Modeling loss and recovery of Zostera marina beds in the Chesapeake Bay: The role of seedlings and seed-bank viability

Aquatic Botany 113 (2014) 32–45

Contents lists available at ScienceDirect

Aquatic Botany journal homepage: www.elsevier.com/locate/aquabot

Modeling loss and recovery of Zostera marina beds in the Chesapeake Bay: The role of seedlings and seed-bank viability Jessie C. Jarvis a,b,∗ , Mark J. Brush a,1 , Kenneth A. Moore a,2 a b

Virginia Institute of Marine Science, College of William and Mary, PO Box 1346, Gloucester Point, VA 23062, USA Centre for Tropical Water and Aquatic Ecosystem Research, James Cook University, P.O. Box 6811, Cairns, QLD 4870, Australia

a r t i c l e

i n f o

Article history: Received 26 August 2011 Received in revised form 1 October 2013 Accepted 11 October 2013 Available online 8 November 2013 Keywords: Zostera marina Sexual reproduction Ecological model Seed

a b s t r a c t Loss and recovery processes following a documented large scale decline in Zostera marina beds in the York River, Virginia in 2005 were modeled by coupling production and sexual reproduction models. The reproduction model included formulations for reproductive shoot production, seed production, seedbank density, seed viability, and seed germination. After the model was calibrated and validated using in situ water quality and plant performance measurements from two different sites, model scenarios were run for three years (1 year pre-decline, 2 years post-decline) to quantify the effects of (1) the presence or absence of sexual reproduction, (2) increases in water temperatures from ambient to ambient +5 ◦ C in 1 ◦ C increments, and (3) the potential interactive effects of light and temperature conditions on bed maintenance and re-establishment. Model projections of Z. marina production following the decline corresponded to in situ measurements of recovery only when sexual reproduction was added. However, a 1 ◦ C increase in temperature resulted in a complete loss of biomass after two consecutive years of temperature stress following the depletion of the viable sediment seed bank. Interactions between light and temperature stress resulted in overall lower production and resilience to declines under lower light conditions due to corresponding decreases in photosynthetic rates and increases in respiration. Model results highlight (1) the need to incorporate sexual reproduction into Z. marina ecosystem models, (2) the projected sensitivity of established beds to consecutive years of stress, and (3) the negative effects of multiple stressors on Z. marina resilience and recovery. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Global declines in seagrass populations have been well documented (Orth et al., 2006; Waycott et al., 2009; Short et al., 2011). The increase in rate of loss of seagrasses has been attributed to coastal development (Short and Wyllie-Echeverria, 1996), eutrophication (Burkholder et al., 2007), and climate change (Short and Neckles, 1999). Within the mid-Atlantic region of the United States in the Chesapeake Bay large scale declines in Zostera marina populations have been attributed to chronic deterioration in water quality compounded by episodic stresses from short term events such as tropical storms or high water temperatures (Orth and Moore, 1983; Moore and Jarvis, 2008). Restoration attempts in the Chesapeake have increased in response to continued declines; however, efforts

∗ Corresponding author at: Centre for Tropical Water and Aquatic Ecosystem Research, James Cook University, P.O. Box 6811, Cairns, QLD 4870, Australia. Tel.: +61 7 4232 2028; fax: +61 7 4781 5589. E-mail addresses: [email protected] (J.C. Jarvis), [email protected] (M.J. Brush), [email protected] (K.A. Moore). 1 Tel.: +1 804 684 7402; fax: +1 804 684 7752. 2 Tel.: +1 804 684 7384; fax: +1 804 684 7752. 0304-3770/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aquabot.2013.10.010

have proven to be primarily unsuccessful (Shafter and Bergstrom, 2008; Orth et al., 2010). In order to increase restoration efficiency, effectiveness, and success a better understanding of bed resilience to perturbations, as well as loss and recovery processes within established seagrass beds is required (Duarte, 2002; Orth et al., 2006). Z. marina populations within the Chesapeake Bay are particularly at risk for temperature stress as they are located near the southern limit of the species distribution in the western Atlantic (Short and Moore, 2006). Temperature stress has been attributed to the development of annual Z. marina populations along the Pacific (Meling-López and Ibarra-Obando, 1999; Santamaría-Gallegos et al., 2000), where summer water temperatures can reach a maximum of 30–32 ◦ C and result in a complete loss of above ground biomass. Average maximum summer water temperatures range from 28 to 30 ◦ C in the Chesapeake Bay (Moore and Jarvis, 2008), which is warmer than temperatures recorded for populations in the Mediterranean (26.9 ◦ C, Plus et al., 2010) and for Z. marina living near its southern limit in the eastern Atlantic in the Ria Formosa, Portugal (27 ◦ C; Newton and Mudge, 2003; den Hartog, 1970). Evidence for the vulnerability of seagrass populations in this region to episodic stressors has been highlighted in the last

