Physiological consequences of biologic state and habitat dynamics on the critically endangered Yangtze finless porpoises (Neophocaena asiaeorientalis ssp. asiaeorientalis) dwelling in the wild and semi-natural environment

This research describes the effect of age, sex, reproductive status and habitat location on serum biochemical analytes in the critically endangered Yangtze finless porpoise (Neophocaena asiaeorientalis ssp. asiaeorientalis). Results highlight conservation relevant effects of biologic stages and suggests that without proper management habitat degradation may adversely affect their long-term sustainability.


Introduction
Effects of climate change and population growth on global freshwater reserves are increasing to such an extent that in some locations this valuable resource is shrinking, and in some instances, may disappear (Coe and Foley, 2001;Oren et al., 2010). Poyang Lake (PL), the largest freshwater lake in China and an extension of the Yangtze River, is an important habitat for critically endangered Yangtze finless 1 porpoises (YFPs, Neophocaena phocaenoides asiaorientalis; Wang et al., 2013). In addition to significant water loss from a prolonged drought (Mei et al., 2015;Zhang et al., 2016), this crucial habitat is under increasing anthropogenic pressures in the form of sand dredging (Li, 2008), vessel trafficking (Zhang, 2007) and water quality degradation due to Yangtze River watershed alterations (Wu et al., 2004;Mueller et al., 2008). In addition to direct impacts on water quality, habitat fragmentation from these alterations may only increase with future hydroelectric power plant projects (Jiangxi Water Conservancy, 2010). Such changes in the Yangtze River are believed to have played a part in the rapid decrease of the YFP population numbers, which have been declining at the average rate of 13.73% per annum (Mei et al., 2014). These declines have resulted in the latest estimated total YFP population numbers to be around 1000 individuals, with almost half (~450) of the living population in PL (Mei et al., 2014). With the current population trajectory, the YFP may soon join the baiji (Lipotes vexillifer), a previous co-inhabitant of the Yangtze River, in extinction (Turvey et al., 2007).
Worldwide deleterious alterations in habitat quality are believed to be the main cause of almost 40% of all mammalian species experiencing severe population declines (Ceballos et al., 2017;Johnson et al., 2017) with freshwater cetaceans, or river dolphins, appearing to be extremely vulnerable toward these pressures He et al., 2017). Habitat destruction in the biodiversity-rich area around the globe can cause mass extinction, especially for threatened and endemic species (Brooks et al., 2002). Therefore, determining current physiological characteristics of individuals within an at-risk population provides baselines for future investigations on the potential effect of anthropogenic activities on their health and well-being (Schwacke et al., 2009). It is well accepted that habitat quality effects body fitness and physiological functions of the animals living within these locations (Pulliam, 1988;Huey, 1991;Carey, 2005). Therefore, monitoring of physiologic traits can help us to understand and predict organismal and population responses to environmental change and stressors, causeeffect relationships, and specificity of management techniques (Christine et al., 2016).
While the population of YFP in PL represents almost half of the remaining animals living within their historic range, an experimental conservation strategy involving relocation of a small number of animals into a natural ex situ reserve called the Tian-E-Zhou Oxbow (TZO) began in 1990. The use of this reserve, a habitat removed from many of the detrimental anthropogenic activity faced by in situ YFP, if successful over the long-term, would serve as a model for the identification of other natural habitats which could provide shelter for relocated YFP until the natural habitat of the Yangtze River could be sufficiently reclaimed for animal reintroduction. While the use of ex situ populations as a conservation strategy has been successfully implemented in terrestrial mammals, it represents the first application of this model toward a cetacean (O'Brien and Robeck, 2010). The initial success of the ex situ TZO population has been demonstrated by its increased from a net introduction of five animals from the wild since 1990 to a total of 25 in 2010. From 2010 to 2015, due to successful animal breeding, the population exceeds to more than 60 individuals with a net 108% increase in the population (Wang, 2009;2015). This growing ex situ population, which has been living in a unique environment, may serve as a physiologic control for changes that may have occurred in wild populations of the Yangtze River, Poyang and Dongting Lake system. However, until recently (Nabi et al., 2017a), most efforts at evaluating the health of the TZO ex situ population has focused on their reproductive health and resource management (Wu et al., 2010;Zhao et al., 2010;Zeng et al., 2018). As a result, no information has been published examining the effects of their habitat relocation on their physiology over time. Therefore, the primary objective of our study was to determine if habitat, while controlling for season, had an adverse effect on the YFP health as indicated by differences in serum biochemical parameters within YFP between each location (PL vs. TZO). Secondarily, to determine if indirect evidence exists that habitats have degraded by evaluating changes in YFP serum biochemical profiles within each location over time. And finally, if changes existed between locations, to detect during what age (immature or mature), sex or physiologic state (pregnancy or lactation) they were occurring. The results of this study will help determine the efficacy of establishing ex situ natural reserves and provide evidence for or against the continued expansion of this conservation strategy.

