Introduction

In arid and semi-arid ecosystems, photodegradation has been recently identified as a key process in ecosystem carbon (C) cycling (King et al. 2012; Song et al. 2013, and references therein). Photodegradation refers to the process through which solar radiation decomposes organic matter. Multiple field experiments have demonstrated that ultraviolet (UV) radiation and visible radiation increase litter mass loss via photodegradation (Austin and Vivanco 2006; Barnes et al. 2011; Brandt et al. 2010; Day et al. 2007; Gallo et al. 2006; Liu et al. 2014). A meta-analysis showed that increased exposure to solar radiation enhanced litter mass loss by 23 % on average (King et al. 2012). Despite the increasing interest in understanding the role of photodegradation in ecosystem C cycling, it remains relatively unclear exactly how photodegradation induces litter mass loss.

Photodegradation can directly contribute to litter mass loss through photochemical mineralization. Laboratory-based studies found that exposure to radiation can induce trace gas emissions (CO2, CO, and CH4) from plant litter (Brandt et al. 2009; Lee et al. 2012; McLeod et al. 2008; Schade et al. 1999). Rutledge et al. (2010) suggested that photodegradation accounted for almost 60 % of CO2 flux from a California grassland during summer. Photodegradation can also indirectly affect litter decomposition by influencing litter chemical composition. Lignin has been found to be preferentially degraded by photodegradation, as lignin strongly absorbs UV and visible radiation (Austin and Ballaré 2010; Day et al. 2007; Henry et al. 2008). Consequently, photodegradation is thought to improve the biodegradability of litter, since lignin often degrades slower than other compounds in litter (Aerts 1997; Meentemeyer 1978). Adding another level of complexity, solar radiation, especially UV radiation, also suppresses microbial activity, as it is known to damage microbial nucleic acids (Hughes et al. 2003; Sinha and Häder 2002). Several studies have found that UV exposure decreases litter nitrogen (N) immobilization (Brandt et al. 2010; Smith et al. 2010), a microbial process through which N is transferred from the environment to litter. There is a significant gap in understanding the relative importance of the direct and indirect contributions of photodegradation, as few studies have attempted to separate and quantify them.

Arid ecosystems characterized by distinct dry and wet periods present an opportunity to separate the direct and indirect contributions of photodegradation to litter mass loss. Radiation exposure during the dry season can “condition” litter for microbial decomposition in the following wet season (Foereid et al. 2010). If photodegradation preferentially degrades lignin in the dry season, it might relieve the inhibitory effect of lignin on subsequent microbial decomposition in the wet season. This conditioning effect of photodegradation can have significant implications at ecosystem scales. For example, severe drought might increase the importance of photodegradation and the loss of lignin during the dry season. If these changes made up for a drought-induced decrease in microbial decomposition, then drought would not suppress decomposition in arid ecosystems. However, mixed results have been reported about the conditioning effect of photodegradation (Brandt et al. 2010; Foereid et al. 2010; Henry et al. 2008; Lambie et al. 2014; Wang et al. 2015). For example, several studies have found that prior exposure of litter to UV radiation facilitates microbial decomposition (Foereid et al. 2010; Henry et al. 2008; Wang et al. 2015). Brandt et al. (2009) and Lambie et al. (2014), on the other hand, reported negligible or even negative effects of UV radiation exposure on subsequent microbial decomposition. The UV exposure in most of the above studies was achieved using UV lamps in the laboratory or greenhouse. Few studies to date have examined whether field UV exposure will facilitate microbial decomposition, particularly as litter experiences distinct dry and wet seasons (except Henry et al. 2008).

