Compensatory light utilization of Phalaris arundinacea within the Phalaris arundinacea–Phragmites australis compound community

ABSTRACT This study investigated the light utilization characteristics of Phalaris arundinacea within the Phalaris arundinacea-Phragmites australis compound community. The investigation was based on the response curve equation of the net photosynthetic rate-photosynthetic photon flux density (Pn-PPFD) of P. arundinacea and on the distribution of PPFD on the surface canopy of P. arundinacea within the compound community. Results showed that P. arundinacea was able to compensate for substantial reduction in available light, and utilize the weak light within the lower part of the compound community. On the one hand, the density of P. australis within the compound community determined the compensatory light utilization rate (CLUR) of P.arundinacea. On the other hand, the CLUR amplitude of P. arundinacea within the compound community with the same P. australis density was relatively small in a day. The results can help people to regulate P. arundinacea-P. australis compound communities to maintain a long-term, stable coexistence in the constructed wetlands. Abbreviations: CK – control; CLUR, compensatory light utilization rate; LER, light extinction rates; Pn, net photosynthetic rate; PPFD, photosynthetic photon flux density.


Introduction
Phalaris arundinacea and Phragmites australis, which possess highly similar biological characteristics, are perennial rhizome clonal plants that often live in wetlands such as river/ lake floods, low-lying lands, swamps, and coasts. P. arundinacea is widely distributed in the north temperate zone of North America, Europe, and Asia (Martina and Ende 2012;Martínková and Klimešová 2017), whereas P. australis is widely distributed almost all over the world (Tucker 1990;Elonger 2009;Kulmatiski et al. 2011). In Europe, P. arundinacea and P. australis are widely used to build constructed wetland plant communities due to their extremely strong decontamination effects (Vymazal and Kröpfelová 2005;Gagnon et al. 2010;Vymazal 2011;Pedescoll et al. 2015). Jan Vymazal evaluated the establishment after planting, seasonal/long-term growth, and growth characteristics (biomass, stem count, and stem length) for P. australis and P. arundinacea planted in constructed wetlands for wastewater treatment in the Czech Republic (Vymazal and Kröpfelová 2005). Recently, Brezinová and Vymazal have evaluated the accumulation of heavy metals in above-ground biomass of P. australis and P. arundinacea in horizontal flow constructed wetlands for wastewater treatment (Březinová and Vymazal 2015;Vymazal and Březinová 2016). However, both P. arundinacea and P. australis are very competitive, and these plants often develop into the mono-dominant population in the natural ecological community by inhibiting the growth of other species (Adams and Galatowitsch 2005;Perkins and Wilson 2005;Kercher et al. 2007;Rodriguez and Brisson 2015).
These two species in North America were even considered invasive species and subsequently controlled by humans Bains et al. 2009;Uddin et al. 2012). In the twentieth century, P. arundinacea considerably increased its range in North America, becoming an invasive weed in many wetlands. As such, this species has been officially identified as a pest in nine American states (Lavergne and Molofsky 2004;Herr-Turoff and Zedler 2007). The distribution and abundance of P. australis in North America have dramatically increased over the past 150 years. The introduction of a non-native genotype of the common reed from other areas is reportedly responsible for the invasive characteristic of P. australis in North America (Saltonstall 2002). However, the two species with such strong competitive abilities are often chosen as dominant species and planted in the same constructed wetland which often operates for a long time, what will happen to the plant community in the constructed wetland?. Some studies reported that P. australis displaces P. arundinacea, although this phenomenon inconsistently occurs in systems where both plants are present (Vymazal and Kröpfelová 2005); even so, the displacement of P. arundinacea by P. australis takes a long time (Březinová and Vymazal 2014). Other studies reported that the two species in the Zhenjiang Waterfront Wetland in the lower reaches of Yangtze River in China could maintain a long-term and stable coexistence (Fu et al. 2011). Scholars have tried to elucidate the mechanisms underlying the long-term stable coexistence between the two species both in constructed and natural wetlands. Considering the extremely low light compensation point (46 µmol m −2 s −1 ) of P. arundinacea, this author previously concluded that P. arundinacea could compensate for substantial reduction in available light, and make use of the weak light within the lower part of the P. arundinacea-P. australis compound community, which can alleviate the competition between the two species for the light resource (Fu et al. 2011). Břzinova and Vymazal evidently concur with this conclusion (Březinová and Vymazal 2014).
Nevertheless, systematic and quantitative studies have yet note determine the compensatory light utilization rate (CLUR) of P. arundinacea within various P. arundinacea-P. australis compound communities with different densities of P. australis. In the present paper, based on the Pn-PPFD response curve fitting equation of P. arundinacea and the distribution of PPFD on the surface canopy of P. arundinacea population within those P. arundinacea-P. australis compound communities with three typical densities of P. australis found in the Zhenjiang Waterfront Wetland in the lower reaches of Yangtze River, this study measured the CLUR of P. arundinacea within such P. arundinacea-P. australis compound communities and identified a maintenance mechanism for the two species to obtain a relatively stable coexistence pattern with respect to the comprehensive utilization of light energy. On the basis of the above mechanism and on other work suggesting that a metapopulation could achieve stable coexistence through population management (Hastings 2003;Dushoff 2004), the P. australis-P. arundinacea compound community in the constructed wetlands could be regulated to maintain a long-term, stable coexistence between the two species.