Animal ethics
The protocols for animal collection and handling were approved under Chinese law and guidelines for wildlife use by the Ministry of Agriculture of China. The Research Ethics Committee of Institute of Hydrobiology, the Chinese Academy of Science reviewed and approved the blood sampling and handling procedures.

Study design
A totals of 168 animals were sampled from TZO (n = 70) and PL (n = 98). Information about animals and sampling dates are summarized in Tables 1 and 2. Data collected in 2002in the TZO and 2009 in the PL have been previously reported (Nabi et al., 2017a. Animals were grouped based on total body length (Gao and Zhou, 1993) and sex as follows: Juvenile male (JM < 138 cm), adult male (AM > 138 cm), juvenile female (JF < 138 cm) and adult females (AF > 138 cm, Table 1). Adult females were further classified as non-pregnant, lactating females (L) and pregnant non-lactating females (P

Animal collection and blood sampling
Both in the PL and TZO, YFPs were captured by using the 'sound chase and net capture' method (Hua, 1987). The detailed information of animal chasing, catching, handling and release are explained in detail by Hao et al. (2009). Briefly, for each capture, animals were randomly selected within different geographical areas of the sampling locations for each collection attempt. The methodology during the capture event, the blood sampling procedure and the timing of blood collection were consistent for both populations. Ten ml of blood were drawn from the major vein of tail fluke aseptically using a disposable 10-ml syringe (Gemtier, G/Ø/ L: 21/0.7/31 mm, 201502, Shanghai, China). The blood was then transferred into serum separator and EDTA Vacutainer ® tubes (BD Vacutainer, Becton Dickinson, Franklin Lakes, New Jersey, USA), and placed immediately on ice. After centrifugation (Eppendorf AG, 22332, Hamburg, Germany) at 1500 × g for 15 min, the obtained serum was then immediately transferred into cryotubes (Fisher Scientific, Pittsburgh, Pennsylvania, USA), and stored in a liquid nitrogen kettle for transportation to the laboratory for immediate analysis. Before each assay, the auto-analyzer was calibrated.

Statistical analysis
Based on sampling methodology, and for the analysis, animals were considered to have been randomly selected during each collection period and no animal was sampled more than once. Statistical analysis was performed by either STATA ( A maximum likelihood (ML) general linear model (GLM, identity link, Gaussian family) was used to determine if the dependent variable (analyte concentration) differed between locations (STATA) and between groups combined across both locations. Many of the analytes evaluated are known to be effective by season (Hall et al., 2007;Macchi et al., 2011;Nollens et al., 2018); therefore, season was added as a covariate Post hoc marginal mean comparison between locations, groups and season were then preformed and data was presented as marginal mean and 95% confidence interval. To determine if changes had occurred for each dependent variable within each location over time (2002-15 for TZO and2009-17 for PL) a linear regression was used with time as a continuous independent variable and group as a covariate (STATA).
An unpaired Student's t-test was used to compare the analyte concentration within one group from PL to its respective group in TZO using Graph Pad Prism, version 5.01 (Graph Pad Software Inc., San Diego, CA, USA). Results of the t-test between locations were presented as mean ± SEM. For all analyses, significance was defined as P ≤ 0.05.

