Effects of Tropical Cyclone Passage on Plankton Community Respiration in a Phosphate-Limited Freshwater Ecosystem

© 2023 Academia Sinica, Taiwan Open Access Effects of Tropical Cyclone Passage on Plankton Community Respiration in a Phosphate-Limited Freshwater Ecosystem Erica Silk P. dela Paz1, Chao-Chen Lai2, Fuh-Kwo Shiah3, Chih-hao Hsieh3,4,5 , Chia-Ying Ko1,5,* 1Institute of Fisheries Science, National Taiwan University, Taipei, 106 Taiwan. *Correspondence: E-mail: [email protected] (Ko) E-mail: [email protected] (dela Paz) 2Department of Science Education, National Taipei University of Education, Taipei 106, Taiwan. E-mail: [email protected] (Lai) 3Research Center for Environmental Changes, Academia Sinica, Taipei 115, Taiwan. E-mail: [email protected] (Shiah) 4Institute of Oceanography, National Taiwan University, Taipei 106, Taiwan. E-mail: [email protected] (Hsieh) 5Master’s Program in Biodiversity, Institute of Ecology and Evolutionary Biology, National Taiwan University, Taipei, 106 Taiwan Received 22 August 2023 / Accepted 22 November 2023 / Published 29 December 2023 Communicated by Benny K.K. Chan Plankton community respiration (CR) in aquatic ecosystems varies with environmental factors, which could be altered during tropical cyclones (TCs). A potential increase in CR resulting from the effects of TCs is generally understudied. Here, we examined the relationship between plankton CR and environmental factors, including during TC-affected periods, in a phosphate-limited freshwater ecosystem. We conducted an intensive in situ sampling in Fei-Tsui Reservoir (FTR) from January 2010 to December 2015 during TC periods and non-TC periods. Our results showed a consistent temporal pattern that plankton CR increased between March to October and declined between November to February. These changes in plankton CR, primarily supported by bacterial biomass, were positively influenced by euphotic depth-averaged temperature. The CR also significantly increased with euphotic depth-averaged NO2 - concentrations and decreased with euphotic depth-averaged NO3 - concentrations. These results indicated that these factors typically influenced CR dynamics in the FTR. During TC periods, plankton CR was increased further due to a higher and ideal euphotic depth-averaged temperature (23–27°C) and increased supply of limiting nutrient resources via stream runoff. Overall, this study showed that a TC positively influences plankton CR by creating favorable water conditions. Notably, with a higher frequency of intense TCs projected for the Western North Pacific in most climate change scenarios, the impact of TCs on CR may increase in the near future. Key words: Bacterial biomass, Total community respiration, Oligotrophic lake, Plankton, Typhoon Citation: dela Paz ESP, Lai CC, Shiah FK, Hsieh Ch, Ko CY. 2023. Effects of tropical cyclone passage on plankton community respiration in a phosphate-limited freshwater ecosystem. Zool Stud 62:58. doi:10.6620/ZS.2023.62-58. BACKGROUND Severe weather events (e.g., storms, tornadoes, thunderstorms, heatwaves, droughts, wildfires) can cause considerable ecosystem challenges. Severe weather events impact some aquatic ecosystems, leading to noticeable negative effects on productivity, water quality, and biodiversity, among other indicators (Poff et al. 2002; Field et al. 2007). For instance, tropical cyclones (TCs), strong weather disturbances that develop as a result of warm air rising over the Pacific Ocean, are regularly observed in the Western North Pacific about 30 times each year (Emanuel 2005; Vecchi and Soden 2007; Ying et al. 2012; Chen et al. 2020). They are characterized by heavy precipitation and strong winds that influence terrestrial areas and have impacted various ecological relationships in lakes (Ko et al. 2016 2017; Huang et al. 2022). Yet, lakes have demonstrated Zoological Studies 62:58 (2023) doi:10.6620/ZS.2023.62-58 1 © 2023 Academia Sinica, Taiwan varying degrees of resilience to TCs and have developed ways to cope with and recover (i.e., nitrogen cycling) from the disturbance (Gao et al. 2021). Although aquatic ecosystems may show resilience in the face of TCs, there is still variability in the subsequent impacts due to factors such as precipitation rate (Jansson et al. 2008; Ko et al. 2016). Precipitation associated with TCs increases water inflow into lakes, causing a rapid rise in water levels and increased sedimentation in the lake, which impacts the phytoplankton and zooplankton