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

decade following sudden large scale declines in 2005 (Moore and Jarvis, 2008) and in 2010 (Moore et al., 2013). In 2005, the Z. marina decline was related to a period of unusually high water temperatures (maximum 32 ◦ C) in July and August 2005. The two month period of temperature stress in the summer of 2005 combined with chronically light stressed Z. marina populations (Moore, 2004) resulted in the loss of the majority of above ground biomass by October (Moore and Jarvis, 2008). The 2010 decline was also attributed to temperature and light impacts; however, the timing of the event was significantly different as the heat event was documented in June at the end of the maximum biomass period prior to stressful peak summer water temperatures (Moore et al., 2013). Z. marina populations may continue to experience these types of large-scale event-driven declines, particularly in the Chesapeake Bay where between 1949 and 2002 Chesapeake Bay winter water temperatures increased 0.8–1.1 ◦ C and are predicted to continue to increase by 2–5 ◦ C over the next century (Preston, 2004; Najjar et al., 2010). In addition, climate change models predict that by the end of the 21st century a 10% increase in rainfall over the Chesapeake Bay watershed will subsequently increase river flow by 30%, thereby increasing the nutrient and sediment input into the bay (Najjar, 1999; Gibson and Najjar, 2000). As a result the amount of light available to Z. marina will likely decrease, further stressing Z. marina populations (Dennison et al., 1993). Quantifying possible interactions between environmental stressors on seagrass beds in the Chesapeake Bay is necessary to accurately predict persistence and resilience of these populations. Ecological models are useful tools in quantitative analysis of complex ecosystems such as seagrass beds. Through models, the response of Z. marina to stressful environmental conditions such as low light, high nutrients, and high temperatures has been quantified under a variety of situations (Wetzel and Neckles, 1986; Bach, 1993; Aveytua-Alcázar et al., 2008). While these models provide insight into the singular and combined effects of environmental stressors on Z. marina production, the capacity to accurately model population responses to stressful conditions is limited by focusing solely on vegetative reproduction and ignoring sexual reproduction (van Lent, 1995). Exclusion of sexual reproduction in carbon based models has been accepted due the dominance of vegetative reproduction in perennial Z. marina beds and the relatively low carbon value of seeds (Harwell, 2000). However, recent research has shown that sexual reproduction plays a significant role in Z. marina bed recovery from large scale declines (Plus et al., 2003; Greve et al., 2005); therefore a key component of the bed loss and recovery dynamic may be missing from Z. marina production models when sexual reproduction is excluded. For perennial Z. marina populations, seeds in the sediment seedbank provide a measure of resilience to large scale loss (Leck et al., 1989; Combroux et al., 2001). Recolonization of Z. marina beds following sudden large scale declines through seed germination and seedling establishment have been documented throughout the species range (Plus et al., 2003; Greve et al., 2005; Lee et al., 2007; Jarvis and Moore, 2010). However, for the seed-bank to provide any function, seed viability must be maintained (Leck et al., 1989). Successful germination of viable seeds is dependent upon environmental cues and the surrounding sediment microenvironment (Hootsmans et al., 1987; Moore et al., 1993; Probert and Brenchly, 1999). Therefore, ecological models need to consider seed production, seed-bank density, seed viability, and germination to accurately incorporate sexual reproduction into seagrass production models. The objective of this study was to evaluate the role of sexual and asexual reproduction in providing resilience to and recovery from disturbance in perennial Z. marina beds in the lower Chesapeake Bay. A sexual reproduction sub-model was developed and coupled to a Z. marina production model building upon established