Study site and study species
The Zhenjiang Waterfront Wetland (32°15 ′ N and 119°28 ′ E) is a 5000-ha intermittent river wetland along the Yangtze River that lies to the north of Zhenjiang City. The Zhenjiang Waterfront Wetland formed from the accumulation of massive silt deposits in the Yangtze River. The primary succession of the plant community in the wetland occurred with gradual elevation of the wetland beach. The vegetation in the wetland has successively passed through three community types over its succession process, namely, an initial community of P. arundinacea, an intermediate transitional community of P. arundinacea-P. australis, and a climax community P. australis. The P. arundinacea-P. australis transitional community took the longest time in the whole succession process, accounting for more than 50% of the total succession process time. During the succession process, the P. australis population gradually displaced the P. arundinacea population through an increase in the density of P. australis. This condition provides the ideal experimental material for us to study the compensatory light utilization characteristics of P. arundinacea within the P. arundinacea-P. australis compound community with different densities of P. australis.

Material selection
In the Zhenjiang Waterfront Wetlands in the lower reaches of Yangtze River, P. arundinacea often germinates in the beginning of February and matures by the end of May, with a growth period of about 130 days. Meanwhile, P. australis often germinates in mid-March and matures by the end of October, with a growth period of about 220 days. The two species have an overlap of about 80 days during the growth process (Fu et al. 2011). Experiments were performed on early May 2016, in which the plant heights of P. australis and P. arundinacea were about 3.0 and 1.5 m, respectively, and the shade effect of P. australis on P. arundinacea was most obvious. First, some P. arundinacea individuals from the P. arundinacea community that were not shaded by P. australis in the wetlands were chosen to study the Pn-PPFD response curve fitting equations of P. arundinacea at different time points in a day. Thereafter, three P. arundinacea-P. australis compound communities with P. australis densities of about 25, 35, and 45 shoot counts m −2 were selected to measure the photosynthetic photon flux density (PPFD) on the surface canopy of P. arundinacea within the lower part of the P. arundinacea-P. australis compound communitis having the three aforementioned densities of P. australis. Those three P. australis densities represent the different developmental stages of the P. arundinacea-P. australis compound communities, namely, (1) the semi-mature, (2) middle and late, and (3) the near-mature stages, respectively. The maximum density of the P. australis community at the fully mature stage is about 50 shoot counts m −2 in this wetland.