communities (Iavorivska et al. 2016; Hoover et al. 2006; LópezLópez et al. 2012; Ko et al. 2016). Additionally, stream runoff from TCs can increase the phosphate (PO4 3-) supply to a nutrient-limited system (Tseng et al. 2010). Furthermore, stronger precipitation has more dramatic effects on aquatic systems and can disrupt thermal stratification (Klug et al. 2012; Doubek et al. 2021; Lin et al. 2022). Thus, TCs accompanied by heavy precipitation have the potential to significantly impact aquatic ecosystems and their constituents (e.g., species and water conditions). Although other freshwater biological communities have been studied in the wake of a TC passage, the effects of TCs on plankton community respiration (CR) in this regard have not been studied. In aquatic ecosystems, plankton CR plays a critical role in carbon cycling, but is sensitive to changing environmental factors, particularly the limiting ones (Del Giorgio and Williams 2005). For example, warmer waters could reduce some enzymatic reactions, resulting in a higher respiration rate (Hall and Cotner 2007; Scofield et al. 2015), while more acidic water may increase cell membrane diffusion, which destabilizes the cell membrane of aquatic species (Russell 1992; Jin and Kirk 2018). Moreover, an increase in the concentrations of nitrogen compounds (i.e., NO2 - and NO3 -), PO4 3-, dissolved oxygen (DO), and dissolved organic carbon (DOC) increases CR (Bueno et al. 2012; Vikström and Wikner 2019). Additionally, CR is represented by the cumulative carbon consumption of constituents such as bacterioplankton, phytoplankton, planktonic protozoa, and zooplankton (Liu et al. 2000), contributing to some of the variability of CR. TCs, as a severe force, can rapidly alter plankton community and environmental conditions, potentially introducing limiting factors that can enhance CR. However, the effects of TCs on the relationship between species’ biological activities and the environment are largely unexplored. The aim of the present study was to examine the relationship between plankton CR and environmental factors, including in TC-affected periods, in the FeiTsui Reservoir (FTR), a phosphate-limited freshwater ecosystem. We hypothesized that (1) plankton CR dynamics resulted from environmental changes in the FTR and (2) TCs affect environmental factors that lead to an increase in plankton CR rate. To test these hypotheses, we analyzed a 6-year intensive in situ monitoring dataset from January 2010 to December 2015. Understanding how strong weather events affect this relationship involving plankton CR can help us better understand how nutrient-limited freshwater ecosystems are impacted, which is important for regulating and mitigating TC effects in lakes. MATERIALS AND METHODS Study site Fei-Tsui Reservoir is a phosphate-limited freshwater reservoir in northern Taiwan (Tseng et al. 2010) (Fig. 1). The catchment area is approximately 303 km2, and the water depth at the dam site is 90–120 m (Chow et al. 2017). It was created to provide drinking water to Taipei City and is surrounded by secondary forests and tea plantations (Ko et al. 2016; Chow et al. 2017). FTR flows to two adjacent streams: the upper stream of the Beishi River and the lower streams of the Xindian River and Nanshi River. Study design We used an intensive in situ monitoring dataset to investigate the TC effects on CR between the averaged depth of 0 to 20 m, i.e., euphotic depth, in the FTR. Before noon, sampling was conducted at the dam site bi-weekly from January 2010 to December 2015, except (i) weekly sampling from July 2012 to September 2012 and July 2014, (ii) monthly sampling in August 2010, October 2010, February to July 2011, October 2011, April 2012, May 2013, October 2013, April 2014, and April 2015, and (iii) no sampling between November and December in 2010 and 2011. Our sampling involved collections of physical parameters (i.e., temperature, pH), DO, bulk water samples for nutrients (NO2 -, NO3 -, and PO4 3-), bacteria, chlorophyll-a, and DOC analyses, which were taken from 10 depth intervals (0, 2, 5, 10, 15, 20, 30, 50, 70, and 90 m) using 5 L Go bottles (General Oceanics, Miami, FL), and zooplankton samples. All these variables were analyzed as described in the following section. Sampling and laboratory analyses Physical measurements Physical properties of water, including vertical page 2 of 14Zoological Studies 62:58 (2023) © 2023 Academia Sinica, Taiwan profiles of temperature, pH, and