Color profile: Disabled Composite Default screen

25

Influences of gap microheterogeneity on the regeneration of Nothofagus pumilio in a xeric old-growth forest of northwestern Patagonia, Argentina Karin Heinemann, Thomas Kitzberger, and Thomas T. Veblen

Abstract: We experimentally examined the influences of within-gap environmental heterogeneity on regeneration patterns of Nothofagus pumilio (Poepp. & Endl.) Krasser near the xeric limit of its distribution in northern Patagonia, Argentina. Results from this xeric old-growth forest are compared with patterns previously described for the same species in mesic forests. Survival of N. pumilio seedlings beneath tree-fall gaps in this relatively xeric forest appears to be strongly influenced by moisture availability. Seedlings and saplings that have survived this demographic bottleneck are found at microsites where soil water potentials are higher, such as in the shady northern edges of tree-fall gaps (Ψ = –0.46 MPa compared with less than –0.6 MPa in other gap positions) and on coarse woody debris (Ψ = –0.29 MPa, compared with –0.51 MPa on the forest floor). Although gap creation in this dry N. pumilio forest is favorable to tree regeneration by releasing light resources, decreased water resources may switch the system from a light- to a waterlimited system in some positions of the gap. This may explain the lack of regeneration of N. pumilio often observed after creation of large gaps towards the xeric end of its range and needs to be considered in the management of this important timber species. Résumé : Les auteurs ont étudié expérimentalement l’influence de l’hétérogénéité environnementale à l’intérieur des ouvertures sur les patrons de régénération du Nothofagus pumilio (Poepp. & Endl.) Krasser près de la limite xérique de sa répartition, dans le nord de la Patagonie, en Argentine. Les résultats provenant de cette vieille forêt xérique sont comparés avec les patrons décrits auparavant pour la même espèce dans des forêts mésiques. La survie des semis de N. pumilio, dans les ouvertures créées par la chute des arbres dans cette forêt relativement xérique, semble fortement influencée par la disponibilité en eau. Les semis et les gaules qui ont survécu à ce goulot d’étranglement démographique se trouvent sur les microsites où le potentiel en eau du sol est plus élevé, comme c’est le cas dans les bordures ombragées au nord des ouvertures créés par la chute des arbres (Ψ = –0,46 MPa comparativement à moins de –0,6 MPa ailleurs dans les ouvertures), ainsi que sur les débris ligneux grossiers (Ψ = –0,29 MPa comparativement à –0,51 MPa sur le parterre de la forêt). Bien que la création d’ouvertures dans cette forêt sèche de N. pumilio soit favorable à la régénération des arbres en procurant plus de lumière, les ressources limitées en eau peuvent modifier le système à certains endroits dans les ouvertures, en passant d’un système limité par la lumière à un système limité par l’eau. Ceci peut expliquer le manque de régénération de N. pumilio souvent observé après la création de grandes ouvertures, près de la limite xérique de son aire, et la nécessité d’en tenir compte dans l’aménagement de cette importante espèce forestière. [Traduit par la Rédaction]

Heinemann et al.

31

Introduction In forests where large-scale disturbances are rare, stand dynamics are controlled by the creation of gaps due to single or multiple overstory-tree mortality events (Brokaw 1985; Runkle 1985; Veblen 1992). Although the role of small treefall gaps in forest dynamics has been an important research Received November 26, 1998. Accepted August 29, 1999. K. Heinemann1 and T. Kitzberger. Departamento de Ecología, Universidad Nacional del Comahue, Quintral 1250, 8400 Bariloche, Argentina. e-mails: [email protected] and [email protected] T.T. Veblen. Department of Geography, University of Colorado, Campus Box 260, Boulder, CO 80309, U.S.A. e-mail: [email protected] 1

Corresponding author.

Can. J. For. Res. 30: 25–31 (2000)

I:\cjfr\cjfr30\cjfr-01\X99-181.vp Monday, February 07, 2000 2:00:22 PM

theme in many forests worldwide (see, for example, the special feature in Ecology 70(3) 1989), the paradigm of finescale gap dynamics has been studied primarily in mesic forests, where the predominant environmental limitation to successful tree regeneration is presumed to be sufficient solar radiation (Veblen 1992). Nevertheless, studies of tree-fall gap dynamics have demonstrated significant variation in both within-gap environmental influences on tree regeneration and species’ responses to such variation in relation to environmental gradients and forest type (Poulson and Platt 1989; Kneeshaw and Bergeron 1998). Even for the same tree species, responses to gaps may vary markedly depending on associated understory species and the abiotic factors of a particular site, as well as position along broad-scale environmental gradients (Spies and Franklin 1989; Veblen 1989). Given such geographical variation in the role of treefall gaps in forest dynamics, it is important to consider the potential for changes in the mechanisms and net effects of © 2000 NRC Canada