Results
Overall, the results of the GLM analysis indicated that 58% (19 of 33) of the analytes were significantly different between locations, while, 55% (18 of 33) demonstrated significant seasonal effects (Table 3). However, a complete picture of the seasonal effects could not be determined due to the lack of sampling during the summer months. Across both locations, 46% (15/33) of the analytes demonstrated significant differences between the different age and sex groups of animals ( Fig. 1). Finally, the linear regression results indicated that within each location, 46% (15 of 33) of the analytes changed over time in PL while only 24% (8 of 33) changed over time in TZO, respectively (Table 3).
Comparison of biochemical parameters from YFP between each habitat (TZO vs. PL)  (Table 3). For TZO animals, TC was increased (P = 0.004) while TG was almost half (P = 0.002) than those observed in PL, but HDL-c was increased (P < 0.0001, Table 3). The AMS was three times higher (P = 0.0005) in PL animals, while CK in TZO YFP was almost double (P = 0.0001) than PL animals. Finally, CO 2 was elevated (P = 0.002) while K + , Na + Ca 2+ PO 4 and Mg 2+ were significantly decreased in TZO animals (Table 3).
Biochemical parameter changes across the years for animals living in PL and TZO YFPs, respectively Poyang Lake Significant decreases in the liver related analytes ALT, AST and GGT were detected. No other significant changes were noted for liver related biochemical values (Table 3). The lipid profile indicated a significant decrease in TC, HDL-c and HDL-c/LDL-c, while a significant increase in LDL-c was detected (Table 3). Both the UA and CK concentrations decreased significantly, while LDH and AMS both significantly increased. For electrolytes, Ca + decreased while Mg 2+ and PO 4 increased (Table 3).

Tian-E-Zhou Oxbow
Significant changes in the hepatic system were characterized by an increase in T-BILI, D-BILI and I-BILI. Similarly, decreases (P = 0.04) in ALB were noted, however, no changes in liver enzymes over time were detected. A decrease (P < 0.001) in LDL-c concentration was also noted. Finally, a significant decrease in the Na + and Clwere noted, with a near significant decrease in K + (P = 0.08; Table 3).

Comparisons of marginal mean biochemical parameters between groups (JM, JF, AM, PF, LF) combined across both locations
The YFPs showed significant decreases in ALP, CK and increased in TP with age (P < 0.05, Fig. 1). The Ca 2+ was non-significantly higher in juveniles (JM, JF) versus AM and significantly higher than PF (Fig. 1). Inter sex differences within adults could not be measured directly since we did not sample any non-pregnant, non-lactating adult females and both pregnancy and lactation are known to affect multiple analytes in killer whales (Robeck and Nollens, 2012). However, within the juveniles, only GLU was significantly increased in JM compared to JF. The GGT was significantly greater in adult males compared to JF, PF and LF (Fig. 1). Pregnant YFP had significantly reduced AST, AST/ALT, CO 2 , Ca 2+ compared to JF, and significantly increased GLB, BUN and TG compared to JM, AM and both juveniles and AM, respectively. The LF had significantly increased HDL-c and K + compared to PF and AM, respectively (Fig. 1).
Comparisons of biochemical parameters in YFP between each location (TZO vs. PL) within each group (JM, JF, AM, P, L)

Juvenile males
The overall results of biochemical parameters are summarized in (Fig. 2). Juvenile male (JM) dwelling in the PL showed significantly higher serum level of GLB, Mg 2+ , Na + ,

Adult males
For AM living in the TZO, the hepatic enzymes (ALP, ALT, AST), lipid profile (TC, HDL-c, LDL-c) and other biochemical parameters (ALB, ALB/GLB, CK, Cl − , CO 2 ) were significantly higher than the adult males (AM) of PL YFPs. On the other hand, AM living in the PL showed significantly higher serum levels of hepatic enzymes (I-BILI, AST/ALT), lipid profile (TG, HDL-c/LDL-c) and other biochemical parameters (Ca 2+ , Cr, GLB, PO4 3− , TP) as shown in (Fig. 2).

