grown on high salt, the highest values were found in

ECOSCIENCE, VOL. 4 (2),1997

.-/

plants that were both clipped and fertilized. Given that greater soil surface evaporation may exacerbate soil salin­ization on grazed swards (Srivastava & Jefferies, 1995), physiological adjustments in mechanisms for salt tolerance are likely essential for maintaining rapid rates of growth by C. ramenskii when grazed.

While plant communities on the Y-K Delta occupy topographic zones with distinct edaphic characteristics (Kincheloe & Stehn, 1991; Bab<;:ock & Ely, 1994), grazing by geese strongly influences the dominance among Carex species within a community. For example, within the C. subspathacea grazing lawns surrounding inland ponds, exclusion of geese for 2 years resulted in complete domi­nance of the sward by C. ramenskii (J. Sedinger, unpubl. data). In contrast, when a C. ramenskii sward is subjected to frequent, intensive grazing, C. subspathacea can become dominant within a growing season (pers. observ.). During spring snowmelt, exposed islands of C. ramenskii are often grazed heavily by early-arriving breeding and non-breeding Pacific black brant geese. The present experiments suggests such intensive, but infrequent defoliation may improve forage quantity and quality to geese, particularly if accompanied by increased soil nitrogen availability. Such plant responses to grazing likely explain why these patches are in many instances regrazed throughout the growing season. Moreover, tolerance of C. ramenskii to salinized soils, resulting from greater surface evaporation on trampled surfaces defoliated by geese, may be linked to improved plant nitrogen balance derived from goose feces.

The number of nesting pairs of black brant at the Tutakoke River colony increased dramatically between 1986 and 1993 (Sedinger et al., 1993; 1994). Our field observations and results from the present experiments indicate that increased grazing pressure on patches of C. ramenskii may provide high rates of biomass and N offtake to geese, which in tum may lead to further grazing and the eventual conversion to more preferred C. subspathacea habitat. This apparent pattern of vegetation change has important consequences for goose population dynamics at the landscape scale, and offers an interesting contrast to the coastal lowlands of west Hudson Bay, where grubbing by lesser snow geese on the roots and rhizomes of graminoid plants causes irreversible destruction of grazing habitat through erosion and hypersalinization of devegetated sediments (Kerbes, Kotanen & Jefferies, 1990).

Acknowledgements

We wish to thank K. Barber and L. Oliver at the UAF Forest Soils Lab for assistance in laboratory analyses. R. L. Jefferies and two anonymous reviewers provided many helpful suggestions to the manuscript. This research was funded by the Department of Biology and Wildlife at the UAF and by NSF OPP 92-14970.

Literature cited

Babcock, C. A. & C. R. Ely, 1994. Classification of vegetation communities in which geese rear broods on the Yukon­Kuskokwim delta, Alaska. Canadian Journal of Botany, 72: 1294-1301.

Bazely, D. R. & R. L. Jefferies, 1985. Goose faeces: A source of

177

RUESS ET AL.: GROWTH RESPONSES IN CAREX RAMENSKII

nitrogen for plant growth in a grazed salt marsh. Journal of Applied Ecology, 22: 693-703.

Bazely, D. R& R L. Jefferies, 1989. Lesser snow geese and the nitrogen economy of a grazed salt marsh. Journal of Ecology, 77: 24-34.

Blits, K. C. & J. L. Gallagher, 1991. Morphological and physio­logical responses to increased salinity in marsh and dune eco­types of Sporobolus virginicus (L.) Kunth. Oecologia, 87: 330­335.

Cargill, S. M. & R. L. Jefferies, 1984a. Nutrient limitation of primary production in a sub-arctic salt marsh. Journal of Applied Ecology, 21: 657-668.

Cargill, S. M. & R L. Jefferies, 1984b. The effects of grazing by lesser snow geese on the vegetation of a sub-arctic salt marsh. Journal of Applied Ecology, 21: 669-686.

Georgiadis, N. J., R. W. Ruess, S. J. McNaughton & D. Western, 1989. Ecological conditions that determine when grazing stimulates grass production. Oecologia, 81: 316-322.

Giblin, A. E., K. J. Nadelhoffer, G. R Shaver, J. A. Laundre & A. J. McKerrow, 1991. Biogeochemical diversity along a riverside toposequence in arctic Alaska. Ecological Monographs, 61: 415-435.