33

Table 1 Governing equations for (1) epiphyte biomass (Cepi ; g C m−2 ); (2) Z. marina vegetative shoot biomass (Czms ; g C m−2 ); (3) Z. marina vegetative root/rhizome biomass (Czmr ; g C m−2 ); (4) Z. marina seed-bank density (Zmseeds ; seeds m−2 ); and (5) Z. marina seedling density (Zmsd ; seedlings m−2 ). Terms include P = production; M = mortality; G = grazing; R = respiration; Td = translocation down; Tczmss = transfer of seedling biomass to vegetative shoot biomass; Tczmsr = transfer of seedling root/rhizome biomass to vegetative root/rhizome biomass; Seedsgerm = germinated seeds; Seedsprod = total seeds produced; Seedsvia = viable seeds; PRseeds = seed predation; Zmsd = germinated seedling density. Differential equations C epi = C epi (t − dt) + (P epi − M epi − Gepi − Repi ) × dt C zms = C zms (t − dt) + (P zms + T czmss − M zms − Rzms − T d ) × dt C zmr = C zmr (t − dt) + (T d + T czmsr − M zmr − Rzmr ) × dt Zmseeds = Zmseeds (t − dt) + (Seedsprod − M seeds − PRseeds ) × Seedsvia × dt Zmsd = Zmsd (t − dt) + (Seedsgerm − M zmsd ) × dt

(1) (2) (3) (4) (5)

models of submerged aquatic vegetation in the Chesapeake Bay (Madden and Kemp, 1996; Buzzelli et al., 1999; Cerco and Moore, 2001). Model simulations were then used to quantify the effects of (1) the presence or absence of sexual reproduction, (2) projected increases in water temperature from ambient to ambient +5 ◦ C, and (3) the potential interactive effects of light and temperature on bed maintenance and re-establishment following a large scale decline. Specifically we quantified percent change between ambient (base model conditions) and model scenarios in Z. marina above and below ground biomass, total seed production, seed-bank density, and seed germination. 2. Methods 2.1. Model description An established perennial Z. marina bed in the lower York River, Virginia USA (Fig. 1) was chosen as a basis for the production and sexual reproduction sub-models due to the documented large-scale die-off of eelgrass around the lower Chesapeake Bay in the fall of 2005 in response to elevated water temperatures (Moore and Jarvis, 2008). The base Z. marina productivity model was modified from the models of Madden and Kemp (1996), Buzzelli et al. (1999), and Cerco and Moore (2001) (Fig. 2). The sexual reproduction submodel was developed based on monthly observations of Z. marina beds in the York River as well as field and laboratory experiments conducted either in this region or using seeds collected from populations from the local area (Jarvis, 2009; Jarvis and Moore, 2010; Jarvis et al., 2012). 2.2. Model formulation 2.2.1. Production model Governing equations for Z. marina vegetative and seedling shoot biomass were balanced between gains through photosynthesis and losses due to mortality, respiration and translocation to roots and rhizomes (Table 1). Epiphytes were balanced similarly with the added loss of grazing but no loss due to translocation. Production terms for both epiphytes and Z. marina shoots were computed as the product of a temperature dependent maximum rate (Pmax ) and a limiting factor in which either nutrients (dissolved inorganic nitrogen, DIN, or dissolved inorganic phosphorous, DIP) or light (PAR) was limiting (Madden and Kemp, 1996; Cerco and Moore, 2001). Maximum epiphyte production rates were taken from Buzzelli et al. (1999) and production rates for Z. marina were determined from Evans et al. (1986). In both cases maximum production is related to ambient (Tw ) and optimum water temperatures (Topt ; Table 2). Z. marina shoot and epiphyte production (Pzms and Pepi , respectively) were limited by available light and nutrient concentrations.

34

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

Fig. 1. Lower York River Goodwin Island site locations for all calibration and forcing function data collection. Goodwin Island (GI) is located at the mouth of the York River and Gloucester Point (GP) is located 10 km upriver. Site locations are denoted with a star.