Color profile: Disabled Composite Default screen

26

tree-fall gaps in the regeneration of tree species, distributed across gradients from mesic to xeric forests. Several sources of within-gap heterogeneity can dramatically affect the availability of microsites for successful establishment and growth to the canopy (Veblen 1992). These include (i) gap size, shape, and orientation, which strongly influence the microclimate of gaps (Poulson and Platt 1989; Canham et al. 1990; Dahir and Lorimer 1996); (ii) withingap heterogeneity of underlying edaphic and topographic gradients, which may create a mosaic of different microsites (Beatty 1984; Schaetzl et al. 1989; Peterson et al. 1990); (iii) characteristics of the gap-creating species, which may influence the initial edaphic and topographic patterns as well as litter properties, and allelopathic interactions (Beatty and Scholes 1988; Harmon and Franklin 1989; Boettcher and Kalisz 1990); (iv) coarse woody debris, which can be important in creating safe sites for seedling establishment by reducing inhibition by understory species, preventing litter burial, or creating more favorable moisture and nutrient conditions in comparison with forest floor sites (June and Ogden 1975; Christy and Mack 1984; Harmon and Franklin 1989); and (v) understory species that survive the gap or nonarboreal gap colonists may greatly influence tree species establishment in gaps (Huenneke 1983; Taylor and Zisheng 1988). Different growth rates of juvenile trees have been associated with different positions within the same gap because of within-gap spatial heterogeneity associated with a variety of combinations of the physical and biotic parameters (Wayne and Bazzaz 1993; Runkle et al. 1995). The effects of within-gap heterogeneity on tree regeneration processes have been extensively studied in mesic temperate and tropical forests, but studies conducted near the xeric limit of continuous forests are rare. In this study, we examine the influences of within-gap environmental heterogeneity on regeneration patterns of Nothofagus pumilio (Poepp. & Endl.) Krasser near the xeric limit of its distribution and compare our results with patterns previously described for the same species in mesic forests. Nothofagus pumilio is a widespread deciduous tree that occurs from 35°35′ to 55°S and from near sea level (at high latitude) to ca. 2000 m elevation in southern Chile and Argentina (Veblen et al. 1996b). At 41°S, mean annual precipitation in monotypic N. pumilio forests ranges from over 5000 mm in the west (Chile) to less than 800 mm near its eastern limit in Argentina. Among the nine South American Nothofagus species it is the most widely used species for timber production, and on Tierra del Fuego it is the target of an extremely large proposed timber harvesting scheme (Arroyo et al. 1996). The importance of small tree-fall gaps for the regeneration of N. pumilio has been widely documented in old-growth mesic forests from the mid-latitudes (ca. 39°S) to the high latitudes of Tierra del Fuego (Schlegel et al. 1979, Veblen et al. 1981; Schmidt and Urzúa 1982; Rusch 1992; Rebertus and Veblen 1993). However, near its drier limit at both mid-latitudes and high latitudes, regeneration of N. pumilio often fails after creation of large gaps by clearing and (or) burning of forests (Roig et al. 1985; Veblen et al. 1996a). Although some of these regeneration failures can be attributed to inadequate seed dispersal or inhibition by herbivores, many appear to be related to unfavorable microenvironments created by gaps. In the present study we

Can. J. For. Res. Vol. 30, 2000

examine N. pumilio seedling establishment patterns in relation to within-gap heterogeneity to assess the relative importance of environmental factors controlling the regeneration of this species near the xeric limit of its distribution.

Study site Data were collected in Chall-Huaco Valley (Nahuel Huapi National Park, 41°S, 71°W, 1200 m above sea level), Argentina, which is a disjunct low outlier of the Andes Mountains. Mean annual precipitation is 800 mm of which only 12% falls during the summer months of January through March (Bariloche Airport climate station, ca. 12 km from the study site, located at similar longitude). The study site is at the boundary with the Patagonian steppe, which indicates very dry conditions for a forest. Most of the precipitation falls as snow, from May to September. Mean annual precipitation increases sharply towards the Andes, so that only 40 km to the west, mean annual precipitation is well over 2000 mm (Barros et al. 1983). Soils in the study area are derived from aeolian volcanic ash deposits. In this single-species, old-growth forest of N. pumilio, the dominant trees are up to 26–28 m tall and 80–120 cm in diameter at breast height (DBH). Increment core sampling of the ages of more than 100 trees indicate maximum tree ages of at least 280 years (and perhaps substantially greater, since the oldest trees had rotten centres). The dominant understory species is Alstroemeria aurea, a rhizomatous forb that forms dense thickets up to 1.20 m tall and covering approximately 60% of the forest floor in midsummer. Sampling was restricted to areas that lacked any history of logging or recent (i.e., past 100 years) stand-replacing natural disturbances. Although the sites may have been significantly affected by livestock several decades ago, this area has shown extremely low impacts from livestock since our observations began here in 1985.