depth, were measured at the time of the collection from the surface to near bottom (90 m depth) by multiparameter sensor equipment (CTD, Idronaut, Brugherio (MB) Italy). Precipitation data, which is 7 days of accumulated precipitation before sampling, were obtained from the Taiwan Central Weather Bureau (CWB, https://www. cwb.gov.tw). These variables represented external factors that impact CR. DOC concentrations DOC was used to represent the energy source for heterotrophic organisms, usually derived from decomposing organic matter. Water samples of 300–500 ml were filtered in the laboratory through Whatman GF/F glass microfiber filters (0.7 µm pore size, Whatman, GE Healthcare Life Sciences, Little Chalfont, United Kingdom). DOC concentrations from the filtrates were determined by acidulating a 60 ml subsample with 0.5 ml of 80% H3PO4 and sprayed with 350 ml min-1 of CO2-free O2 for at least 10 minutes. Then, the concentration of DOC was analyzed by a Shimadzu TOC-5000 high-temperature catalytic oxidation analyzer. Nutrients in the water Dissolved inorganic nutrients such as NO2 -, NO3 -, and PO4 3were used to represent the resources available for the metabolic demands of plankton. Concentrations of NO2 -, NO3 - and PO4 3from a filtrate subsample of 60 ml were analyzed by a brucine sulfate method and a molybdenum blue method using a 10-cm detection cell in a fabricated flow injection analyzer (FIA) with a cadmium-copper column with material detection limits of the three nutrients (0.05 µM, 0.05 µM, and 0.03 µM, respectively) (Parsons et al. 1984). Plankton community respiration CR was determined using the DO method. We measured the initial DO concentrations from 2, 10, and 20 m depths using spectrophotometry with a precision ratio of 0.5 μM following the standard methods of Pai et al. (1993). Then, duplicate 350 ml subsamples were incubated in BOD bottles per depth for 24–48 hours in a dark chamber at in situ temperature (±1°C) (see Fig. 1. Map of the study area, including (a) Taiwan’s geographical location, (b) Fei-Tsui Reservoir, and (c) the dam site. N page 3 of 14Zoological Studies 62:58 (2023) © 2023 Academia Sinica, Taiwan Chen et al. 2006). Next, 60 ml subsamples were taken to estimate DO concentrations (see Chen et al. 2007). Finally, CR was calculated from the difference between the initial DO and the DO concentrations treated in the dark chamber. Bacterial biomass Bacterial biomass was used to represent the abundance of microbial communities actively metabolizing and respiring in the FTR. Bacterial samples of approximately 100 ml were fixed with 0.1% glutaraldehyde and kept in a container at -80°C with liquid nitrogen. Then, samples were stained by SYBRGREEN I and incubated at room temperature in the dark for 15 minutes. Stained samples were analyzed through a flow cytometer of CyFlow® Space (PARTEC) with < 1000 events sec-1 rate. To obtain bacterial biomass, bacterial cell count was transformed to carbon mass using a conversion factor of 20 fgC per cell (Ducklow and Carlson 1992). Chlorophyll-a concentrations Chlorophyll-a (Chl-a) served as an indicator of phytoplankton abundance and biomass in the FTR. The chl-a samples on the GF/F filter, collected from 300–500 ml filtered water samples, were extracted using acetone, and the concentrations were estimated using an in vitro fluorometer (TD-700 Laboratory Fluorometer, Turner Designs, Sunnyvale, CA, USA) following the standard protocols provided by Parson et al. (1984). Zooplankton biomass Zooplankton biomass was used to represent standing stocks of actively respiring and metabolizing zooplankton (Uye et al. 1998; Gonçalves et al. 2015). Zooplankton samples were collected by filtering approximately 1000 ml of depth-integrated samples between the surface and 50 m depth through a 50 µm Norpac net with a net mouth radius of 0.225 m and attached with a HYDRO-BIOS flowmeter with backrun stop, Model 438-115 (HYDRO-BIOS, Altenholz, Germany). A previous observation revealed that zooplankton density was significantly lower at depths less than 50 m. The results of net hauling indicated a filtration efficiency of 53% based on water volume (net opening area * hauling depth) (Chang et al. 2014). In the laboratory, zooplankton samples were split, and each subsample was condensed to 120 ml with a CO2 effervescing fixative agent, stored at 4°C for an hour, and preserved with a 2% formalin solution. Then, zooplankton samples were taxonomically identified, their lengths were measured, and