Construction of CO₂ application system.

Food-grade liquid CO₂ was housed on the property in a 10-ton tank measuring 20 × 8 × 8 ft, which rested atop four thick planks. The tank was filled to capacity about every 45 d at a cost of about $300 per ton (Airgas West, Lakewood, CA). The tank was connected to an ambient vaporizer that warmed the liquid, vaporizing it into gas. The gas flow rate was measured using a root gas meter and electronic gas volume corrector (Mini-AT; Mercury Instruments, Cincinnati). The gas moved through 550 ft of polyvinyl chloride (PVC) pipe to the three raspberry tunnels. At each tunnel entrance, the PVC pipe was connected to black plastic tubing and then to ultralow-flow drip irrigation tape with 8-inch emitter spacing (T-tape; T Systems, San Diego), suspended within the raspberry canopy for the entire length of the tunnel. To ensure gas was flowing through the length of the tunnel, we checked the ends of the drip tape on a weekly basis by feeling the pressure of gas flow on the walls of the tape [about 2 psi (without gas flow, drip tape walls were not taut)] and by listening for the sound of gas flow through the tape. We applied about 25 tons of CO₂ between 15 May and 24 Oct.

In two of the three tunnels receiving CO₂, we placed gas analyzer sensors (WMA-4; PP Systems, Amesbury, MA) within the raspberry canopy at about 120 cm aboveground. The gas analyzers recorded CO₂ concentration every 30 s and were connected to a datalogger (CR 10X; Campbell Scientific, Logan, UT) that output an average reading each minute.

Photosynthesis parameters.

We measured photosynthetic photon flux (PPF) (LI-189; LI-COR, Lincoln, NE), leaf stomatal conductance (g_S), leaf CO₂ assimilation, air temperature (TPS-1, PP Systems), and leaf fluorescence (Handy PEA; Hansatech Instruments, King's Lynn, UK) in January, before CO₂ was applied to determine where to locate the tape within the canopy for CO₂ application, in June, at the start of CO₂ application and second harvest (floricanes), and in September, at the third harvest, which was 1 month before ending CO₂ application. Three young, fully expanded, healthy green leaves were selected at three canopy heights—40, 80, and 120 cm from the ground—on each of six plants per treatment in the center row of each tunnel or at the outer surface of the plant canopy in the case of air temperature. In January, photosynthetic parameters were determined for all leaves every hour from sunrise to late afternoon on the side of the row with the highest PPF. For subsequent sampling dates, measurements were conducted during the morning hours only because we determined in our full-day measurements that the morning hours had the highest rates of g_S and CO₂ assimilation.

Measuring plant baseline conditions and response to CO₂ application.

Before CO₂ application, we counted the total number of floricanes and measured their height in two randomly selected 20-ft subplots in the center row of each tunnel. We also determined fruit yield in each plot, the average berry weight of 25 randomly selected fruit, and the number of 6-oz baskets harvested from the full tunnel length for the duration of commercial harvest from 26 May to 15 June. We repeated the cane height measurements after 2 months of CO₂ application in July, before the pruning of floricanes that had finished fruiting. We also measured the pruned floricane biomass (fresh weight) from each plot. For the third harvest from 12 Sept. to 30 Oct., we repeated all yield measurements. Yields for the CO₂ treatment are reported as percentages of the untreated control in accordance with an agreement made with our grower collaborator. We also pureed the 25 fruit from each plot and measured soluble solids using a portable refractometer (model RHB-32ATC; Westover Scientific, Mill Creek, WA). The remaining puree was frozen before analysis of free glucose, fructose, and sucrose using high-performance liquid chromatography (HPLC), total glucose by enzymatic hydrolysis followed by HPLC, and total nonstructural carbohydrates and starch derived by calculations from analysis results (University of California Division of Agriculture and Natural Resources Analytical Laboratory, Davis). Finally, in December, we measured cane diameter in the research plots using calipers at the soil surface.

Statistical analysis.

Data were analyzed using the General Linear Model (GLM) Procedure in SAS (version 9; SAS Institute, Cary, NC). The overall error rate was controlled by Tukey-Kramer adjustment. We validated the assumptions of normal distribution and equal variance using diagnostic plots generated by the GLM procedure.