PAR availability was calculated similarly to Madden and Kemp (1996) where forced incident irradiance was reduced in successive stages. Initial light availability was reduced exponentially to depth (z) with a Beer Lambert equation (Kirk, 1983). Down-welling light attenuation coefficient (Kd ) accounted for additive effects of

chlorophyll a, TSS, and the water itself on light availability in the water column (Madden and Kemp, 1996). Total light available to Z. marina leaves was further attenuated as a function of simulated epiphyte biomass on the leaf blades (Madden and Kemp, 1996). For both epiphytes and Z. marina light limitation of Pmax values was

Table 2 Parameter estimates for the Z. marina production model. References: 1 = calibration within the model, 2 = Buzzelli et al. (1999), 3 = Cerco and Moore (2001), 4 = Madden and Kemp (1996); 5 = Bach (1993); 6 = Fishman and Orth (1996). Abbrev. BMRepi JD Kgepi Khnepi Khnszm Khnwzm Khpepi Khpszm Khpwzm KPARepi KPARzm KtBepi MRepi MRzmr MRzms RRzmr Toptepi Toptzm Tzms WD zmr

Description Epiphyte basal metabolic rate Julian Day Epiphyte grazing constant Epiphyte N half saturation constant Z. marina N half saturation constant sediment Z. marina N half saturation constant water Epiphyte P half saturation constant Z. marina P half saturation constant sediment Z. marina P half saturation constant water Epiphyte PAR half saturation constant Z. Marina PAR half saturation constant Epiphyte respiration constant Epiphyte mortality constant Z. marina root mortality constant Jan–July Z. marina root mortality constant July–Dec Z. marina shoot mortality constant Z. marina root respiration at 20 ◦ C Epiphyte optimum temperature for production Z. marina optimum temperature for production Z. marina shoot to root transfer Water Depth Z. marina root respiration constant

Units −1

d d−1 d−1 ␮mol N m−3 ␮mol N m−3 ␮mol N m−3 ␮mol P m−3 ␮mol P m−3 ␮mol P m−3 ␮E m−2 s−1 ␮E m−2 s−1 ◦ C d−1 d−1 d−1 d−1 d−1 ◦ C ◦ C Unitless m Unitless

Value

Reference

0.047 0–365 0.0094 1.79E−09 2.86E−09 7.14E−10 7.14E−11 7.14E−09 4.35E−10 90 57.5 0.069 0.0085 0.000085 0.0095 0.007 0.0005 25 22.5 0.3 0.5 1.25

2 1 3 3 3 3 3, 4 3 4 3 2 1 1 1 1 1 2 3 3 5

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

35

Fig. 2. Conceptual diagram for Zostera marina production and sexual reproduction model. Circles = forcing functions, triangles = modifiers, squares = state variables, thick arrows = flows, and thin arrows = iterations. State variables included epiphyte biomass (Cepi ) and Z. marina vegetative shoot biomass (Czms ), vegetative root biomass (Czmr ), seed-bank density (Zmseeds ); seedling density (Zmsd ). Forcing functions included water temperature (◦ C), photoperiod (F), photosynthetically active radiation (PAR, ␮E m−2 s−1 ), water column chlorophyll a (␮g l−1 ), total suspended solids (mg l−1 ), water column and sediment dissolved inorganic nitrogen (DINWC , DINS –NOx + NH4 , ␮mol l−1 ), water column and sediment dissolved inorganic phosphorus (DIPWC , DICS –PO4 −3 , ␮mol l−1 ), sediment organic content (SO, % organic), and seed burial depth (BD, cm). Temp, JD, and F affect multiple processes so are not connected to minimize diagram complexity.