Pn-PPFD response curve equations
Measurements were conducted on a clear day (May 5). The third fully expanded leaf from the top of three uniform and vigorous individuals (as three repetitions) of P. arundinacea from the P. arundinacea community were chosen. The net photosynthetic rate -photosynthetic photon flux density (Pn)-PPFD response curve was measured once every 2 h from 8:00 to 16:00 using a LI-6400 Portable Photosynthesis System (LI-COR6400, USA). The Pn-PPFD response curves of P. arundinacea were determined across a PPFD coverage range of 0-2500 µmol m −2 s −1 using a light emitting diode (LED) light source and under ambient CO 2 concentrations. Thereafter, the optimal Pn-PPFD response curve equations of P. arundinacea at different time points were obtained.

Light extinction rates
With the ST285 automatic range illuminometer (produced by the Photoelectric Instrument Factory of Beijing Normal University), the PPFD on the surface canopy of P. arundinacea within the P. arundinacea-P. australis compound communities with three densities of P. australis were measured and abbreviated as PPFD' at 8:00, 10:00, 12:00, 14:00, and 16:00 on May 5. The PPFD on the surface canopy of P. arundinacea within the P. arundinacea community that was not under the shade of P. australis was also measured simultaneously and abbreviated as PPFD max . Ten values of PPFD for every surface canopy were recorded, and then the mean values were obtained. The light extinction rates (LERs) caused by the shade of the P. australis populations on the P. arundinacea populations at different time points were calculated as LER % = (1-PPFD'/PPFD max ) × 100.

Compensatory light utilization rate
With PPFD on the surface canopy of P. arundinacea within the compound community as an independent variable, the forecast values of Pn' (independent variable of PPFD') and Pn max (independent variable of PPFD max )of P. arundinacea at the different measurement time points were obtained on the basis of the Pn-PPFD response curve fitting equations of P. arundinacea at the corresponding measurement time points. The CLURs of P. arundinacea within the P. arundinacea-P. australis compound community were calculated as CLUR % = Pn'/Pn max × 100.
With this method, the CLURs of P. arundinacea within the P. arundinacea-P. australis compound communities with the three densities of P. australis at the different time points in a day were obtained.

Statistical analysis
We analyzed LERs and CLURs of P. arundinacea in the communities among the three densities of P. australis using repeated measures analysis of variance (rmANOVA). For the data of LERs and CLURs analyses, the mean ± S.E. were reported and the SPSS 17.0 statistical software was used by one-way ANOVA.LSD multiple comparison test was used at α = 0.05.