their biomass was estimated. Using a stereomicroscope (Olympus SZX16 AnalSIS®) with an attached camera (OlympusDP71), we analyzed 300 zooplankton individuals per water sample. Species of zooplankton such as Rotifera, Cladocera, and Copepoda were identified using taxonomic keys by Li (2005), Wang (1961), Chiang and Du (1979), Korovchinsky (2000), Tuo and Young (2002 2011) and Shen et al. (1979). Subsequently, the length and width of Rotifera, the size of Cladocera, and the prosomal and urosomal length of Copepoda were measured using a charge-coupled device (CCD) imaging system. To obtain the biomass, each taxon-specific dry weight was calculated, and the resulting values were converted to organic carbon of 0.48 (Dumont et al. 1975; Bottrell 1976; Ruttner-Kolisko 1977; Pauli 1989; Pace and Orcutt 1981; Andersen and Hessen 1991; EjsmontKarabin 1998; Michaloudi 2005; de Azevedo et al. 2012). Finally, in this study, we used zooplankton biomass with a body size of < 177 μg (hereafter small zooplankton biomass) to better quantify the relationship of plankton CR with zooplankton. Tropical cyclones To accurately measure TC effects on FTR, only TCs with their centers impacting the Taiwan region were considered in this study. Considering the period of time for phytoplankton and bacteria to respond to an abrupt nutrient injection by TCs (Collos 1986; Arteaga et al. 2020), a TC period is defined in this study as the day the TC entered the Taiwan region up to 7 days after the last day the TC was over Taiwan. FTR was impacted by 24 TCs between 2010 and 2015, considering TC Tembin as two independent TCs as its track crossed twice into the Taiwan region (Table 1). TC residence time was determined based on the days or duration of stay of TCs within Taiwan, and data were obtained from Ventusky Meteorological (https://www.ventusky. com), supported by Deutscher Wetterdienst (DWD) and National Oceanic and Atmospheric Administration (NOAA). The minimum center pressures and maximum sustained wind speeds of individual TCs were acquired from the CWB. TC intensity was determined based on maximum sustained wind speed classified as minor, 17.2–32.6 m s-1; moderate, 32.7–50.9 m s-1; or intense ≥ 51.0 m s-1 (Ko et al. 2016). Statistical analyses A trapezoidal rule was used to calculate averaged euphotic depth-integrated values (i.e., 0–20 m depthaveraged) to allow comparisons between each variable of the water column, including temperature, DO, DOC page 4 of 14Zoological Studies 62:58 (2023) © 2023 Academia Sinica, Taiwan concentrations, nutrient concentrations, plankton CR, bacterial biomass concentrations, and chl-a concentrations (Hornbeck 1975). Small zooplankton biomass was set to the depth-averaged values between 0–50 m. Data were tested for equal variances and normality using Levene’s and Shapiro-Wilk tests, respectively (Table S1). Factors were analyzed for autocorrelation and a 0.6 threshold estimation was used to select factors for this study (Table S2). After autocorrelation tests, the selected variables were euphotic depth-averaged temperature, DOC, DO, NO2 -, NO3 -, PO4 3-, chl-a, and 0–50 m depth-averaged small zooplankton. To test Hypotheses 1, which provides a baseline connection between CR and environmental factors in the FTR, we examined significant relationships between the factors and CR using simple linear regression. Also, to estimate which biological factors greatly contributed most to the plankton CR, the relationships of bacterial biomass, chl-a, and small zooplankton biomass concentration versus plankton CR were tested using Pearson’s correlation test. To test Hypothesis 2, we analyzed the relationship between the factors and CR using simple linear regression during TC and nonTC periods and compared their change of rate using slope values. To accurately identify the impact of TCs on resource availability, we compared the factors during TC and non-TC periods using the data within May– November. In all hypothesis testing procedures, the 5% significance level (based on a = 0.05) was used. RESULTS Time-series pattern and variation of plankton CR in the Fei-Tsui Reservoir Plankton CR gradually increased between March–October and decreased between November– February (Fig. 2). Moreover, using all datasets, CR had significantly increased in TC periods (2.28 ± 1.03 mM/d) compared to non-TC periods (1.69 ± 1.07 mM/d) (p = 0.01, Fig. 2). But, CR dynamics analyzed using only the May–November dataset showed no significant difference