The objective of this study was to examine how field UV radiation exposure affected subsequent microbial decomposition of litter of an abundant grass in California, Bromus diandrus. Litter was exposed in the field to two levels of UV radiation (ambient vs. reduced) for different periods: summer, summer-winter, or 1 year. Then the litter was incubated with microbial inoculum for a period of 25 days under laboratory conditions to evaluate its biodegradability. We asked the following questions: 1) does intensive UV exposure during a Mediterranean summer increase subsequent microbial decomposition by increasing loss of persistent substrates, such as lignin? and 2) does the conditioning effect of UV exposure differ among exposure periods?

Materials and methods

Litter collection and UV treatments

Litter samples of B. diandrus were collected from the University of California’s Sedgwick Reserve in Santa Ynez, California, USA (43°42′N, 120°2′W; 25 km north of Santa Barbara). A detailed description of the site can be found in Lin and King (2014). Briefly, the site is dominated by European annual grasses, particularly B. diandrus, and it experiences a Mediterranean climate of distinct wet and dry seasons with average annual precipitation of 380 mm, mostly occurring between November and April. Annual grasses typically fully senesce by late April. Senesced litter lying across the ground surface forms a litter layer of 5 to 15 cm thickness, the surface of which is exposed to intensive solar radiation during the dry season from May to September. To manipulate UV radiation (280–400 nm) received by litter samples, 20 pairs of steel frames (l × w × h: 75 × 150 × 25 cm) were constructed with plastic louvers that either block or pass UV radiation. A subset of the screens were used in Lin and King (2014), which reported the technical details of these screens, including dimensions, placement, optical properties, and effects on air temperature and relative humidity. In short, the “UV block” screens eliminated 93 and 85 % of UV-A (315–400 nm) and UV-B (280–315 nm) radiation, respectively, whereas the “UV pass” screens transmitted 80 and 79 % of UV-A and UV-B radiation, respectively. Screens allowed penetration of rainfall and controlled for heating by having a louvered design. There was no difference between UV block and UV pass screens in their effects on photosynthetically active radiation (PAR), air temperature, or relative humidity.

Litter samples were under either UV block or UV pass treatments in the field for three periods (summer, summer-winter, and one year) (Table 1). For litter that received UV treatments in summer, 10 pairs of UV block and UV pass screens were placed over areas dominated by litter of B. diandrus in mid August, 2011. During the set-up of the screens, some standing litter was pushed over by hand so that it would fit underneath the screens. In late October 2011, litter was removed from under the screens resulting in UV treatment that lasted for 2.5 months. Only litter at the very top of the thatch layer and constantly exposed to solar radiation was collected for this study. The other two sets of litter were obtained from the experiment reported in Lin and King (2014). In short, B. diandrus litter was sealed in 20 × 20 cm aluminum bags of 1.5-mm mesh size and suspended at 5 cm beneath the louvers of 10 pairs of UV pass and UV block screens and above the thatch layer in the field in mid August, 2011. The bags were supported from below by a stainless steel screen. The UV screens were not the same ones used for treating litter during summer, but they were identical in design. Litter samples (n = 10) were collected both in early March 2012 and early September 2012 to achieve UV radiation exposure periods of summer-winter and one year, respectively. These three sets of litter all originated from the 2010–2011 growing season at the same field site. Even though that aluminum mesh bags were not used for samples exposed during summer, we believe the use of mesh bags was not a confounding factor to the exposure period. The aluminum mesh material transmits greater than 70 % of UV radiation, and its mesh size is big enough for microbial decomposers to colonize the litter inside the mesh.

Table 1 Bromus diandrus litter samples and their field UV exposure characteristics
Full size table

We monitored UV radiation at 1.7 m above the soil surface with a broadband UV radiometer (CUV5, Kipp & Zonen) at a meteorological station adjacent to the site. After considering light transmission of screens and aluminum mesh, as well as length of exposure, we estimated the amount of UV radiation received by each treatment during field exposure (Table 1).

Sample processing and chemical analysis

After collection of the litter from the field site, green plants, visible soil, and arthropods were removed from the litter. Litter was then oven-dried at 55 °C for 2 days. Four out of ten replicates were randomly taken from each combination of UV treatment and exposure duration for chemical analysis and measurement of biodegradability. These samples were ground using a Wiley mill with U.S. standard #20 mesh.