Materials and methods Characterization of within-gap heterogeneity After extensive reconnaissance, 15 spatially discrete tree-fall gaps were selected for sampling within-gap heterogeneity. All gaps sampled had a northerly aspect and were on gentle slopes (<10°). In each gap, canopy gaps and expanded gaps (sensu Runkle 1982) were sketched and measured using polar coordinates from a central reference point. Areas were calculated by fitting the gap perimeters to ellipses. Measurements of understory vegetation, snow depth, and soil seed content were made at 3-m intervals along two perpendicular transects in each gap. The perpendicular transects ran north–south and east–west and intersected at midpoints at the gap centre. Soil volumes of 10 cm × 10 cm × 5 cm (depth) were collected in January 1997 to quantify by directly counting the amount of seeds dispersed to different points in the gaps. Mean height and cover of A. aurea were estimated in quadrats of 0.25 m2 in December 1996. Snow depth was measured to the nearest 10 cm in spring of 1996 and 1997. Soil samples for measurements of water potential were taken at a depth of 10 cm, including humus layer but not litter, at nine positions in each gap: N, S, W, and E at the edges of the expanded gap, at the intersection of the transects with the projected gap periphery, and in the center. Soil water potential was measured with WESCOR dew point psycrometers in spring (October 1996), midsummer (January 1997), and late summer (March 1997). A single gap of typical size and aspect was selected to examine the spatial variation of solar radiation within a gap. Photosynthetic active radiation (PAR) measurements were made at the northern and southern edges of the north–south transect and in the gap centre. Data were collected using LI-COR quantum sensors (LI-190) and dataloggers (LI-1000) on a cloudless midsummer day. © 2000 NRC Canada

I:\cjfr\cjfr30\cjfr-01\X99-181.vp Monday, February 07, 2000 2:00:22 PM

Color profile: Disabled Composite Default screen

Heinemann et al.

27

Fig. 1. (a) Mean (±SE) water potential values during the growing season for nine positions along the north–south and east–west transects in four gaps: the periphery of the expanded gap (the external circle), the periphery of the canopy gap (the internal circle) and the centre. (b) Mean soil water potential (solid dots) and standard deviation values for spring, midsummer and late summer, corresponding to all nine positions in four gaps, as shown in Fig. 1a. Differences are statistically significant for the three dates (Mann– Whitney U test, p < 0.005). Three individual values (open circles) of soil water potential are shown for March to illustrate extremes corresponding to microsites with woody debris.

Seedling microsite assessment In 31 gaps, which included the 15 selected for sampling withingap heterogeneity, seedlings (<50 cm tall) and saplings (>50 cm tall but <5 cm DBH) of N. pumilio were counted, and their positions in quadrants within the gap were recorded. Beneath a representative closed canopy, 50 seedlings in a 30 × 30 m plot were monitored for survival over 1 year. Substrate (soil or log), degree of decomposition of logs (on a scale from 1 to 4 of least to maximum decomposition), and height of the rooting surface above ground level were recorded for each seedling. Density and mean height of A. aurea were measured in 1-m2 plots centred at each seedling and at 50 random points. The following summer (January 1997), soil water potential and instantaneous PAR (expressed as the percentage of PAR received at an adjacent open site) at ground level were measured at microsites corresponding to surviving seedlings and to an equal number of random points (n = 30).

Understory removal experiment Five gaps among the 15 sampled were selected according to similar size and orientation for experimental removal of the understory vegetation. Pairs of plots were established at the northern edge, the southern edge, and in the centre of each gap along the north–south transect. Each pair consisted of a 1-m² plot with all its understory removed by clipping aboveground parts and an adjacent 1-m² unclipped control plot. The plots were 50 cm apart. Repeated cuttings were required because of regrowth of A. aurea from rhizomes. At the beginning of fall (April) 1996, ten naturally estab-

lished seedlings of N. pumilio (<10 cm tall) were transplanted to each treatment plot and its adjacent control plot. Seedlings were transplanted in fall, so that they could acclimate during the wet season. Only the seedlings that were still alive in the spring (75%) were included in the study. Instantaneous light measurements were taken at five systematically located points in each 1-m2 plot and repeated twice during a 2-h period in the middle of a cloudless day during late summer. Seedling survival was recorded once a month during the following growing season, from October 1996 to April 1997.