Pregnant females
The same as observed in other groups, serum levels of TP, PO4 3 − , Ca 2+ , Na + and TG along with TIBILI were significantly higher in the PL YFPs. Only serum ALP and Cr were significantly higher in the TZO individuals (Fig. 2).

Discussion
Biochemical differences between locations and across time Liver parameters The hepatobiliary system of the TZO YFPs appeared to be under increased activation as demonstrated by a significant increase in bilirubin (T-BILI, D-BILI and I-BILI) over time and significant increases in liver associated analytes including ALP, AST, ALP, T-BILI, TBA and decreased TP compared to animals located at PL. In addition, during the entire study period, animals at PL had significant decreases in ALT, AST and GGT. Bile acids and bilirubin are typically used as indicators of liver clearance and can be elevated during states of parenchymal liver disease and biliary obstruction (Limdi and Hyde, 2003). Similarly, AST and ALT are typically increased during hepatocellular injury due to multiple clinical etiologies such as viral hepatitis, toxic hepatitis, cholestatic hepatitis and chronic active hepatitis (Rosalki and Mcintyre, 1999;Limdi and Hyde, 2003). Alternatively, significantly higher serum ALT, AST and ALP paired with the significantly increased serum cholesterol concentrations in the TZO animals suggests steatosis (Bayard et al., 2006). Total protein was reduced, albumin, which is produced by the liver and decreased in liver dysfunction, was increased for animals at TZO verses PL. The exact cause for elevated albumin in TZO needs further investigations. Complicating this determination is the fact that cetaceans are known to have heavy reserve capacity for hepatic albumin production. In addition, dehydration can elevate albumin production. Therefore, these factors limit the use of albumin as an early indicator of hepatic disorders (Bossart et al., 2001). The GLB were half and ALB/GLB was double those at TZO verse PL and were the cause of the observed decrease in TP.
Water quality in the TZO has recently been reported as being of reduced quality when compared to PL (Nabi et al., 2017a). We are aware (our unpublished work) that the TZO is primarily influenced by agricultural non-point pollution, natural input, poultry excrement and decayed organic matter pollutants. In addition to pesticides, poultry discharge from local poultry farms have been entering the TZO for over 20 years. Poultry discharge is known to possible contain viruses, bacteria, parasites, veterinary pharmaceuticals and heavy metals (EPA, 1998;University of Iowa and Iowa State Study Group, 2002;Boxall et al., 2003;Wei et al., 2010). All of these toxic chemicals or biologics are metabolized by or can directly affect the liver and may account for the increase in hepatic associated enzymes. Consequently, the significant changes observed in the hepatobiliary system of TZO over time are of important concern.

Lipid profile
In Poyang Lake YFPs, we observed a significant increase in LDL-C, and a significant decrease in TC, HDL-C and HDL-C/ LDL-C as compared to TZO, while a significant decrease in LDL-C in the TZO YFPs over time. (Table 3). Variation in the YFPs lipid profile may be affected by changing fisheries resources, overall animal nutritional health, habitat utilization and the reproductive cycle (Pethybridge et al., 2011a, b;Kim et al., 2012;Nabi et al., 2017b). In cetaceans, such as bottlenose dolphins (Tursiops truncates), beluga whales (Delphinapterus leucas) and pantropical spotted dolphins (Stenella attenuata), the effects of diet on the lipid profile has been reported (Asper et al., 1990;Cook et al., 1990;St. Aubin et al., 2013). A significant variation in the lipid profile in response to habitat dynamics may indicated changing prey availability between the two locations (Miller et al., 2012). In the PL, there are ample of evidences that overfishing (Chen et al., 2002;Wei et al., 2007;Li, 2008), illegal fishing, using illegal fish gears (Wang, 2009;Schelle, 2010), and removal of large numbers of fish and shrimps by sand mining machine (Yu et al., 2001), are reducing the availability or diversity of prey for the YFPs. Furthermore, water pollution, acoustic pollution and habitat degradation (Warner, 2008;Chen et al., 2009) have already threatened several fish species within this environment (Chen et al., 2004;Hvistendahl, 2008). Similarly, in the TZO, the increasing population of YFPs (Wang, 2015) combined with possible fish mortality and morbidity by various toxic chemical pollutants in the reserve (Nabi et al., 2017a) deplete or change the availability of prey for the YFPs.