We analyzed litter carbon fractions, including the cell solubles fraction (which includes soluble carbohydrates, proteins, and lipids; hereafter, cell solubles), hemicellulose, cellulose, and lignin, using a sequential extraction technique (Van Soest 1963). Subsamples were treated with neutral fiber detergent, acid fiber detergent, and sulfuric acid digestions using an ANKOM fiber analyzer (Type 2000, ANKOM Technology). We refer to the fraction left after sulfuric acid digestion as ‘lignin’ so that our results can be compared with many previous studies that have adopted the same method in examining litter decomposition and photodegradation (Austin and Vivanco 2006; Brandt et al. 2010; Rozema et al. 1997). We recognize that this lignin fraction also includes cutin, suberin, and waxes (von Lützow et al. 2007). For litter C and N concentrations, subsamples were ground to powder using a roller mill and analyzed using an elemental analyzer (Fisons NA1500, Fisons Instruments) with acetanilide standards. Each sample was analyzed in duplicate, and the average value was used. For extraction, a 100 mg subsample was soaked in 50 ml deionized water at 4 °C for 24 h. Extracts were filtered through glass fiber filter paper (Type A/E, Pall Corporation) and analyzed for water extractable C (WEC) and N (WEN) using a total organic C/total N (TOC/TN) analyzer (Series V, Shimadzu Corporation). Potassium hydrogen phthalate and potassium nitrate were used to prepare the standards for WEC and WEN, respectively. WEC and WEN were calculated as the average of three measurements. All litter chemical characteristics were reported on a dry litter mass basis.

Litter biodegradability

Litter biodegradability was evaluated by measuring microbial respiration in a 25-day laboratory incubation experiment on subsamples of the coarsely ground litter (n = 4, #20 mesh). Subsamples (250 mg each) were first placed into 50-mL plastic beakers. Microbial inoculum was added to introduce a uniform community of decomposers to all of the litter samples and to offset potential effects of UV exposure on the microbial community on the litter itself. To make the microbial inoculum, soil from the field site was mixed with water at 1:3.5 (soil:water, mass:volume ratio) and extracted at 50 rpm on a bench shaker for 2 h. After shaking, the extract was filtered through Whatman 40 filter paper to remove soil particles and then used as microbial inoculum. For each plastic beaker, 250 μL of microbial inoculum was added with 2 mL deionized water to fully soak the litter sample. The TOC measurements revealed that there was approximately 20 μg C in the inoculum for each plastic beaker, which represents less than 0.3 % of total CO2-C produced during the incubation. The 50-mL beakers were then placed into 473 mL glass jars, sealed, and incubated at 20 °C in the dark. Microbial respiration was estimated by measuring CO2 production during the incubation. For each glass jar, a 1 mL headspace sample was obtained through a butyl stopper in the lid using a needle and syringe, and its CO2 concentration was measured using an infrared gas analyzer (IRGA, LI-COR 820, LI-COR Corporation) every 1 or 2 days. The IRGA was calibrated at each measurement time point using four CO2 standards ranging from 500 to 25,000 ppm (Scott Specialty Gases, Plumsteadville, PA). The CO2 concentration was converted to grams CO2-C using the ideal gas law. All glass jars were vented when any single headspace CO2 concentration exceeded 2 %. Average microbial respiration rate between two measurements was calculated as the increase of CO2-C in each glass jar between the two time points per hour incubated per dry mass of litter. Cumulative microbial respiration (CMR) for the 25-day incubation period was calculated as the sum of CO2-C production in each glass jar per dry mass of litter and was used to represent litter biodegradability.

