Abstract

Photon flux densities > 500 m E m^-2× s^-1 are beyond the light saturation point for net photosynthesis, as determined in the laboratory from field-collected seedlings. Absorption of light >500m E m^-2× s^-1set by seedlings will substantially increase the energy load without corresponding increases in carbon gain. We suggest seedlings are absent from high-light plots because of temperature and water stresses induced by higher irradiances early in the growing season.

Key words: Abies magnifica; California forests; irradiance photon flux density; photosynthesis; red fir; seedling demography; Sierra Nevada Mountains; sunflecks.

Introduction

The importance of the light regime for plant growth in the highly variable environment of understory plants has been recognized in several studies (Evans 1956, Bray 1958, Anderson 1966, Anderson et al. 1969). In the present study we examine the role of light in natural regeneration of Abies magnifica A. Murr. (red fir). Although the effect of variation in irradiance is readily observable in the mosaic of different communities as aspect varies, the effects within a community are less apparent. In some cases, higher light intensities might be expected to favor seedling establishment. Red fir seedlings have been widely reported to occur under small openings in the canopy (Sudworth 1908, Oosting and Billings 1943, Emmingham and Waring 1973, Rundel et al. 1977), and Seidel (1977) found release from shade increased red fir sapling growth. This suggests that low irradiance conditions characteristic of the understory might be inhibitory to red fir seedlings and that some minimal irradiance conditions are required for growth and survival. In other cases, however, higher light might inhibit seedling survival. In the northern hemisphere, the high total insolation on south-facing exposures, especially at midday, may result in higher temperatures and greater water stress relative to other exposure aspects. During the period of initial root penetration of the soil, such stresses might have a significant effect on seedling establishment. These conditions should be especially pronounced in California due to the strongly Mediterranean climate, where seed germination coincides with the onset of annual summer drought.

Red fir is a dominant tree in the upper montane forests of the Sierra Nevada Mountains, southern Cascade Range, Klamath Mountains, and southern Yolla Bolly Mountains of California (Hemphill 1952, Griffin and Critchfield 1972, Rundel et al. 1977, Sawyer and Thornburgh 1977). Previous analysis revealed significant aggregation of young red fir seedlings in an apparently homogeneous understory at our study site in the central Sierra Nevada. Similar patchiness of seedling and sapling regeneration has been recognized in other parts of the red fir range (Pitcher 1981). Our focus was to determine if a correlation existed between the irradiance distribution on the forest floor and the aggregation of seedlings.

Materials And Methods

Description of research site

The environment, flora, and vegetation of the Experimental Forest have been described by Talley (1977), Smith (1978a, b), and Palmer et al. (1983). Briefly, mean annual precipitation is » 1.3 m, >90% of which falls from mid-October to early June, mostly as snow. Average snowpack duration is 192 d and depth is 3.1 m. Mean annual temperature is » 4°C, mean daily maximum July temperature is 28°, and mean daily minimum January temperature is - 10°. The soil is a moderately shallow (» 0.5 m) Entisol in the Ahart soil series, derived from weathered rhyolitic tuff. There is a continuous duff layer ~50 50 mm thick, with scattered limbs and twigs on the surface. More than 95% of the trees in every size class are red fir. Density of overstory trees (>0.2 m dbh) is » 190 per ha; we estimate they are 100-600 yr old. Cover by shrubs and herbs is only » 1-2%, and species richness is low (15 taxa). In comparison to forests described by Oosting and Billings (1943), we found less basal area per hectare of intermediate-age and mature red fir, a greater density of sapling fir, and fewer species of shrubs and herbs, but otherwise the same community physiognomy and overwhelming dominance of red fir in all age categories.
 

Description of study plots at the research site

After permanent marking (July 1981), microenvironmental measures of overstory and understory plant cover, litter depth, soil texture, and soil chemistry (top 0.2 m) were taken. Soil texture, pH, electrical conductivity, cation exchange capacity, and concentrations of soil cations and nitrate were determined by the University of California Agricultural Extension Service in Davis. All seedlings were counted and aged, and the presence of other species noted. Seedlings were aged by counting the terminal bundle scars, previously found to have a high correlation with tree rings.

Seven evenly dispersed locations within each plot were chosen as permanent sample points for recording light and temperature data. Within S plots, the points were located adjacent to 1-yr-old seedlings.

During the first summer we noted that seedlings were more abundant and more continuously distributed on north than on south slopes. Therefore, in 1982 we established thirty 5 x 5 m contiguous plots along five transects where seedling density varied on a north exposure. These plots were located at the same elevation as those on the south slope and about I km from them, within the same continuous stand of red fir. Here, transect locations were subjectively chosen to cross areas where seedling densities varied. Subsequently, 14 plots were selected to span the range of canopy cover and seedling densities, including 3 having high and 3 low numbers of seedlings and 8 having intermediate seedling densities.
 