Other enzymes
The significantly higher levels of serum CK and nonsignificantly higher concentration of LDH in TZO YFPs could be due to rhabdomyolysis linked to capture stress as these animals are occasionally exposed to chasing during capture (Williams and Pulley, 1983;Gasper and Gilchrist, 2005). Despite using the same capture method for both populations, animals in the TZO were apparently more active when compared to the PL (Nabi et al., 2017a). Therefore, the overall significantly higher CK level in the adult male of TZO might be due to the hyper-muscular activities (Paola et al., 2007).

Electrolytes
The significant decrease in serum levels of Na + , K + and Cl − in TZO as compared to PL, and a significant increase in the serum levels of Mg 2+ and PO4 3− while decrease in the Ca + levels of PL YFPs over time may reflect endocrine and gastrointestinal conditions (Bossart et al., 2001) of YFPs in response to a changing habitat (Sun et al., 2012;Dong, 2013). Overall, the electrolytes (Ca 2+ , PO4 3− , Mg 2+ , Na + ) were significantly higher in the PL YFPs compared to TZO YFPs where only Clwas

9
significantly higher. In addition to dehydration, hyperaldosteronism, liver and renal dysfunctions elevate Na + levels in cetacean (Bossart et al., 2001). Similarly, dehydration, renal disease, hypoadrenocorticism and primary hyperparathyroidism increase the serum Ca 2+ concentration (Bossart et al., 2001). The Mg 2+ , PO4 3− and Cl − levels are affected by renal problems. However, PO4 3− is also affected by rhabdomyolysis, dietary phosphorus excess, osteolytic bone disease, and hypoparathyroidism and hypercalcemia with normal glomerular filtration (Bossart et al., 2001). While differentiating clinical conditions from normal homeostatic variations in response to dietary, or environmental differences would require further diagnostics, these significant changes are worth noting.

10
Biochemical profile changes during maturation and reproductive state Serum ALP was significantly higher in the juveniles compared to the adults. While Ca 2+ was also increased in juveniles compared to adults with significant differences found when compared to PF. Both ALP, and Ca 2+ have been reported to be increased in juveniles of multiple cetaceans including killer whales, bottlenose dolphins and beluga (Andersen, 1968;St Aubin et al., 2001;Venn-Watson et al., 2011;Nollens et al., 2018). Higher ALP and Ca 2+ concentrations in young animals is generally associated with active bone growth (Andersen, 1968;Kovacs, 2001) and has been used to indicate physical maturity in other mammals including finless porpoises (Andersen, 1968). Increased CK in young animals was also observed in killer whales, bottlenose dolphins and pigs (Thorén-Tolling, 1982;Venn-Watson et al., 2007;Nollens et al., 2018). This increase during growth may be an indicator of muscle development in both species, but without isoenzyme identification a similar phenomenon occurring in YFP is only unknown.
Similar to what was observed in killer whales (Robeck and Nollens, 2013), Pregnant YFP had significantly decreased AST, AST/ALT ratio and increased GLB, TG. Liver enzyme decreases were attributed to volume expansion during pregnancy in killer whales and mare (Harvey et al., 2005). However, contrary to our results in YFP whereby BUN increased, this volume expansions also results in increased renal clearance and decreased BUN (Robeck and Nollens, 2013). Therefore, volume expansion may not be as significant in the relatively short gestation of the YFP (~<12 month) as compared to the killer whale (17.5 month, Robeck et al., 2015). Since BUN is considered metabolic waste of protein metabolism, the increase during YFP gestation may simply reflect increase food intake during gestation. The increase in TG were also observed in the killer whale (Robeck and Nollens, 2013) and in humans and horses this change has been attributed to estrogen mediated increase in Very Low Density Lipoprotein (VLDL) (Alvarez et al., 1996;Harvey et al., 2005).
Only HDL-C was elevated during lactation, while this change hints at the well-documented changes in lipid mobilization during lactation in most mammals (Koopman et al., 2002;Struntz et al., 2004), changes in TG, and TC post-partum are commonly observed in other species (Struntz et al., 2004;Robeck and Nollens, 2013). This lack of observed changes in TG and TC during lactation may indicate differences in YFP physiology or is most likely due to extremely small sample set of animals from which detecting significant deviations from the other groups was not possible.