Results Characterization of within-gap heterogeneity Canopy gap area and extended gap area were 95 ± 51 m2 (mean ± SD) and 413 ± 134 m2, respectively. During spring the mean height of A. aurea was slightly less in gap centres and intermediate positions (23.1 ± 8.5 cm) than sites beneath the canopy edge surrounding the gaps (27.6 ± 8.9 cm; p = 0.06). This tendency towards taller A. aurea at the gap perimeter coincides with the pattern of snow accumulation; snow was significantly deeper in the centre of the gap and at the southern and eastern edges (31.5 ± 9.0 cm) than beneath the surrounding canopy (15.4 ± 9.7 cm; p < 0.001). The mean number of N. pumilio seeds found in soil samples collected along the north–south and east–west transects © 2000 NRC Canada

I:\cjfr\cjfr30\cjfr-01\X99-181.vp Monday, February 07, 2000 2:00:24 PM

Color profile: Disabled Composite Default screen

28 Fig 2. Frequency of seedlings (<50 cm) and saplings (>50 cm, <5 cm DBH) in each of the quadrants of all gaps (χ2 = 42.6 for saplings, χ2 = 63.7 for seedlings; p < 0.001 in both cases). Total number of seedlings was 68 in 9 gaps, and total number of saplings was 93 in 31 gaps.

Can. J. For. Res. Vol. 30, 2000 Table 1. Instantaneous readings of PAR (photosynthetic active radiation), measured along the north–south transect of gaps on clear days. Position in the gap

Midsummer PAR (µmol·m–2·s–1)a

Late-summer PAR (µmol·m–2·s–1)b

Northern edge Centre Southern edge

81±86 466±736 130±269

228±384 142±265 786±670

Note: Values are means ± SD. Note that conditions of maximum radiation shifted from gap centres in midsummer to southern edges in late summer. a Midsummer PAR averages correspond to measurements during a 4.5-h period (11:00–15:30). b Late-summer averages of four gaps correspond to instantaneous readings repeated twice during a 2-h period (12:00–14:00).

in gaps was 40 seeds/sample (4000/m2), with high standard deviations along the transects. Seed numbers were significantly lower in canopy gaps (27.5 ± 13.9 seeds/sample; n = 34) compared with sites beneath the closed canopy (47.7 ± 22.9 seeds/sample; n = 46; Mann–Whitney U test, p < 0.0001). In October, when the snow cover has recently melted, mean water potential was similar for all gap positions. During the dry spring and summer, different gap positions demonstrate substantial differences in moisture availability (Fig. 1a). The most negative values of growing-season soil water potential occurred in the intermediate eastern positions of the gaps in March (Fig. 1a). Other positions in gaps (N, E, and S edges and centres) also showed low values of soil water potential at the end of summer. In contrast, at intermediate positions towards the northern and western edges of the gaps, mean water potential remained relatively constant from January to March (Fig. 1a). Measurements of soil water potential indicate that microsite differences in water availability become more extreme later in the growing season as moisture deficits become more intense. In the spring (October), there was little difference in soil moisture among microsites in gaps as reflected by low standard deviations (Fig. 1b). Variability in soil moisture increased in midsummer (January), and peaked in March (Fig. 1b). Water potential values decreased from spring to the end of summer and were significantly different when comparing spring with midsummer values, as well as midsummer to late-summer values (p < 0.005, Mann–Whitney U test). The high standard deviation of water potential in March for all gap positions combined (Fig. 1b) reflects microsite variability in water-holding capacity.