calculated using Michaelis–Menten kinetics with a half saturation constant for light (KPAR Table 2; Madden and Kemp, 1996). Nutrient limitation in epiphytes was computed similarly to light limitation using forced water column nutrient concentrations (Nw ) and half saturation constants for nutrients (Table 2). For Z. marina multiple sources of nutrients (sediment (Ns ) and water column) were taken into account with a Monod-like function for nutrient limitation (Madden and Kemp, 1996; Cerco and Moore, 2001). Losses from epiphytes and Z. marina shoots were a function of mortality (leaf sloughing in Z. marina) and respiration (Madden and Kemp, 1996; Buzzelli et al., 1999). Z. marina shoots had an additional loss term due to translocation (Td ; Buzzelli et al., 1999) to the roots at a constant rate (Tzms ; Table 2), while epiphyte production was also lost to grazing (Gepi ) using a quadratic function of biomass and rate constant Kgepi (Table 2; Buzzelli et al., 1999). Epiphyte mortality (Mepi ) was a function of a density dependent mortality constant (MRepi ) and the ratio of epiphyte and Z. marina shoot carbon (Table 2; Buzzelli et al., 1999). Z. marina shoot mortality (Mzms ) was a combination of a constant mortality term over time (MRzms ; Table 2) and a temperature dependent function (Buzzelli et al., 1999). Epiphyte respiration (Repi ) was a combination of a constant basal respiration rate (BMRepi ) and a temperature dependent function (with constant KtBepi ; Table 2; Buzzelli et al., 1999). Z. marina shoot respiration (Rzms ) was a temperature-dependent function of Z. marina production per day (PRzm ; Buzzelli et al., 1999), and was reduced to 0 at water temperatures below 14 ◦ C (Nejrup and Pedersen, 2008). Z. marina root and rhizome respiration (Rzmr ) was based on an Arrhenius relationship between respiration and water temperature (Bach, 1993). Respiration at an optimum temperature of 22 ◦ C (RRzmr ; Table 2) was scaled to daily temperatures with an Arrhenius constant (zmr ; Table 2; Buzzelli et al., 1999). Root and rhizome mortality (Mzmr ) was computed as a constant fraction of

biomass (MRzmr ; Table 2) which increased after water temperatures became stressful to Z. marina (>25 ◦ C) in June of each model run (Setchell, 1929; Marsh et al., 1986; Nejrup and Pedersen, 2008; Hosokawa et al., 2009; Höffle et al., 2010). 2.2.2. Reproduction model A factor for converting shoot carbon to density (VegC:D ; Table 3) was computed based on Z. marina above ground biomass samples collected from Goodwin Island in 2006 and 2007 (n = 560) and used to initiate the sexual reproduction model. Flowering was considered to be limited by water temperature and time (Julian day) and was based on the optimum conditions for Z. marina flowering observed in the York River (Silberhorn et al., 1983). When water temperatures were <21 ◦ C and Julian Day was <182 (July 1), 10% of total shoot density was converted to flowering shoots (Zmrsf ; Table 2; Silberhorn et al., 1983). In addition, flowering was initiated only when vegetative shoots dominated the above ground carbon pool as Z. marina seedlings in the Chesapeake Bay do not flower during their first year of growth (Silberhorn et al., 1983). Subsequent loss of flowering shoots was considered to be inherently included in the above ground biomass mortality term. Initial seed-bank densities (Seedsprod ) were derived from the product of total flowering shoot densities and the average number of seeds per reproductive shoot (Seedssh , Table 3). Once produced, seeds were deposited into the sediment seed-bank. While in the seed-bank a portion of the seeds were removed via mortality and predation (MRseeds and PRseeds , respectively, Table 3; Fishman and Orth, 1996). The number of germinable seeds remaining in the seed-bank was further reduced by a loss of viability (VRseeds ; Table 3). The seeds remained in the seed-bank until water temperatures decreased below 20 ◦ C as this is when germination is initiated in Chesapeake Bay populations (Moore et al., 1993). Due to the transient nature of Chesapeake Bay seed-banks those seeds

36

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

Table 3 Parameter estimates for the Z. marina reproduction model. References: 1 = calibration within the model, 2 = Silberhorn et al. (1983); 3 = Harwell (2000); 4 = Fishman and Orth (1996); 5 = Bintz and Nixon (2001). Abbrev

Description

Units

Value

Reference

MRseeds Mzmsd PRseeds SeedlingRD:C SeedlingSD:C Seedssh VegC:D VRseeds Zmrsf

Seeds mortality rate Z. marina shoot mortality rate Seeds predation rate Z. marina seedling density to roots conversion factor Z. marina seedling density to shoots conversion factor Seeds per reproductive shoot Z. marina shoot carbon to density Seeds viability rate Reproductive shoot density

d−1 Unitless d−1 g C shoot−1 g C shoot−1 seeds shoot−1 g C shoot−1 d−1 Unitless

0.1 0–1 0.33 0.0384 0.0374 10 0.0168 0.4 0.10

1 5 4 n = 120 shoots n = 120 shoots 3 n = 560 shoots n = 100 seeds 2

that did not germinate by day 365 (December 31) where then lost from the system via seed mortality (Orth et al., 2000; Jarvis and Moore, 2010). Germination of viable seeds (Seedsvia ; Table 1) was determined by a relationship between sediment organic content (SO, %) and seed burial depth (BD, cm) which was held constant at 3 cm and a viability rate constant (VRseeds ; Table 3): Seedsgerm =