Repeated-measures analysis was conducted also using the GLM procedure to determine effects of canopy height on PPF, g_S, CO₂ assimilation, and leaf fluorescence (F_v/F_m) from 0830 hr to 1630 hr, with time as the repeated-measure factor and means separation by Duncan's multiple range test. Air temperature was analyzed using the GLM procedure with means separation by Fisher's least significant difference test.

Optimization of CO₂ application system.

Photosynthesis parameters for plants in the tunnels were determined before CO₂ was applied. PPF was typically 20% lower in the tunnels than outside (data not shown). In other research, photosynthetically active radiation under plastic that was reused in the second year of the season, similar to this study, was 50% of outdoor irradiance (Daugovish and Larson, 2009). As light within the tunnels approached 50% full sunlight, PPF decreased proportionally from the uppermost leaves 120 cm aboveground to the lower leaves 40 cm aboveground (Fig. 2A). Averaged between 0830 and 1630 hr, PPF at 120 cm was 56% greater than at 80 cm; PPF at 80 cm was 58% greater than for leaves at 40 cm (Table 1, P < 0.0001). Mid and lower leaves were likely shaded by leaves higher in the canopy, which received the most sun.

Table 1.

Photosynthesis parameters averaged from 0830 hr to 1630 hr at three heights in the canopy of raspberry plants under high tunnels near Camarillo, CA, in Mar. 2007 before carbon dioxide (CO₂) application. Data were analyzed using repeated-measures analysis with time as the repeated measure; the same letters within a column denote nonsignificant comparisons at P < 0.05 using Duncan's multiple range test.

View Table

PPF density (A), leaf g_S (B), leaf CO₂ assimilation (C), and leaf fluorescence (D) at 120 cm (▲), 80 cm (■), and 40 cm (◆) aboveground (1 cm = 0.3937 inch) in the canopy of raspberry plants under high tunnels before CO₂ application near Camarillo, CA. Values reported at each hour are the mean result for measurements made on one leaf at each height on each of three plants along the trellis in two separate tunnels (n = 6). The same letters at each time denote nonsignificant comparisons at P < 0.10 using Duncan's multiple range test.

Citation: HortTechnology hortte 20, 1; 10.21273/HORTTECH.20.1.213

• Download Figure

View Full Size

PPF density (A), leaf g_S (B), leaf CO₂ assimilation (C), and leaf fluorescence (D) at 120 cm (▲), 80 cm (■), and 40 cm (◆) aboveground (1 cm = 0.3937 inch) in the canopy of raspberry plants under high tunnels before CO₂ application near Camarillo, CA. Values reported at each hour are the mean result for measurements made on one leaf at each height on each of three plants along the trellis in two separate tunnels (n = 6). The same letters at each time denote nonsignificant comparisons at P < 0.10 using Duncan's multiple range test.

Citation: HortTechnology hortte 20, 1; 10.21273/HORTTECH.20.1.213

• Download Figure

PPF density (A), leaf g_S (B), leaf CO₂ assimilation (C), and leaf fluorescence (D) at 120 cm (▲), 80 cm (■), and 40 cm (◆) aboveground (1 cm = 0.3937 inch) in the canopy of raspberry plants under high tunnels before CO₂ application near Camarillo, CA. Values reported at each hour are the mean result for measurements made on one leaf at each height on each of three plants along the trellis in two separate tunnels (n = 6). The same letters at each time denote nonsignificant comparisons at P < 0.10 using Duncan's multiple range test.

Citation: HortTechnology hortte 20, 1; 10.21273/HORTTECH.20.1.213

• Download Figure

Leaves at the top and middle of the canopy had similar g_S rates, about 25% higher than for lower leaves when averaged for the day (Table 1, P = 0.0048). Midday decline in g_S was observed for leaves at all three canopy heights, including a 48% reduction between 1130 and 1230 hr for the topmost leaves (Fig. 2B). This midday depression in g_S has been documented in ‘Titan’ red raspberry and other woody plants (Fernandez and Pritts, 1994). Stomates likely close to prevent excessive transpiration in response to high irradiance (Fig. 2A), causing increased leaf temperature; air temperatures during this period increased by more than 3 °C (Table 2). Compared with outside the tunnel, air temperatures under high tunnels in south coastal California were 1 to 3 °C greater at all hours of the day during 1 year but 2 to 4 °C lower during the day and 2 to 3 °C higher at night in the second year, probably because of reduced irradiance levels in the tunnel due to reuse of the weathered polyethylene tunnel cover (Daugovish and Larson, 2009).