Statistical analysis

Preliminary two-way analysis of variance (ANOVA) found significant interaction effects between UV treatment and exposure period on most of the studied variables, suggesting that the effects of UV treatment should be examined for each exposure period separately. Therefore, we conducted Student’s t-test to compare differences in litter carbon fractions, C and N concentrations, WEC, WEN, and CMR between the UV block and UV pass treatments for each period of UV treatment separately. Before applying the t-test, samples were checked for equality of variances using Levene’s test. If equal variances could not be assumed between two treatments, the degrees of freedom of the t-statistic were adjusted using the Welch-Satterthwaite method. Pearson correlation was used to examine the relationship between litter chemical characteristics and CMR. All statistical analyses were carried out in SPSS (Version 20, IBM Corporation).

Results

Litter chemical quality

For litter exposed to UV treatments during summer, litter N concentration was lower in the UV pass than in the UV block treatment (n = 4, P = 0.013, Table 2). Its C concentration was higher under UV pass than under UV block (n = 4, P = 0.021). Litter WEN also tended to be lower under UV pass compared to UV block (n = 4, P = 0.066). Litter lignin concentration and other measured chemical characteristics were not affected by the summer UV treatments. For litter exposed to UV treatments over summer-winter, UV pass did not affect litter lignin concentration (n = 4, P = 0.139) or other measured chemical characteristics. After 1 year of UV treatments, litter hemicellulose concentration was lower under UV pass compared to UV block (n = 4, P = 0.009). This decrease in hemicellulose corresponded to a trend of higher cell solubles under UV pass than under UV block (n = 4, P = 0.082). No other litter chemical characteristics, including lignin concentration, were affected by 1 year of UV treatments.

Table 2 Effects of UV treatments on chemical characteristics of Bromus diandrus litter
Full size table

Litter biodegradability

For litter exposed to UV treatments during summer, the UV block treatment increased its biodegradability (represented by cumulative microbial respiration (CMR)) by 28 % during the 25-day incubation period compared to UV-exposed litter (Fig. 1, n = 4, P = 0.046). The positive effect of blocking UV radiation on litter biodegradability was most pronounced at the peak of microbial activity (Fig. 2, 2nd day since the start of the incubation) when the microbial respiration rate associated with litter from the UV block treatment was 35 % higher than that associated with litter from the UV pass treatment (374.7 ± 26.8 μg C g−1 litter hr−1 vs. 279.5 ± 25.2 μg C g−1 litter hr−1; n = 4, P = 0.041). The litter from the UV block treatment also showed consistently higher microbial respiration rates during the second half of the incubation. Exposure to UV radiation treatments did not affect litter biodegradability when the exposure occurred over summer-winter (Fig. 1, n = 4, P = 0.972) or 1 year (n = 4, P = 0.367), and microbial respiration for those exposure durations was not affected by UV treatments at any time point throughout the incubation (data not shown).

Fig. 1
figure 1

Effects of UV manipulation and exposure periods (summer, summer-winter, and 1 year) on subsequent cumulative microbial respiration from Bromus diandrus litter measured in a laboratory incubation. Mean and standard errors are shown (n = 4). **P < 0.05

Full size image
Fig. 2
figure 2

Subsequent microbial respiration rate from Bromus diandrus litter as a function of time for litter exposed during summer. Mean and standard errors are shown (n = 4). *P < 0.1 and **P < 0.05

Full size image

For litter in the summer UV treatments, its biodegradability was strongly positively correlated with litter N concentration (Fig. 3a, n = 8, r = 0.928, P < 0.001). When UV treatments lasted over summer-winter, the correlation between biodegradability and N concentration was marginally significant (Fig. 3b, n = 8, r = 0.669, P = 0.070). When UV treatments lasted for 1 year, the correlation between biodegradability and N concentration was no longer significant (Fig. 3c, n = 8, r = 0.575, P = 0.136). Similarly, correlations between litter biodegradability and WEN were significant when UV treatments occurred over summer (data not shown, n = 8, r = 0.938, P < 0.001) and summer-winter (n = 8, r = 0.858, P = 0.006), but not significant for litter exposed to 1 year of UV treatments (n = 8, r = 0.341, P = 0.408). In fact, none of the measured litter chemical characteristics had a significant correlation with litter biodegradability for litter exposed to 1 year of UV treatments.