Hasegawa, P. M., R A. Bressan, D. E. Nelson, Y. Samaras & D. Rhodes, 1994, Tissue culture in the improvement of salt tolerance in plants. Pages 83-125 in A. R Yeo & T. J. Flowers (ed.). Soil Mineral Stresses, Approaches to Crop Improvement. Monographs on Theoretical and Applied Genetics, Volume 21. Springer-Verlag, Berlin.

Hik, D. S. & R. L. Jefferies, 1990. Increases in the net above­ground primary production of a salt-marsh forage grass: A test of the predictions of the herbivore-optimization model. Journal of Ecology, 78: 180-195.

Jefferies, R L, 1988. Vegetational mosaics, plant-animal interac­tions and resources for plant growth. Pages 340-361 in L. D. Gottliev & S. K. Jain (ed.). Plant Evolutionary Biology. Chapman & Hall, London.

Jefferies, R L. & T. Rudmik, 1991. Growth, reproduction and resource allocation in halophytes. Aquatic Botany, 39: 3-16.

Kerbes, R H., P. M. Kotanen & R L. Jefferies, 1990. Destruction of wetland habitats by lesser snow geese: A keystone species on the west coast of Hudson Bay. Journal of Applied Ecology, 27: 242-258.

Kielland, K, 1990. Processes controlling nitrogen release and turnover in arctic Alaska. Ph.D. Dissertation, University of Alaska, Fairbanks, Alaska.

Kielland, K, 1994. Amino acid absorption by arctic plants:

Implications for plant nutrition and nitrogen cycling. Ecology, 75: 2373-2383.

Kincheloe, K. L. & R. A. Stehn, 1991. Vegetation patterns and environmental gradients in coastal meadows on the Yukon­Kuskokwim delta, Alaska. Canadian Journal of Botany, 69: 1616-1627.

Maschinski, J. & T. G. Whitham, 1989. The continuum of plant responses to herbivory: The influence of plant association, nutrient availability, and timing. The American Naturalist, 134: 1-19. .

McNaughton, S. J. & F. S. Chapin, III, 1985. Effects of phosphorus nutrition and defoliation on C4 graminoids from the Serengeti plains. Ecology, 66: 1617-1629.

Nadelhoffer, K. J., A. E. Giblin, G. R. Shaver & J. L. Laundre, 1991. Effects of temperature and substrate quality on element mineralization in six arctic soils. Ecology, 72: 242-253.

Ruess, R.W., 1988. The interaction of defoliation and nutrient uptake in Sporobolus kentrophyllus, a short-grass species from the Serengeti plains. Oecologia, 77: 550-556.

Ruess, R. W., D. H. Hik & R. L. Jefferies, 1989. The role oflesser snow geese as nitrogen processors in a sub-arctic salt marsh. Oecologia, 79: 23-29.

SAS, 1985. SAS User's Guide: Statistics. Version 5 ed. SAS Institute Inc., Cary, North Carolina.

Sedinger, 1. S., C. 1. Lensink, D. H. Ward, R M. Anthony, M. L. Wege & G. V. Byrd, 1993. Current status and recent dynamics of the black brant Branta bernicla breeding population. Wildfowl, 44: 49-59.

Sedinger, J. S., D. H. Ward, R M. Anthony, D. V. Derksen, C. J. Lensink, & K. S. Bollinger, 1994. Management of Pacific brant: Population structure and conservation issues. Pages 50­62 in Transactions of the 59th North American Wildlife and Natural Resources Conference. Wildlife Management Institute, Washington D.C.

Snow, A. A. & S. W. Vince, 1984. Plant zonation in an Alaskan salt marsh. II. An experimental study of the role of edaphic conditions. Journal of Ecology, 72: 669-684.

Srivastava, D. S. & R L. Jefferies, 1995. Mosaics of vegetation and soil salinity: A consequence of goose foraging in an arctic salt marsh. Canadian Journal of Botany, 73: 75-83.

Tande, G. F. & T. W. Jennings, 1986. Classification and mapping of tundra near Hazen Bay, Yukon Delta National Wildlife Refuge. Unpublished Report, U.S. Fish and Wildlife Service.

Vince, S. W. & A. A. Snow, 1984. Plant zonation in an Alaskan salt marsh. Journal of Ecology, 72: 651-667.

178