Table 2.

Ambient air temperature at three heights on the outer surface of the canopy of raspberry plants under high tunnels before CO₂ application. Values reported at each hour are the mean result for measurements made on one leaf at each height on each of three plants along the trellis in two separate tunnels (n = 6).

View Table

Leaves at 80 cm aboveground had an average CO₂ assimilation rate about 60% higher than leaves at 120 or 40 cm (Table 1, P = 0.0038). CO₂ assimilation at this height was maintained at a higher rate compared with the other canopy heights throughout most of the day (Fig. 2C). In the early afternoon, when assimilation and conductance rates were similar at all leaf heights, air temperatures at the outer surface of the canopy were the highest of the whole day and similar between 80 and 120 cm height (Table 2).

Although the ratio of variable fluorescence to maximum fluorescence (F_v/F_m) was highest (F_v/F_m = 0.8) for leaves at 40 cm height compared with leaves at 80 and 120 cm heights when averaged for the day [F_v/F_m = 0.78 (Table 1, P = 0.032)], this 3% difference is likely not physiologically significant because a typical leaf has an efficiency of about 80% (F_v/F_m = 0.8) (Carpentier, 2005; Schreiber, 2004). The ratio of F_v/F_m provides an estimate of the energy conversion efficiency in Photosystem II. In this study, only leaves at 120 cm during midday had leaf fluorescence values (F_v/F_m = 0.75) (Fig. 2D) and corresponding reduced rates of CO₂ assimilation (Fig. 2C) that suggest leaf stress and potential damage from excess light energy, possibly as heat (Ralph, 2005; Schreiber, 2004). Air temperature at the top of the canopy was generally higher throughout the day than at mid and lower canopy heights (Table 2).

Based on these results, we chose to apply CO₂ at a height of 100 cm, where leaves had the highest rate of CO₂ assimilation and were efficient at using available light energy. Furthermore, applying CO₂ in the middle of the canopy allowed for interception by lower leaves during periods of minimum air movement when gas would move downward with gravity and also accommodated increasing plant height through the growing season. Due to the observed midday reduction in g_S for leaves at all heights in the canopy and reduced efficiency of Photosystem II during the midday, we chose to apply CO₂ between 0700 and 1100 hr again between 1400 and 1800 hr.

Once CO₂ application began in June, we maintained an average concentration of 436 ppm (se = 19 ppm) or about 18% above ambient conditions (370 ppm) for 4 months. CO₂ concentration under the tunnel averaged 471 ppm (se = 27), or 18% higher in the morning hours compared with the afternoon which was 401 ppm (se = 11), possibly because air movement within the tunnel often increased in the afternoon. In an attempt to increase CO₂ concentration under the tunnel, we increased the duration of CO₂ application from 8 h to 12 h in mid-September for 1 month, but average CO₂ concentration under the tunnel remained unchanged [438 ppm (se = 26)], probably because the flow rate of the gas itself was unchanged. After CO₂ application was stopped in mid-October, CO₂ concentration under the tunnel averaged 398 ppm (se = 15).

Plant conditions before CO₂ application.

Before CO₂ application, raspberry cane height and the number of canes in tunnels designated to receive CO₂ were similar to tunnels designated as untreated (data not shown). Yield in tunnels designated to receive CO₂ was also similar to that of tunnels designated as controls (Table 3, P = 0.87), but average berry weight and the number of baskets filled was 7% and 11% lower, respectively, in tunnels that would later receive CO₂ compared with tunnels that would not (Table 3; P = 0.0001 and 0.001, respectively). The smaller fruit and fewer baskets in tunnels designated to receive CO₂ corresponded with the 27% reduction in average CO₂ assimilation rates for these plants compared with those in tunnels randomly selected to be controls (Table 4, P = 0.03); reduced photosynthesis may have reduced carbon availability for increasing fruit size. By contrast, g_S and leaf fluorescence were similar among all tunnels (Table 4). The relative productivity of each tunnel before CO₂ application was not considered when treatments were randomly assigned.

Table 3.

Relative yield, fruit weight, and number of fruit baskets harvested from raspberry plants grown under tunnels before and after carbon dioxide (CO₂) application in 2007. Data for the CO₂ treatment are reported as percentages of the untreated control in accordance with an agreement made with our grower collaborator.