1 1 + e(−0.1432+(1.1261×BD)+(−1.3964×SO))



× VRseeds

(Jarvis, 2009). Once germinated, seedlings were then converted back to above and below ground carbon values using fixed conversion factors (SeedlingSD:C , SeedlingRD:C , Table 3). When above ground vegetative Z. marina biomass was <0.44 g C m−2 all above and below ground seedling biomass was transferred to the vegetative shoot and root stocks. If vegetative shoot carbon was >0.44 g C m−2 then seedling mortality was 100% (Mzmsd , Table 3). This relationship was based on the inhibitory effect of shading by established vegetative above ground biomass on the survival of seedlings (Phillips et al., 1983; Robertson and Mann, 1984; Bintz and Nixon, 2001). Seedling above and below ground biomass was not tracked separately through the first year of growth due to a lack of information on all seedling parameters. 2.3. Calibration, verification, and testing The initial model (production and seed sub-model) was developed using the STELLA® (isee systems, inc.) modeling platform with a simulation period of 1.5 years (April 1, 2006 through December 31, 2007) and a time step (dt) of 0.125 days. Parameter values for epiphytes and Z. marina were selected from the literature and field data, and revised to increase model fit to calibration data within ecological limits (Table 2). The model was forced with water column (PAR ␮E m−2 s−1 , temperature ◦ C, turbidity mg l−1 , chlorophyll a ␮g l−1 , DIN NOx + NH4 ␮mol l−1 , DIP PO4 −3 , ␮mol l−1 ) and sediment data (percent organic content, DIN NH4 ␮mol l−1 , DIP PO4 −3 , ␮mol l−1 ) and calibrated to Z. marina biomass data collected at biweekly to monthly intervals from April to October in 2006 and 2007 from the Goodwin Island National Estuarine Research Reserve (37◦ 13 N; 76◦ 23 W; Figs. 3 and 4). Once calibrated, the model was independently verified using similar data from a site located 10 km up river adjacent to the Virginia Institute of Marine Science Gloucester Point campus (37◦ 14.8 N, 76◦ 30.3 W; Figs. 3 and 4). Parameter values were left unchanged for verification, but forcing functions were updated to reflect Gloucester Point data. Due to a lack of data on epiphytic biomass during this time period, simulated epiphyte biomass was compared solely to literature values (Cerco and Moore, 2001). Comparisons were made between computed and observed values of all state variables on a monthly average basis using regression analysis. The sensitivity of base model conditions to all parameter estimates and forcing functions was analyzed by sequentially varying values by ±5, 10, and 20%. The average percent

change in all state variables between the base model and sensitivity simulations was then calculated and tests that resulted in >10% change in state variable concentrations were considered to have the greatest impact on model results. 2.4. Model scenarios After calibration, the model was extended to run from April 1, 2005 to December 31, 2007 for both the downriver (Goodwin Island) and upriver (Gloucester Point) sites and run under three different scenarios based on recovery from the 2005 decline. First, the extended model run format was used to quantify the potential role of seeds and seedlings in the recovery and re-establishment of Z. marina beds at both sites, by running the model with and without sexual reproduction. In the second scenario, the role of sexual reproduction in providing resilience to established beds at both the Goodwin Island and Gloucester Point sites via the production of a viable sediment seed-bank to repeated stress was quantified by increasing water temperatures in 1 ◦ C increments from ambient (base model) conditions to ambient + 5 ◦ C. Finally, during the third scenario, the singular and combined effects of light and temperature stress on Z. marina reestablishment due to site differences in water column light attenuation were quantified by comparing Goodwin Island Z. marina bed resilience to temperature stress (ambient to ambient + 5 ◦ C) under ambient (Goodwin Island) and elevated (Gloucester Point) values of light attenuation (Kd ). For all scenarios we quantified percent change between ambient (base model conditions) and model scenarios in Z. marina above and below ground biomass, total seed production, seed-bank density, and seed germination. 3. Results 3.1. Model calibration The model captured the overall seasonal trends in above ground biomass and under typical conditions (i.e. no temperature or light stress) it produced repeatable annual biomass cycles with or without the inclusion of sexual reproduction (data not shown). Field estimates of biomass (reported as mean ± S.E.) in 2006 and 2007 at Goodwin Island ranged seasonally from 2.4 ± 0.8 g C m−2 to 38.6 ± 8.3 g C m−2 while the model output ranged from 0.3 g C m−2 to 38.5 g C m−2 (Fig. 5). The model was the most accurate in describing the initial re-colonization of Goodwin Island between April and July 2006. Following the maximum peak in above ground biomass the model diverged substantially from field measurements and the percent error from August to October averaged 87 ± 3%. The divergence during this time period was due to an underprediction of above ground biomass in the model. A similar pattern in model prediction was observed in 2007 (Fig. 5). Belowground biomass also varied seasonally, but less than aboveground biomass. Observed belowground biomass at Goodwin Island ranged from