In midsummer, PAR was higher in the centre of gaps than at the northern or southern edges of the gap (t test, p < 0.0001 for both comparisons; Table 1), as indicated by light measurements. Midsummer PAR was also higher along the southern edge than along the northern edge of the gap (t test, p < 0.005), as expected for a gap in the southern temperate latitudes. In contrast, at the end of the growing season (March), instantaneous readings of PAR showed that the southern edges of gaps received significantly more light than gap centres and the northern edges (Mann–Whitney U test, p < 0.01 and p < 0.05 respectively; Table 1). Gap centres and northern edges did not differ significantly in the amount of light measured in late summer. Seedling microenvironments In tree-fall gaps, seedlings and saplings were more frequent in the northwestern and northeastern quadrants of gaps (χ2 = 63.7, p < 0.001 and χ2 = 42.6, p < 0.001 for seedlings and saplings respectively; Fig. 2). The 50 monitored N. pumilio seedlings under a closed canopy were more frequent on decaying logs than on the randomly selected forest floor sites (χ2 = 54.2, p < 0.001). Height and density of A. aurea were significantly lower in the 1-m2 plot centred on the N. pumilio seedlings (40 ± 16 cm and 10 ± 5 individuals/m2) than at the randomly selected points (58 ± 14 cm and 18 ± 9 ind./m2; Mann–Whitney U test, p < 0.001 in both cases). Soil water potential values measured at microsites under closed canopy were not significantly different for seedling microsites and random points. However, decaying logs had significantly higher values of water potential (–0.29 ± 0.12 MPa) than seedling microsites on the forest floor (–0.51 ± 0.13 MPa; Mann–Whitney U test, p < 0.001). Percentage open sky PAR at seedling microsites was significantly higher (2.36 ± 7.52%) than at random points (0.94 ± 2.01%; Mann–Whitney U test, p < 0.001), but no significant differences were found between percent open sky PAR measured on logs and soil microsites. Gap microsite and understory influence on seedling survival In five experimentally manipulated gaps, where understory vegetation was removed in treatment plots, 46% of the seedlings in the control plots and 42% in the treatment plots survived one growing season. In general, survival was highest towards the northern and southern gap edges and lowest © 2000 NRC Canada

I:\cjfr\cjfr30\cjfr-01\X99-181.vp Monday, February 07, 2000 2:00:26 PM

Color profile: Disabled Composite Default screen

Heinemann et al.

in gap centres (Fig. 3). Differences in survivorship were statistically significant for all three possible comparisons of treatment plots (p < 0.005 for southern–centre plots, p < 0.0001 for centre–northern plots, and p < 0.05 for southern– northern plots). Most of the mortality in gap centres occurred in mid-summer (December–January), whereas northern and southern edge mortality remained relatively low during the summer and, for southern edges increased towards late summer (March). For control plots, differences in survivorship among locations (Fig. 3) were not statistically significant (p > 0.05). Survival of seedlings was greater in the centres and southern edges of gaps in control plots, whereas in the northern edges of gaps, seedlings survived better in treatment plots. However, in all cases, differences between treatments and controls were not statistically significant (p > 0.05).

29 Fig 3. Survival of seedlings in experimental plots in five gaps in which the understory had been removed (a) or not removed (b) according to plot position within each gap: southern (solid line), central (dotted line), and northern (broken line) position along the a north–south transect within each gap. Survivorship curves were compared pairwise using Gehan–Wilcoxon tests.

Discussion Nothofagus pumilio regeneration Within the range of the tree-fall gap sizes sampled at Chall-Huaco, we did not find evidence that seed dispersal limitations accounted for the relative scarcity of N. pumilio seedlings. Although seeds are dispersed mainly by gravity and wind and most seeds fall beneath the seed-producing tree crowns (Rusch 1987), seed dispersal does not seem to be a limiting factor, as the tree-fall gaps are small or narrow enough to receive adequate numbers of seed from adjacent trees. Despite the high temporal variability both in seed production and viability in N. pumilio, large seed crops occurring at 3- to 5-year intervals appear to guarantee sufficient seed (Schlegel et al. 1979). Results of the understory removal experiment did not indicate inhibition of N. pumilio seedling establishment in tree-fall gaps by the dominant understory plant A. aurea. Survivorship of seedlings in the centres and southern edges of gaps was higher in control plots, indicating that the net effect of the understory cover was positive, probably because of increased soil moisture associated with the shading effects of A. aurea. Positive influences of nonarboreal species on the survival of tree seedlings is not unusual, especially in relatively dry environments (Callaway 1995), but rarely have understory plants in tree-fall gaps been found to have positive influences on tree regeneration. In comparison with studies conducted in more mesic forests, seedling abundances in the relatively xeric forest at Chall-Hauco are low (Veblen et al. 1981; Rebertus and Veblen 1993). Nevertheless, regeneration of N. pumilio in association with small tree-fall gaps allows self-replacement of this species in this old-growth stand. Survival of N. pumilio seedlings beneath these tree-fall gaps appears to be strongly influenced by moisture availability, particularly during the first summer after establishment. Seedlings and saplings that have survived this demographic bottleneck are found at microsites where moisture levels are higher, such as in the shady northern edges of tree-fall gaps. In our study, both the plots where the understory was removed as well as the control plots showed a strong pattern of minimum seedling mortality at northern (i.e., moister) edges of gaps and maximum seedling mortality in gap centres (i.e., drier). Similarly, ca. 40 km west, where mean annual precipitation is