Dates of measurements

Monitoring of temperature and soil moisture

Monitoring of photosynthetically active radiation and total irradiance

An additional, integrated measure of irradiance was obtained during both summers by use of the ozalid paper method. Peak spectral sensitivity for this technique is 410 nm. Stacks (20 sheets) of ozalid paper (Dietzgen) in seven covered Petri dishes with 100-mm² apertures were placed in each plot pair and collected weekly, following the method of Friend (1961).
 

Monitoring of sunfleck size

Description of overstory canopy cover

Measurement of seedling net photosynthesis

Net photosynthesis was measured with an open gas exchange system described elsewhere (Pearcy and Ustin 1984). An attached shoot was placed inside a glass-windowed, temperature-controlled chamber. Vapor pressure and CO₂ concentration within the chamber were controlled with a dew point condenser and a Wosthoff G 27 mixing pump. The air within the chamber was thoroughly mixed. Light was supplied from a 1500-W metal halide lamp, and irradiance was varied with aluminum screens of different mesh size. Light was measured with a photosynthetic photon flux density (PPFD)-calibrated silicon cell mounted on the chamber lid. CO₂ concentrations were determined with a differential infrared gas analyzer (Beckman Instruments, model 865), and water vapor concentrations with a relative humidity probe (Weathermeasure, model HMP 11). During the experiments, needle temperatures were maintained at 25.5° ± 1°, vapor pressure deficits at < 1.0 kPa, and ambient partial pressure of CO₂ at 37 ± 2 Pa. Fir needles were later removed and dried. Net photosynthesis was expressed on a dry mass basis.

Results

Microenvironment of the south aspect study plots

Nearly half of all seedlings in both S and NS plots in 1981 were estimated to be in their 1st yr (Fig. 1). There was a nearly continuous distribution of seedling ages in the S plots but not in the NS plots: plot INS contained only three lst-yr seedlings, 2NS only two 16-40 yr old suppressed saplings, and 3NS five seedlings up to 7 yr old. The span of ages in S plots indicated that these plots had been suitable for germination and establishment for many years.

Even in S plots, there were some years during which few or no seedlings became established. This was the case during our 2nd yr of observation, despite a large cone crop in the previous year. In fact, in 1982 there was no statistically significant difference in the numbers of 1 st yr seedlings between S and NS plots (means = 10.3,6.7). However, in 1982 the pattern of germination and mortality appeared to be quite different on S and NS plots. Germination began 2-3 wk after snow melt on all plots, but further germination ceased and significant seedling mortality began » 1 mo earlier on NS plots than on S plots. By mid-July the resulting patchiness in seedling distribution was again apparent.

Root systems of all but the oldest seedlings were within the upper 0.2 m of mineral soil. Mean root length (+SD) of Ist-yr seedlings in August was 52.2 + 22.6 mm (N= 25 seedlings). Minimum soil moisture (10-11%) and the pattern of soil moisture depletion within this depth during the latter part of the growing season were not significantly different (ANOVA) for S and NS plots in either year. In 1982 the soil moisture declined rapidly from 25 to 10% 30-45 d after the snowpack melted. There was no evident difference in snowpack depth or date of snow melt between plot pairs.

Midday air and soil temperatures in both 1981 and 1982 were similar and were markedly and consistently warmer on NS plots (Fig. 2), by an average of 5°C in the air and just below the soil surface, 2° in the subsoil ( - 0.1 m).
 

Quantum flux density on the forest floor

Results of the ozalid paper technique, which integrates irradiance over time, indicated consistently higher total irradiance on the NS plots. For S plots a mean of 8.9 ozalid sheets were exposed per week for 10 wk in 1981, and for NS plots a mean of 10.0 sheets. In 1982, a 1-wk period showed means of 7.7 and 8.4, respectively. These differences between pairs are statistically significant (Student's t test, P < .01) and, because increasing amounts of light are required to expose additional ozalid sheets through the packet, probably underestimate the ecological significance of the considerably greater total irradiance on the NS plots.

The daily pattern of PPFD during three randomly selected days appears to show a timing difference in all three plot pairs. For example, the daily course of mean PPFD at 10-min intervals on 4 September, a representative day (Fig. 4) indicated that plot 1 NS received 2-4 times as much irradiance as 1 S between 0600 and 1300. These differences are greatest during midday; however, in the late afternoon differences were less pronounced, and IS received more light than INS.
 