Pn-PPFD response curve and its fitting equation of P. arundinacea
The Pn-PPFD response curves of P. arundinacea at 8:00, 10:00, 12:00, 14:00, and 16:00 were obtained on the same day by using a LI-6400 Portable Photosynthesis System. The curves are plotted in Figure 1(A,B,C,D,E), respectively. The Pn-PPFD response curves of P. arundinacea showed similar trends among the five measurement time points. With increasing PPFD, Pn initially increased rapidly and then gradually peaked before gradually decreasing, and the Pn-PPFD response curves of P. arundinacea at all measurement time points showed a parabolic trend, which is suitable for fitting with quadratic function equations. The coefficient of determination (R 2 ) of the fitting equation at each measurement time point was more than 0.94 (Table 1), which showed that a satisfactory fit resulted. Therefore, the corresponding Pn of P. arundinacea at any PPFD (within the range of the fitted PPFD) could be calculated on the basis of the above Pn-PPFD response curve fitting equations.
Distribution of PPFD on the surface canopy of P. arundinacea and its LER The daily dynamics of the PPFD on the surface canopy of P. arundinacea growing in different environments are shown in Figure 2. All PPFD values showed single-peak curves, with peak values appearing at 12:00. However, the decline in PPFD on the surface canopy of P. arundinacea among the P. arundinacea-P. australis compound communities with the three densities of P. australis was considerably different from that on the surface canopy of P. arundinacea within the P. arundinacea community not under the shade of P. australis. The largest decline in daily average PPFD occurred within the compound community with a P. australis density of 45 shoot counts m −2 , with a decline of 82.4%. The daily average PPFDs of the communities with P. australis densities of 35 and 25 shoot counts m −2 declined by 55.3% and 45.4%, respectively. Figure 3 shows the LER caused by the shade of the P. australis populations on the P. arundinacea populations within the P. arundinacea-P. australis compound communities at the different measurement time points during the day. As shown by Figure 4, the LERs on the surface canopy of P. arundinacea within the P. arundinacea-P. australis compound communities increased to varying degrees along with the increase in the density of P. australis. When the densities of P. australis were 25, 35, and 45 shoot counts m −2 within the P. arundinacea-P. australis compound communities, the variation ranges of LER in a day were 41.11%-47.81%, 51.36%-56.50%, and 80.78%-83.39%, respectively. Statistical analysis showed that the LERs on the surface canopy of P. arundinacea within the compound community with a P. australis density of 45 shoot counts m −2 were significantly greater than those on the surface canopy of P. arundinacea within the compound community with P. australis densities of 25 (p = 0.016, 0.018, 0.019, 0.016 and 0.021 at 8:00, 10:00, 12:00, 14:00, and 16:00, respectively) and 35 (p = 0.032, 0.038, 0.041, 0.043, and 0.047 at 8:00, 10:00, 12:00, 14:00, and 16:00, respectively) shoot counts m −2 at each measurement time point; however, no significant differences in LERs were found between the two compound communities with P. australis densities of 25 and 35 shoot counts m −2 at each measurement time point (p = 0.082, 0.094, 0.092, 0.066, and 0.071 at 8:00, 10:00, 12:00, 14:00, and 16:00, respectively) of the day. In addition, no significant differences in LER were observed among the different time points in a day within the compound community having the same P. australis density (all p > 0.05).

CLURs of P. arundinacea within the compound community
The daily Pn dynamics of the surface canopy leaf of P. arundinacea growing in the different environments are shown in Figure 4. Similar to those of PPFD, the daily dynamics of the Pn of the surface canopy leaf of P. arundinacea growing in the different environments were all single-peak curves, although the peak values appeared at 10:00, one hour earlier than the appearance of PPFD peak values. With an increase in the density of P. australis within the compound community, the leaf Pn of P. arundinacea decreased to varying degrees at the same measurement time point in a day. As compared with that of the leaf of P. arundinacea within the P. arundinacea community not being shaded by P. australis, the daily average leaf Pn of P. arundinacea within the compound communities with P. australis densities of 25, 35, and 45 shoot counts m −2 declined by 25.3%, 33.5%, and 63.7%, respectively. The CLURs of P. arundinacea within the compound communities at different time points in a day are shown in Figure 5. The CLURs of P. arundinacea within the compound communities decreased to varying degrees with an increase in the density of P. australis. When the densities of P. australis were 25, 35, and 45 shoot counts m −2 within the compound communities, the daily average CLURs of P. arundinacea were 74.72%, 66.43%, and 36.18%, respectively; the daily CLUR variation ranged from 70.36% to 79.89%, 63.01% to 68.99%, and 33.27% to 39.98%, respectively; and the corresponding CLURs amplitudes were 9.53%, 5.98%, and 6.71%, respectively. The amplitude of the CLURs of P. arundinacea within the compound community with the same P. australis density was relatively small during the day. Statistical analysis showed that the CLURs of P. arundinacea within the compound community with a P. australis density of 45 shoot counts m −2 were significantly less than those of P. arundinacea within the compound community with P. australis densities of 25 (p = 0.017, 0.019, 0.022, 0.013, and 0.012 at 8:00, 10:00, 12:00, 14:00, and 16:00, respectively) and 35 (p = 0.033, 0.038, 0.041, 0.032, and 0.036 at 8:00, 10:00, 12:00, 14:00, and 16:00, respectively) shoot counts m −2 at each measurement time point. However, no significant differences in CLUR were found between the two compound communities with P. australis densities of 25 and 35 shoot counts m −2 at each measurement time points (p = 0.087, 0.091, 0.092, 0.062, and 0.089 at 8:00, 10:00, 12:00, 14:00, and 16:00, respectively) in a day. Finally, statistical analysis showed no significant differences in CLUR among the different measurement time points in a day within the compound community having the same P. australis density (all p > 0.05).