twice as great as that at Chall-Huaco, Rusch (1992) also found that moisture availability was the main limitation to N. pumilio seedling survival in large (>30 m diameter) canopy gaps. Rusch (1992) found that soil moisture decreased steeply from gap edges towards the centre of large gaps (>30 m2), whereas soil moisture increased from the edge towards the centre in small gaps (<10 m2). Rusch (1992) also recorded soil surface temperatures as high as 45°C in the centre of large gaps (40 m2) where ambient air temperature was only 30°C. This high surface temperature is likely accompanied by a significant loss of soil moisture. Consistent with the pattern of moisture-limited seedling survival at Chall-Huaco is the greater abundance of N. pumilio seedlings on microsites of decaying coarse woody debris. Under a closed canopy, the great majority of seedlings were found on decaying logs where superior water retention resulted in greater water availability in late summer. Light levels at seedling microsites were highly variable, which suggests that seedlings may establish under a wider variety of light environments. Once the critical stages of seedling emergence and early survival have been surpassed, it is likely that higher light levels are needed for seedling growth into the sapling size class. This is indicated by the greater abundance of saplings on the perimeters of gaps than beneath closed canopies. In more mesic forests, stem analyses of N. pumilio saplings indicate rapid growth of seedlings © 2000 NRC Canada

I:\cjfr\cjfr30\cjfr-01\X99-181.vp Monday, February 07, 2000 2:00:28 PM

Color profile: Disabled Composite Default screen

30

following gap formation, suggesting that solar radiation levels beneath a closed canopy are too low for growth into the small tree size class (Rebertus and Veblen 1993). This probably also explains the relative abundance of seedlings associated with logs in contrast to the scarcity of saplings beneath closed canopies at Chall-Huaco. In contrast, in temperate rain forest environments where moisture availability is unlikely to be limiting, enhanced seedling establishment on coarse woody debris has been attributed to various other factors such as more favorable light environments, reduced competition, and negative influences associated with fine litter on the forest floor (Christy and Mack 1984, Veblen 1985, Harmon and Franklin 1989). Although gap creation in this dry N. pumilio forest is favorable to tree regeneration by releasing light resources, in some positions of the gap, decreased water resources may switch the system from a light- to a water-limited system. Thus, sapling development is best along the partially shaded edges of gaps. This is an important contrast with mesic N. pumilio forests where sapling abundance tends to be associated with highest light levels (Schlegel et al. 1979; Veblen et al. 1981; Rebertus and Veblen 1993). Given that the results of this study are based on a single year of observations of seedling survival, there are potentially two confounding influences that need to be considered. First, the effects of transplanting seedlings into the experimental site may not be fully manifested in a single year. However, we mitigated for these effects by transplanting at the beginning of the dormant season (fall) when moisture stress is minimal. Furthermore, the 25% mortality of transplanted seedlings that occurred during the dormant season was evenly distributed across treatment plots, whereas the subsequent mortality patterns indicate differential survival in relation to water deficits. Secondly, atypical weather during the experimental period could bias the results. However, during the experiment the 1996–1997 growing season precipitation of 169 mm was near the 1984–1998 mean of 182 mm. Spring precipitation was slightly above average and summer precipitation slightly below average; neither deviated from the 1984–1998 by a standard deviation. Nevertheless, continuation of this experiment may reveal influences of extreme weather on seedling survival. Regeneration of N. pumilio near its eastern limit is much more likely to be affected by climatic variation than in its mesic distribution. Recent studies in northern Patagonia have related periods of no or little regeneration of the conifer Austrocedrus chilensis (D. Don) Pic. Ser. & Bizz to droughts occurring at annual to decadal time scales (Villalba and Veblen 1997; Kitzberger et al. in press). Analogous to N. pumilio, Austrocedrus chilensis occurs near the ecotone with the Patagonian steppe but occupies lower elevations than N. pumilio. The regeneration of both species may be affected by trends towards higher temperatures (and increased summer moisture deficits) in northern Patagonia associated with more frequent warm phases of the El Niño – Southern Oscillation since the mid-1970s (Villalba et al. 1997; Villalba and Veblen 1998). Management implications The spatial heterogeneity of N. pumilio regeneration found in tree-fall gaps in this relatively xeric forest has im-

Can. J. For. Res. Vol. 30, 2000

portant implications for forest management. In mesic N. pumilio forests, small- to moderate-sized gaps (i.e., <400 m2) typically result in abundant regeneration if sites are protected from large herbivores. This observation has been important in developing management strategies of N. pumilio forests that are believed to be applicable across a wide range of habitats but have only been investigated experimentally in relatively mesic forests (Schmidt and Urzúa 1982). Nevertheless, the importance of broad-scale as well as local-scale variation in seedling establishment and seedling banks has also been recognized (Veblen et al. 1996a, 1996b). Our study identified microenvironmental conditions associated with gaps that need to be taken into account in designing silvicultural systems. Management practice in drier N. pumilio forests must consider the potential that large openings create microenvironments too dry for successful regeneration. Even small clearings (<100 m2) can lead to regeneration failures over large parts of the cut area if protection from high solar radiation is inadequate and if coarse woody debris is too scarce to provide more favorable moisture conditions for seedling establishment.