between TC and non-TC periods (p > 0.05, Fig. S1). The results revealed the complex dynamics of plankton CR in the context of TCs and highlighted the need for further analysis to explore the underlying mechanisms. Table 1. List of Pacific tropical cyclones that hit Taiwan from 2010 to 2015 and their characteristics Year Name Residence time Minimum center pressure (hPa) Maximum sustained wind speed (m/s) Intensity grade* 2010 Fanapi 1 (19 September) 940 45 Moderate Megi 2 (22–23 October) 935 48 Moderate 2011 Nanmadol 1 (28 August) 920 53 Intense 2012 Talim 3 (19–21 June) 985 26.7 Minor Doksuri 2 (28–29 June) 992 20.1 Minor Saola 5 (30 July–3 August) 960 37.9 Moderate Haikui 2 (6–7 August) 965 33.5 Moderate Kai-Tak 2 (14–15 August) 970 33.5 Minor Tembin 5 (21–25 August) 945 45 Moderate Tembin 3 (27–28 August) 965 35 Moderate Jelawat 2 (27–28 September) 910 55.9 Intense 2013 Soulik 2 (12–13 July) 925 51.4 Intense Usagi 2 (21–23 September) 910 55.9 Intense Fitow 2 (5–6 October) 960 37.9 Moderate Krosa 3 (1–3 November) 970 37.9 Moderate 2014 Hagibis 3 (13–15 June) 996 20.1 Minor Matmo 2 (22–23 July) 965 35.8 Moderate Fung-wong 3 (20–22 September) 985 22.4 Minor 2015 Noul 1 (11 May) 920 55.9 Intense Linfa 2 (10–11 July) 980 26.8 Minor Soudelor 2 (7–8 August) 900 58 Intense Goni 2 (22–23 August) 930 51.4 Intense Dujuan 3 (27–29 September) 925 55.9 Intense Koppu 3 (19–21 October) 925 51.4 Intense *The intensity grade was based on the maximum sustained wind speed near the center: minor, < 32.7 m/s; moderate, 32.7–50.9 m/s; intense, > 50.9 m/s. page 5 of 14Zoological Studies 62:58 (2023) © 2023 Academia Sinica, Taiwan Relationship between plankton CR and environmental factors Simple linear regression analysis showed that euphotic depth-averaged temperature had a strong positive influence on CR in FTR (p < 0.001, Fig. 3a). Euphotic depth-averaged DO and DOC did not show a significant influence on CR (both p > 0.05, Fig. 3b–c). Euphotic depth-averaged NO2 - and NO3 - strongly influenced CR; the former had a positive influence, and the latter had a negative influence (both p < 0.001, Fig. 3d–e). Euphotic depth-averaged PO4 3did not show a significant influence on CR (p = 0.10, Fig. 3f). These results indicated that CR variability in the FTR was primarily influenced by euphotic depth-averaged temperature and nutrients (NO2 - and NO3 -), among other environmental variables. The different planktonic components’ contributions to CR are shown in figure 4. CR was highly positively correlated with bacterial biomass (p < 0.001, Fig. 4a), while it was weakly correlated with chlorophyll a (p = 0.09, Fig. 4b) and small zooplankton biomass (p = 0.37, Fig. 4c). Thus, plankton CR was supported by bacterial biomass in the FTR. Relationship of environmental factors and plankton CR in TC and non-TC periods Temperature substantially influenced CR during the TC and non-TC periods (Fig. 5), complementing the findings above. An increasing euphotic depth-averaged temperature enhanced CR rate in TC periods and in nonTC periods using all datasets and the May–November dataset (all p < 0.05, Fig. 5a); however, a higher value of regression slope was found in TC periods (b = 0.48, Fig. 5a) compared to non-TC periods (both b = 0.15, Fig. 5a). Accumulated precipitation during the TC periods had no direct effect on CR (p = 0.59, Fig. 5b). Changes in environmental factors between TC and non-TC periods Accumulated precipitation was significantly higher during TC periods than non-TC periods, confirming that TCs were accompanied by heavy precipitation (p = 0.002, Fig. 6a, Fig. S2a). Euphotic depth-averaged DOC concentration was higher during the TC period, but this change was not significant (p = 0.28, Fig. 6b, Fig. S2d). Euphotic depth-averaged NO2 - and NO3 - did not significantly change between TC and non-TC periods (p = 0.26 and p = 0.42, Fig. 6c–d, Fig. S2e–f). Euphotic depth-averaged PO4 3concentration was limited and did not significantly change between TC and non-TC periods (p = 0.35, Fig. 6e, Fig. S2g). DISCUSSION We examined how plankton CR is affected by environmental factors, particularly during periods when TCs are impacting FTR, a phosphate-limited freshwater ecosystem. Plankton CR was typically affected by euphotic depth-averaged temperature, NO2 -, and NO3 - in the FTR, supporting Hypothesis 1. It was found that bacteria significantly contribute to CR, among other biological factors. Moreover, during TC periods, the increase in plankton CR was higher than