View Table

Table 4.

Carbon dioxide (CO₂) assimilation, g_S, and leaf fluorescence (ratio of variable fluorescence to maximum fluorescence) before CO₂ application to raspberry grown under high tunnels (January), at the start of CO₂ application (June), and at the end of CO₂ application (September) in 2007. Leaves were measured at three canopy heights and averaged.

View Table

Effect of CO₂ enrichment.

After CO₂ application, yield from small plots and individual fruit weight in CO₂ tunnels was 12% and 5% higher, respectively, than in untreated tunnels (Table 3; P = 0.03 and 0.0005, respectively). By contrast, CO₂ application did not affect the number of baskets (Table 3, P = 0.43), raspberry cane height, pruned cane biomass, cane diameter, or fruit carbohydrate content (data not shown), suggesting that increased carbon was used to increase fruit size rather than fruit quantity or chemical properties or cane size. Carbohydrate content of strawberry fruit grown in pots was similarly unaffected by CO₂ enrichment to 1000 ppm (Acuña-Maldonado and Pritts, 2008). CO₂ enrichment also did not affect the height of two herbaceous skullcap (Scutellaria spp.) grown for medicinal purposes, but aboveground dry weight was increased 44% at a CO₂ concentration of 1200 ppm (Stutte et al., 2008). The CO₂ concentration of about 400 ppm in this study may have been too low to affect noncrop biomass accumulation. Interestingly, average strawberry size was not increased by CO₂ concentrations higher than we achieved in our study for plants grown in pots in semiopen chambers (Acuña-Maldonado and Pritts, 2008) or field-grown under tunnels (Hartz et al., 1991).

It is important to view these modest increases in yield and berry weight in the context of previous harvests in this semiperennial cropping system: if ‘Titan’ red raspberry was prevented from producing a full crop in 1 year, yields tended to be higher than normal in the following year (Fernandez and Pritts, 1996). The larger fruit size and similar number of fruit baskets for CO₂-treated plants compared with untreated therefore may be a result of smaller fruit size and fewer baskets filled before treatment. However, yield in small plots by weight rather than by number was similar in all tunnels before treatment in May and June but was higher in September and October after receiving CO₂, suggesting some influence of treatment rather than simply prior yield.

PPF, g_S, CO₂ assimilation, and leaf fluorescence were similar for plants in CO₂ and untreated tunnels in June at the start of CO₂ application and in September near the end of CO₂ application (Table 4). These static measures could not account for the differences in yield and fruit size. Floricane leaf photosynthetic rates measured in a similar fashion for ‘Titan’—before, during, and after fruiting—were also unaffected by treatments as extreme as full canopy shading, primocane removal, or flower removal (Fernandez and Pritts, 1996). Instead, if plant physiological parameters were measured during the course of the entire day as they were before CO₂ was applied and for the full growing cycle, differences between treatments may have been captured. A carbon budget estimating net seasonal CO₂ assimilation in the plant canopy may be much more closely correlated to crop yield compared with instantaneous measurements (Zelitch, 1982). We also note that other raspberry varieties, whether public or private, may respond differently to CO₂ application compared with the proprietary variety used in this study: Stafne et al. (2000) measured a wide range of CO₂ assimilation rates among raspberry genotypes sourced from the southern and northwestern United States, Japan, China, and South Korea, and Keep (1972) found wide variation in physiological and morphological traits was among wild populations at 12 locations across Europe.

Increasing CO₂ application to 12 h per day with the same gas flow rate did not significantly increase CO₂ concentration under the tunnel, but may increase CO₂ assimilation, which, unlike g_S, was not reduced during the midday. Most, if not all, controlled studies increased the CO₂ concentration rather than the duration of application, which increased raspberry photosynthesis as long as light was not limiting (Fernandez and Pritts, 1994; Percival et al., 1996): other methods to increase CO₂ concentration inside the tunnel, such as closing the ends or using multiple delivery lines, could be future avenues of investigation.

Our preliminary study demonstrates some potential benefits of CO₂ application for field-grown raspberry under high tunnels. We hope these results and subsequent research stimulate further development of waste-CO₂ recycling technology, enabling progressive agricultural producers to improve the quality of their crops while contributing to a viable solution to greenhouse gas emissions and to the sustainability of their operations.