Discussion
Relationship between the CLUR of P. arundinacea and the density of P. australis within the compound community The characteristic of compensatory light utilization is prevalent in plant species possessing a very low light compensation point, and it not only often affects the vertical distribution patterns of plant species in some compound communities (Xue et al. 2008;Wang et al. 2015), but is also widely used to guide the stereoscopic cultivation of crops in agriculture (Lou et al. 2014;Wang et al. 2015). However, it is the density of plant species occupying the upper space in the compound community that ultimately determines the CLUR of plant species possessing very low light compensation points and occupying the lower space in the compound community. In the present paper, by varying the distribution of PPFD on the surface canopy of P. arundinacea and the subsequent Pn of P. arundinacea, we found that the density of P. australis determined the CLUR of P. arundinacea in the P. arundinacea-P. australis compound community. The CLUR of P. arundinacea decreased to varying degrees with an increase in the density of P. australis within the compound community. However, the downward trend in the CLUR of P. arundinacea did not decrease uniformly. At first the CLUR decreased slowly; only when the density of P. australis reached a certain threshold did the CLUR of P. arundinacea within the compound community sharply decrease. This explains the phenomena that accelerated displacement of P. arundinacea population by P. australis population appeared in the later stage of population displacement in some natural wetlands (Fu et al. 2007) and constructed wetlands (Březinová and Vymazal 2014).
Relationship between the regulation of plant community and the density of P. australis in the constructed wetland Another study showed that P. australis tended to displaces P. arundinacea in some constructed wetlands, although this phenomenon inconsistently occurred in systems where both plants were present (Vymazal and Kröpfelová 2005). We note that Vymazal and Kröpfelová did not mention the density of P. australis, so it's reasonable to suspect that the inconsistencies in observed plant displacement were due to the difference in the densities of P. australis within the P. arundinacea-P. australis compound community among the different constructed wetlands. On the basis of the above results, we can deduce that some moderate or greater density of P. australis must exist, and the CLUR of P. arundinacea is high enough to maintain its healthy growth within the P. arundinacea-P. australis compound community. In the present study, when the densities of P. australis within the compound communities were 25 and 35 shoot counts m −2 , the CLURs of P. arundinacea still maintained high levels with 74.72% and 66.43%, respectively. On the basis of the growth of P. arundinacea, P. arundinacea within the compound communities with the above two densities of P. australis were able to maintain healthy or relatively healthy growth conditions. In addition, during the course of the study, P. arundinacea was in the later grain-filling period.
Therefore, the shade resulting from P. australis can exert a certain degree of adverse impact on its grain filling; however, for a cloned plant such as P. arundinacea whose main mode of reproduction is rhizome propagation rather than seed propagation (Martina and Ende 2013;Chen et al. 2014;Martínková and Klimešová 2017), the aforementioned adverse impact on its grain filling will not significantly affect the normal growth and development of the P. arundinacea population. By contrast, when the density of P. australis within the compound communities rose to 45 shoot counts m −2 , the CLUR of P. arundinacea fell to as low as 36.18%. Furthermore, P. arundinacea individuals within the compound community were no longer able to maintain healthy or even relatively healthy growth conditions, and some growth stress traits, such as the yellow leaves, thin stems, low density, and few branches, generally occurred. Therefore, the density of P. australis within the compound community should be controlled at 35 shoot counts m −2 or slightly higher but must not reach 45 shoot counts m −2 to maintain the stable coexistence of the two species in the P. arundinacea-P. australis compound community in the Zhenjiang Waterfront Wetland of the lower reaches of the Yangtze River. However, this recommended density of P. australis applies only to constructed wetlands where the cultivar of planted P. australis must be the same as that of P. australis growing in the Zhenjiang Waterfront Wetland. In the studied wetland, the maximum density of the P. australis community at