Acknowledgements Research support was provided by Grant No. B036 from Universidad Nacional del Comahue and the National Science Foundation of the United States. We thank P. Alaback for facilitating light measurement equipment; Clemente Arko for providing transportation; and L. Daniels, D. Lorenz, and A. Premoli for commenting the manuscript.

References Arroyo, M.T.K., Donoso, C., Murua, R.E., Pisano, E.E., Schlatter, R.P., and Serey, I.A. 1996. Toward an ecologically sustainable forestry project. Universidad de Chile Press, Santiago. Barros, V., Cordon, V., Moyano, C., Mendez, R., Forquera, J., and Pizzio, O. 1983. Cartas de precipitacion de la zona oeste de las provincias de Rio Negro y Neuquen. Unpublished report. Facultad de Ciencias Agrarias, Universidad del Comahue, Cinco Saltos, Argentina. Beatty, S.W. 1984. Influence of microtopography and canopy species on spatial patterns of forest understory plants. Ecology, 65: 1406–1419. Beatty, S.W., and Scholes, O.D. 1988. Leaf litter effect on plant species composition of deciduous forest treefall pits. Can. J. For. Res. 18: 553–559. Boettcher, S.E., and Kalisz, P.J. 1990. Single-tree influence on soil properties in the mountains of eastern Kentucky. Ecology, 71: 1365–1372. Brokaw, N.V.L. 1985. Treefalls, regrowth, and community structure in tropical forests. In The ecology of natural disturbance and patch dynamics. Edited by S.T.A. Pickett and P.S. White. Academic Press, New York. pp. 53–69. Callaway, R.M. 1995. Positive interactions among plants. Bot. Rev. 61: 306–349. Canham, C.D., Denslow, J.S., Platt, W.J., Runkle, J.R., Spies, T.A., and White P.S. 1990. Light regimes beneath closed canopies and tree-fall gaps in temperate and tropical forests. Can. J. For. Res. 20: 620–631. Christy, E.J., and Mack, R.N. 1984. Variation in demography of juvenile Tsuga heterophylla across the substratum mosaic. J. Ecol. 72: 75–91. © 2000 NRC Canada

I:\cjfr\cjfr30\cjfr-01\X99-181.vp Monday, February 07, 2000 2:00:28 PM

Color profile: Disabled Composite Default screen

Heinemann et al. Dahir, S.E., and Lorimer, C.G. 1996. Variation in canopy gap formation among developmental stages of northern hardwood stands. Can. J. For. Res. 26: 1875–1892. Harmon, M.E., and Franklin, J.F. 1989. Tree seedlings on logs in Picea–Tsuga forests of Oregon and Washington. Ecology, 70: 48–59. Huenneke, L.F. 1983. Understory response to gaps caused by the death of Ulmus americana in central New York. Bull. Torrey Bot. Club, 110: 170–175. June, S.R., and Ogden, J. 1975. Studies on the vegetation of Mount Colenso, New Zealand: 3. The population dynamics of red beech seedlings. Proc. N.Z. Ecol. Soc. 22: 61–66. Kitzberger, T., Steinaker, D.F., and Veblen, T.T. Establishment of Austrocedrus chilensis in Patagonian forest-steppe ecotones: facilitation and climatic variability. Ecology. In press. Kneeshaw, D.D., and Bergeron, Y. 1998. Canopy gap characteristics and tree replacement in the southeastern boreal forest. Ecology, 79: 783–794. Peterson, C.J., Carson, W.P., McCarthy, B.C., and Pickett, S.T.A. 1990. Microsite variation and soil dynamics within newly created treefall pits and mounds. Oikos, 58: 39–46. Poulson, T.L., and Platt, W.J. 1989. Gap light regimes influence canopy tree diversity. Ecology, 70: 553–555. Rebertus, A.J., and Veblen, T.T. 1993. Structure and tree-fall gap dynamics of old-growth Nothofagus forests on Tierra del Fuego, Argentina. J. Veg. Sci. 4: 641–654. Roig, F.A., Anchorena, J., Dollenz, O., Faggi, A.M., and Méndez, E. 1985. Las comunidades vegetales de la transecta botánica de la Patagonia austral. In Transecta botánica de la Patagonia austral. Edited by O. Boelcke, D.M. Moore, and F.A. Roig. Consejo Nacional de Investigaciones Científicas y Técnicas, Buenos Aires, Argentina. pp. 350–519. Runkle, J.R. 1982. Patterns of disturbance in some old-growth mesic forests of eastern North America. Ecology, 63: 1533–1546. Runkle, J.R. 1985. Disturbance regimes in temperate forests. In The ecology of natural disturbance and patch dynamics. Edited by S.T.A. Pickett and P.S. White. Academic Press, New York. pp. 17–33. Runkle, J.R., Stewart, G.H., and Veblen, T.T. 1995. Sapling diameter growth in gaps for two Nothofagus species in New Zealand. Ecology, 76: 2107–2117. Rusch, V. 1987. Estudio sobre la regeneración de la lenga en la Cuenca del Río Manso Superior, Río Negro. Informe final, Beca de iniciación, Conicet. Rusch, V. 1992. Principales limitantes para la regeneración de la lenga en la zona N.E. de su área de distribución. Variables en claros del bosque. In Actas del Seminario de Manejo Forestal de