Fig. 3
figure 3

Relationship between litter N concentration and cumulative microbial respiration when the treatments were applied during a summer, b summer-winter, and c 1 year. r, Pearson correlation coefficient

Full size image

Discussion

Contrary to our hypotheses, we did not find positive effects of UV exposure on litter biodegradability for any of the exposure periods (Fig. 1). Lignin concentration was also not affected by up to 1 year of UV treatments (Table 2). In this study, we used B. diandrus, a common invasive species found in California grasslands. This species has lower lignin concentrations (2–5 %) than many other grasses or woody species (Jung et al. 1999; McLauchlan et al. 2006; Van Soest 1963). Thus, it could be difficult to detect changes in lignin concentration induced by photodegradation. However, UV treatments had limited effects on all of the other litter chemical characteristics as well, suggesting that UV exposure did not improve litter biodegradability through breakdown of recalcitrant substrates.

Surprisingly, we found that exposure to UV treatments during summer decreased litter biodegradability (Figs. 1 and 2). This result is consistent with a laboratory study in which Lambie et al. (2014) found that exposure to UV radiation decreased subsequent microbial respiration from pine (Pinus radiata) and mānuka (Leptospermum scoparium) litter. However, the results of Lambie et al. (2014) did not demonstrate the mechanism behind this negative conditioning effect of UV radiation. Photodegradation could increase litter mass loss and decrease the quality and biodegradability of the remaining litter. Exposure to UV radiation did increase litter mass loss when exposure occurred over summer-winter and one year (Lin and King 2014). Litter mass loss was not measured during summer only, but the same UV exposure effect was likely. However, if UV exposure decreased litter biodegradability mainly through reducing litter quality, then a negative effect of UV exposure on biodegradability would have been found in all exposure durations, and this effect would have been strongest in litter with the longest UV exposure (1 year). Instead, UV exposure only decreased litter biodegradability in summer, the shortest UV exposure. We found a strong positive relationship between litter biodegradability and N concentration only when UV treatments occurred during summer (Fig. 3), suggesting that the early stage of litter decay is limited by N availability in our incubation. This N limitation to short-term microbial respiration has been commonly observed (e.g. Allen and Schlesinger 2004; Vance and Chapin 2001). Given the strong correlation between biodegradability and N concentration, we speculate that the UV-induced decrease in litter N concentration (Table 2) led to lower biodegradability in the UV pass treatment.

Several studies have reported reduced N immobilization on photodegraded litter (Brandt et al. 2010; Lin and King 2014; Smith et al. 2010; Song et al. 2011). It is likely that UV exposure over summer decreased litter N concentration through suppression of microbial N immobilization. This inhibitory effect of UV on N immobilization was temporary, as litter N concentration was no longer different between UV treatments for litter exposed during summer-winter and 1 year (Table 2, Lin and King 2014). Litter N immobilization presumably occurs during early stages of decomposition (e.g. the first summer after B. diandrus senesces), when litter N cannot meet the N requirements of microbial decomposers. The UV effect on N immobilization should be much stronger in summer than in winter, as high moisture availability and low UV intensity in winter favor microbial activity (Johnson 2003; Xiang et al. 2008). Therefore, favorable environmental conditions during the wet season likely mask the difference in N immobilization induced by UV during summer.