usual, attributed to a higher and ideal euphotic depth-averaged temperature, supporting Hypothesis 2. Furthermore, the concentration of limiting nutrient factors was supplied by the indirect effect of precipitation and ultimately increased the CR. These findings show that plankton CR in the FTR alters due to environmental changes. With the TC effects, CR tends to increase as the water conditions during TC periods are favorable. This implies that strong weather events indirectly affect biological activities in nutrient-limited freshwater ecosystems. Fig. 2. Timeseries and boxplot displaying differences in euphotic depth-averaged plankton community respiration in tropical cyclone (TC) and nonTC periods in the Fei-Tsui Reservoir from January 2010 to December 2015. The euphotic depth spans from 0–20 m on average. Each bar represents individual sampling. Different bar colors indicate the TC and Non-TC periods. periods. The p-value estimated by the Mann-Whitney U test is shown. page 6 of 14Zoological Studies 62:58 (2023) © 2023 Academia Sinica, Taiwan Factors affecting plankton CR dynamics Euphotic depth-averaged water temperature is a significant environmental factor in regulating plankton CR in the FTR, supporting Hypothesis 1. We find that euphotic depth-averaged water temperature positively influenced the seasonal variation of CR, which is consistent with many studies (Hall and Cotner 2007; Scofield et al. 2015; Pace et al. 2021; García et al. 2023). This information has significant implications for how biological activities respond to increasing global warming. Additionally, our study further Fig. 3. Relationships between euphotic depth-averaged plankton community respiration versus (a) euphotic depth-averaged temperature, (b) dissolved oxygen (DO), (c) dissolved organic carbon (DOC), (d) NO2 -, (e) NO3 -, and (f) PO4 3concentrations in the Fei-Tsui Reservoir from January 2010 to December 2015. The euphotic depth spans from 0–20 m on average. Black and white circles represent tropical cyclone (TC) and non-TC periods, respectively. Estimated R2 and p-values by simple linear regression analysis are shown. The solid line indicates a significant relationship (p ≤ 0.001) and the broken line indicates no significant relationship (p > 0.05). page 7 of 14Zoological Studies 62:58 (2023) © 2023 Academia Sinica, Taiwan Fig. 5. Relationships between euphotic depth-averaged plankton community respiration versus (a) euphotic depth-averaged temperature during tropical cyclone (TC) periods and non-TC periods (all datasets and May–November dataset) and (b) precipitation during TC periods in the Fei-Tsui Reservoir. The euphotic depth spans from 0–20 m on average. Black and white circles represent TC and non-TC periods, respectively. Estimated R2 and p-values by simple linear regression analysis are shown. The solid line indicates a significant relationship (p ≤ 0.05). Fig. 4. Relationships between euphotic depth-averaged plankton community respiration versus (a) euphotic depth-averaged bacterial biomass, (b) chlorophyll a, and (c) log10 0–50 m depth-averaged small zooplankton biomass in the Fei-Tsui Reservoir from January 2010 to December 2015. The euphotic depth spans from 0–20 m on average. Black and white circles represent tropical cyclone (TC) and non-TC periods, respectively. The correlation coefficient (r) and p-values estimated by Pearson’s correlation test are shown. page 8 of 14Zoological Studies 62:58 (2023) © 2023 Academia Sinica, Taiwan demonstrated that plankton CR may be enhanced during TC passages. Previous findings showed that TCs from the Pacific Ocean typically crossed and/or hit Taiwan between May and November (Ko et al. 2017), with increased euphotic depth water temperatures during this period; this explains a higher plankton CR. Given the significant relationship of temperature with plankton CR, understanding the dynamics of these factors is crucial for predicting CR variability during TC periods. Therefore, additional analyses were conducted to examine the interactions between CR and water temperature and their implications for ecosystem Fig. 6. Boxplots comparing (a) precipitation, (b) euphotic depth-averaged dissolved organic carbon (DOC), (c) NO2 -, (d) NO3 -, (e) PO4 3concentrations between tropical cyclone (TC) periods and non-TC periods using the May–November dataset. The euphotic depth spans from 0–20 m on average. Estimated p-values by the Mann-Whitney U test are shown. page 9 of 14Zoological Studies 62:58 (2023)