Sunfleck size

Data on sunfleck size and number between 0800 and 1700 show that sunflecks in NS plots were similar in number (S = 18.2 ± 8.4 SD, NS = 15.9 ± 6.5 SD) but larger by 40% than sunflecks in S plots (S = 0.20 ± 0.04 m SD, NS = 0.28 ± 0.16 m SD, ANOVA). This size difference was not significant. When considered at individual hourly intervals or over the entire day, mean (n = 6 transects, two transects per plot) total percentage ground area covered by all sunflecks [S (number sunflecks x length of sunfleck)] was not significantly different between S and NS plots (S = 24.2% ± 12.3 SD, NS = 28.8% ± 16.7 SD), probably because of the large variance and the inverse relationship between sunfleck size and number (ANOVA). A comparison of mean individual sunfleck size (Fig. 5) did reveal a daily pattern consistent with the differing pattern of irradiance in the 1S and 1NS plots shown in Fig. 4. Although the differences are not statistically significant, S plots had larger sunflecks than NS plots during early morning and late afternoon hours when irradiance was lower. In contrast, NS plots had larger sunflecks than S plots at midday when irradiance is higher. These differences in sunfleck size, consistent on all transects, are significant between 1000 and 1400 (ANOVA, P < .01).
 

Canopy structure and solar tracks on the south exposure

Seedling density and canopy structure on the north exposure

Seedling net photosynthesis

Discussion

Because large new gaps in the overstory, caused by losses of major limbs and of entire trees, occur infrequently, microsite differences in the pattern of the light regime should be consistent over several years. Consistent with this expectation, seedlings of different ages occur within the aggregations, indicating that these sites have been suitable for establishment for many years.

No other factors account for the clumped distribution of seedlings as readily as sunfleck distribution. The presence of cone-bearing trees throughout the south exposure and of seedlings of different ages within the aggregations suggest that irregularities of dispersal are not responsible for the observed seedling distributions. The close spatial proximity between areas where seedlings are present and absent and the lack of apparent differences in physical soil characteristics indicate that topographic or edaphic factors do not account for the differences in seedling abundance.

Limited data suggest that aggregation of seedlings is the result of differential germination and seedling survival. In 1982 it appeared that germination ceased earlier and rates of mortality were higher on plots on the south exposures with few seedlings, but because seed germination was extremely low, we could not demonstrate that these differences were statistically significant. However, a study by Selter (1983) on A. magnifica shastensis (Shasta fir) on similarly paired plots in the North Coast Range of California showed significantly later cessation of germination, higher germination, and higher survival on plots with lower irradiance. Selter predicted from differential rates of seedling mortality, that by the end of the growing season seedling survival would be three times higher on the high-density plots than on the low-density plots.

In contrast to the situation on the south slopes, the results of our second-year study indicate that aggregations of red fir seedlings are not correlated with sunfleck distribution on north exposures. North exposures often had high seedling densities under canopies as open as those of the plots with few seedlings on the south-facing slope. However, these differences in seedling densities are consistent with the lower total direct-beam irradiance on the north aspect.

Possible mechanisms for the observed relationship may be envisioned in terms of the direct effects of sunflecks on plant physiological response through their influence on energy balance and water relations, and on photosynthesis and respiration (Rackham 1975). Recently, Young and Smith (1979, 1980, Smith 1981) found that stomata! response, transpiration, leaf temperatures, and lower xylem water potentials were closely coupled to changes in irradiance in several understory species. They related differences between the understory herb species Arnica cordifolia and A. latifolia in withstanding water stress, to the latter's distribution in microhabitats where intensity, duration, and frequency of sunflecks were lower. While our study provides no direct evidence for these effects, Gordon (1970) reported that shading increased survival in I st and 2nd-yr red fir seedlings, and Nobel and Alexander (1977) presented experimental evidence that shade improved seedling establishment in Picea engelmannii on the north aspect and was essential on south exposures.

Our results show that a significantly larger portion of total PPFD received on south-aspect plots without seedlings is of high irradiance. This is particularly pronounced during morning and midday when irradiance would have the most severe effect on water relations through early stomata! closure and consequently lowered carbon gain and increased leaf temperatures. In fact, midday temperatures above, at, and below the soil surface were all significantly higher in plots with few seedlings, indicating higher levels of thermal stress. These differences are expected to have a significant effect on seedling survival. Limited data of S. R. Radosevich (personal communication) suggest that it is not high levels of irradiance per se that reduce seedling survival, but induced drought stress. Additional experiments, particularly on water relations, will be necessary to identify and characterize the mechanisms by which seedling survival and distribution are influenced by the light regime under the forest canopy. Because of the low photosynthetic rates measured and because most surviving seedlings occur in sites where light levels are usually below photosynthetic saturation, it would appear that these seedlings may not maintain a positive net carbon balance, suggesting that low light intensities may also be a factor in seedling mortality. A lower irradiance limit, although not demonstrated, is also necessary.

Acknowledgments

Literature Cited