the fully mature stage was only about 50 shoot counts m −2 , while the density of P. australis planted in the some constructed wetlands in Europe ranged from 11 to 332 (Vymazal and Kröpfelová 2005). The huge difference in the density of P. australis in the different regions mainly resulted from the difference in the cultivars of planted P. australis. Therefore, the difference in cultivars should be taken into account when the appropriate density of P. australis maintaining a stable P. arundinacea-P. australis compound community in the constructed wetlands is recommended. In addition, studies have shown that Figure 3. LER caused by the shade of the P. australis populations on the P. arundinacea populations at different time points in a day. Roman numerals I, II, and III stand for the P. arundinacea-P. australis compound communities with P. australis densities of 25, 35, and 45 shoot counts m −2 , respectively. CK refers to the P. arundinacea community not being shaded by P. australis. Error bars are ± standard error, n = 10. PPFD-photosynthetic photon flux density. Figure 4. LERs caused by the shade of the P. australis populations on the P. arundinacea populations at different time points in a day. Roman numerals I, II, and III stand for the P. arundinacea-P. australis compound communities with P. australis densities of 25, 35, and 45 shoot counts m −2 , respectively. Different small letters indicate significant difference between the communities with three densities of P. australis according to LSD test (p < 0.05). Error bars are ± standard error. LER-light extinction rates.
lowering light availability with a cover crop in restored prairie pothole wetlands might slowP. Arundinacea invasion in some regions where P. arundinacea was regarded as an invasive species, but its effect is extremely limited Galatowitsch 2004, 2006), and this outcome is precisely due to the characteristic of the compensatory light utilization of P. arundinacea.
Other factors affecting the stable coexistence of P. arundinacea and P. australis in the compound community The growth and development of plants are influenced by plant-intrinsic factors and external environmental factors. Likewise, the stable coexistence of P. arundinacea and P. australis in the Zhenjiang Waterfront Wetland must be the result of the combined effect of many factors, including the density effect of P. australis. In addition to the density effect of P. australis, many other obvious differences were observed between the two species, including the growth period, plant height, light compensation point, and root depth (Lavergne and Molofsky 2006;Zhang and Luo 2008;Fu et al. 2011;Ge et al. 2011;Nelson and Anderson 2015). In particular, Fu Weiguo suggested that despite the stronger competitiveness of P. australis compared with P. arundinacea, the latter's earlier germination of about 30 days resulting from the difference in growth period may have allowed this species to establish rapidly and preempt the establishment of P. australis, and the absence of shading by P. australis would also have been very beneficial for the early growth of P. arundinacea during this period Fu et al. 2011). In addition, some scholars have pointed out that an obvious niche segregation exists between the two species, which could reduce the interspecific competition and promote the sharing of various resources between them (Fu et al. 2015). This paper suggests a maintenance mechanism by which the two plant species can achieve a relatively stable coexistence, by considering the density effect of P. australis on the ability of P. arundinacea to compensate for reduced light levels.

Conclusion
P. arundinacea can compensate for reduced light and make use of the weak light within the lower part of the P. arundinacea-P. australis compound community, provided that there are reasonable limits on the density of P. australis. The density of P. australis within the compound community determined the CLUR of P. arundinacea. In addition, the amplitude of the CLURs of P. arundinacea within the compound community with the same P. australis density was relatively small during the day. This result indicates that the compensatory light utilization of P. arundinacea functions better throughout the day. Finally, regardless of which cultivar of P. australis is selected, an appropriate density range of P. australis within the P. arundinacea-P. australis compound community is required to maintain the relative stability of the compound community according to the ability of P. arundinacea to effectively utilize weak light. Nevertheless, the difference in cultivars should be taken into account when the appropriate densities of P. australis are determined.