Our results suggest a new mechanism through which photodegradation affects litter mass loss: alteration of biodegradability through changes in microbial N immobilization patterns (Fig. 4). This mechanism can potentially explain the negative conditioning effect of UV on litter mass loss found in Lambie et al. (2014). Given that photodegradation can both positively and negatively affect litter mass loss, it is critical to understand the controls of these mechanisms. Our study indicates that the relative importance of different photodegradation pathways (Fig. 4) is affected by seasonal patterns of environmental factors, such as solar radiation and moisture. As discussed above, the negative effect of UV on litter biodegradability is likely to occur during early stages of litter decomposition when N immobilization is necessary and during summer when environmental conditions favor photodegradation. The cumulative dose of radiation could also regulate the balance among photodegradation pathways (Foereid et al. 2010); however, the strong seasonal variation in solar radiation (Table 1) limits our ability to separate its effect. Future studies are needed to specifically characterize the mechanistic controls of different mass loss pathways during photodegradation.

Fig. 4
figure 4

Conceptual model of solar radiation effects on litter mass loss. Rectangles indicate litter decomposition pathways. Ellipses indicate factors that affect litter decomposition. Radiation induces photochemical mineralization that increases litter mass loss. Radiation also affects litter microbial decomposition through either suppressing microbial activity or altering litter chemistry. This study suggests that radiation-induced changes in microbial activity (e.g. reduced N immobilization) can influence litter chemistry (dashed arrow), which further affects litter mass loss

Full size image

Furthermore, there are several alternative mechanisms behind the conditioning effect of photodegradation that require further examination. For example, even though this experiment did not find positive effects of UV radiation exposure on subsequent microbial decomposition of litter, microbial decomposers on our litter samples might have already consumed the labile substrates released by photodegradation before the samples were collected from field. In other words, the conditioning effects of UV radiation on biodegradability might operate at a much shorter time scale than that measured in this experiment. Specifically, during summer in California grasslands, photodegradation likely dominates litter decomposition during daylight hours and may condition organic matter for microbial decomposition at night. Another alternative mechanism is that exposure to UV radiation may also induce physical fragmentation of litter and increase its biodegradability. We ground our litter samples prior to the incubation study; therefore, our results did not evaluate the impacts of UV exposure on litter physical characteristics.

In arid and semi-arid ecosystems, it has been suggested that C and N dynamics during decomposition are decoupled, as observations have shown that litter decomposition does not depend on litter C:N ratio, and N immobilization is not observed regardless of initial litter N content (Parton et al. 2007; Vanderbilt et al. 2008). Several abiotic processes have been proposed to explain this decoupling of C and N dynamics, such as photodegradation and soil-litter mixing (Brandt et al. 2010; Hewins et al. 2013; Throop and Archer 2007). Our results, however, suggest that C and N dynamics during litter decomposition can be coupled by photodegradation, as photodegradation likely decreased microbial decomposition by altering N immobilization. Similarly, a combination of photodegradation and N addition was shown to decrease the overall decomposition rate of Pinus massoniana litter (Song et al. 2014b). Song et al. (2014a) also found that the interaction between photodegradation and N addition induced faster litter mass loss than the sum of their individual effects. Photodegradation appears to either positively or negatively affect litter decomposition through interaction with litter N dynamics. More work is needed to fully understand the mechanisms behind these seemingly contradictory results. Nevertheless, impacts of photodegradation on the interaction between C and N dynamics during litter decomposition are much more complex than a single “decoupling” effect.

In summary, our study shows that up to one year of conditioning with UV radiation does not facilitate microbial decomposition of B. diandrus litter. In fact, UV exposure decreased the subsequent microbial respiration rate when the exposure occurred during summer and had no significant effects when exposure was longer. We suggest that UV radiation suppressed N immobilization and consequently limited subsequent microbial decomposition of litter. Together with previous studies (Foereid et al. 2010; Lambie et al. 2014), our results imply that photodegradation may influence subsequent microbial decomposition through altering microbial activity and/or affecting litter chemical composition. Instead of decoupling C and N dynamics, photodegradation may affect litter C loss by interacting with litter N turnover. Further studies are required to closely examine the nature and controls of these mechanisms to better understand photodegradation, as well as its contribution to decomposition processes in general.