Seasonality in biochemical parameters
While complete evaluation of seasonal changes could not be conducted due to the lack of sampling in the summer months, multiple parameters exhibited significant differences between the remaining three seasons. Serum creatinine in the YFPs showed seasonality with significantly higher levels in spring vs fall suggesting the effects of nutrition as observed in captive and wild bottlenose dolphins (Terasawa et al., 2002;Hall et al., 2007). We are aware (unpublished work) that in both populations of YFPs, prey availability is higher in the spring and summer and therefore these changes could reflect higher intake of prey. Furthermore, increased creatinine concentration during spring and summer months has been attributed to seasonal alteration in muscle mass in bottlenose dolphins (Hall et al., 2007;Macchi et al., 2011). In addition, higher concentrations of creatinine have been associated with increased and prolonged physical exertion (Gasper and Gilchrist, 2005) and this increase in spring for the YFP may be due to the increased socially driven physical activity associated with seasonality of reproduction in both males and females. For YFP, a seasonal spring increase in GLU was also observed and may also indicate metabolic changes in response to increased physical activity. Seasonal changes in activity can also be associated with changes in both demand and types of prey availability. For example, in the polar bear, concentrations of GLU correlate with seasonal changes in the percentage of dietary protein and fat (Bossart et al., 2001). The synthesis of ALB is directly associated with protein intake (Hall et al., 2007), therefore, higher serum ALB and ALB/GLB in spring provides additional support for increased prey consumption during the spring. The increased serum UA concentrations observed in winter for the YFP has also been reported in bottlenose dolphin and could be due to seasonal variation in the nutrient composition of their diet (Hall et al., 2007) as a diet rich in purines can cause hyperuricemia (Villegas et al., 2012). Similarly, seasonal variations in the lipid profile, electrolytes and hepatobiliary parameters of YFPs could be due to changes in the water components, changes in body condition, water temperature or quality, diet, photoperiod and other factors (Domingo-Roura et al., 2001;Terasawa et al., 2002;Sergent et al., 2004;Norman et al., 2013).

Conclusions and future recommendations
In summary, our findings provide indirect evidence for potential changes in fisheries resources in both populations as indicated by a significantly different and changing lipid profiles within and between the two populations. If the cause is from a declining fisheries resources in the TZO, it may be due to pesticides induce fish mortality, morbidity and low productivity. To regulate the Oxbow, fisheries resources and the total population numbers should be closely managed. Similarly, in the PL, illegal fishing and the use of nonselective fishing gears should be strictly avoided. While significant age specific differences were noted, all could be attributed to normal physiologic changes with age that have also been observed in other cetacean species. Across age groups, however, both the populations in general and TZO YFPs specifically showed hepatic dysfunction as indicated by higher hepatobiliary parameters which might be in response to various pollutants such as pesticides and poultry. The TZO also showed differing concentrations of various electrolytes as compared to PL which suggest gastrointestinal and endocrine changes whose etiology requires further investigation. Our findings suggest that YFPs in the TZO have not been removed from the negative effects of escalating anthropogenic activities especially, pesticides pollutions. Therefore, to conserve the YFPs in the TZO, special and immediate attention is required to improve the water quality and control the discharge from agriculture runoffs. In addition, future efforts at quantifying current concentrations of PCBs and other hydrocarbon bioaccumulations in fat and blood in both populations should be prioritized.