31 la Lenga y Aspectos Ecológicos Relacionados. Centro de Investigación y Extensión Forestal Andino Patagónico. Esquel, Argentina, Bol. No. 8. Schaetzl, R.J., Burns, S.F., Johnson, D.L., and Small, T.W. 1989. Tree uprooting: review of impacts on forest ecology. Vegetatio, 79: 165–176. Schlegel, F.S., Veblen, T.T., and Escobar, R.B. 1979. Estudio ecológico de la estructura, composición, semillación y regeneración del bosque de Lenga (Nothofagus pumilio) XI Región. Universidad Austral de Chile, Valdivia. Fac. Ing. For. Ser. Tec. Schmidt, H., and Urzúa, A. 1982. Transformación y manejo de los bosques de lenga en Magallanes. Universidad de Chile, Santiago. Cienc. Agric. No. 11. pp. 1–62. Spies, T.A., and Franklin, J.F. 1989. Gap characteristics and vegetation response in coniferous forests of the Pacific Northwest. Ecology, 70: 543–545. Taylor, A.H., and Zisheng, Q. 1988. Regeneration patternsin oldgrowth Abies–Betula forests in the Wolong Natural Reserve, Sichuan, China. J. Ecol. 76: 1204–1218. Veblen, T.T. 1985. Forest development in tree-fall gaps in the temperate rain forests of Chile. Natl. Geogr. Res. 1: 162–183. Veblen, T.T. 1989. Tree regeneration responses to gaps along a transandean gradient. Ecology, 70: 541–543. Veblen, T.T. 1992. Regeneration dynamics. In Plant succession: theory and prediction. Edited by D.C. Glenn-Lewin, R.K. Peet, and T.T. Veblen. Chapman & Hall, London. pp. 152–187. Veblen, T.T., Donoso, C., Schlegel, F.M., and Escobar, B. 1981. Forest dynamics in south-central Chile. J. Biogeogr. 8: 211–247. Veblen, T.T., Donoso, C., Kitzberger, T., and Rebertus, A.J. 1996a. Ecology of southern Chilean and southern Argentinean Nothofagus forests. In Ecology and biogeography of Nothofagus forests. Edited by T.T. Veblen, R.S. Hill, and J. Read. Yale University Press, New Haven, Conn. pp. 293–353. Veblen, T.T., Kitzberger, T., Burns, B.R., and Rebertus, A.J. 1996b. Perturbaciones y regeneración en bosques andinos del sur de Chile y Argentina. In Ecología del bosque nativo de Chile. Edited by J.J. Armesto, M.K. Arroyo, and C. Villagrán. Universidad de Chile Press, Santiago. pp 169–198. Villalba, R., and Veblen, T.T. 1997. Regional patterns of tree population age structures in northern Patagonia: climatic and disturbance influences. J. Ecol. 85: 113–124. Villalba, R., and Veblen, T.T. 1998. Influences of large-scale climatic variability on episodic tree mortality at the forest–steppe ecotone in northern Patagonia. Ecology, 79: 2624–2640. Wayne, P.M., and Bazzaz, F.A. 1993. Birch seedling responses to daily time courses of light in experimental forest gaps and shadehouses. Ecology, 74: 1500–1515.

© 2000 NRC Canada

I:\cjfr\cjfr30\cjfr-01\X99-181.vp Monday, February 07, 2000 2:00:29 PM