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45 800

37

PAR

PAR (µE m-2 min-1)

700 600 500 400 300 200 100 0

Water Temperature (C)

35

Temperature

Date (month)

30 25 20 15 10 5 0 50

Turbidity

45

TSS (mg l-1)

40 35 30 25 20 15 10 5 0 180

Chlorophyll a (µg l-1)

160

Chlorophyll a

140 120 100 80 60 40 20

Sediment Organic Content (% organic)

0 3.0

Sediment % Organic 2.5 2.0 1.5 1.0 0.5 0.0 Jan 05

Jul 05

Jan 06

Jul 06

Jan 07

Jul 07

Jan 08

Date (month) Fig. 3. Forcing functions for Goodwin Island (solid line) and Gloucester Point (dashed line) for 2005–2007.

9.2 ± 2.1 g C m−2 to 75.7 ± 10.3 g C m−2 while modeled values were similar and ranged from 4.1 g C m−2 to 38.5 g C m−2 (Fig. 5). As with aboveground biomass, the model under-predicted belowground biomass values throughout the calibration period as percent error from August to October averaged 50 ± 17% in 2006 and 46 ± 11% in 2007. During model calibration both above and below ground biomass respiration, mortality and carbon translocation constants were altered systematically by increasing and decreasing all factors

individually and combined to reduce the biomass error. The smallest error corresponded to the literature values reported in Table 2. 2007 observed seed production averaged In 6922 ± 778 seeds m−2 while the model predicted germinable seed densities of 13,034 seeds m−2 . Maximum viable seed-bank densities predicted by the model were also greater than observed values. No viable seeds were found in the ambient sediment seedbank in the Goodwin Island site in 2006 or 2007. In the calibration

38

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

30

Water Column DIN DIN (µ mol N l-1)

25 20 15 10 5 0 2.0

DIP (µmol PO4-3 l-1)

Water Column DIP 1.5

1.0

0.5

0.0 30

Sediment DIN DIN (µmol NH4 l-1)

25 20 15 10 5 0 10

DIP (µmol PO4 l-1)

Sediment DIP 8 6 4 2 0 Apr 05

Aug 05

Dec 05

Apr 06

Aug 06

Dec 06

Apr 07

Aug 07

Dec 07

Date (month) Fig. 4. Water column and sediment nutrient forcing functions for Goodwin Island (solid line) and Gloucester Point (dashed line) for 2005–2007.

model runs 0 seeds m−2 were produced in 2006; however, in 2007 the model produced maximum seed viable seed-bank densities of 3136 seeds m−2 . 3.2. Model verification The model accurately predicted lower overall above and below ground biomass values in Gloucester Point Z. marina beds compared to Goodwin Island (Fig. 6). Similar to the base model runs, the verification runs were the most accurate in describing the initial re-colonization in 2006 and the spring growth period (April–July) in 2007 and under-predicted biomass during the fall (August–October) period in both

years. As with above ground biomass, the model under predicted below ground biomass throughout the verification period. No flowering shoots were observed at Gloucester Point in 2006 or in 2007. However, the model predicted germinable seed densities at 1030 seeds m−2 . Maximum viable seed-bank densities predicted by the model were also greater than observed values. No viable seeds were found in the ambient sediment seed-bank in Gloucester Point in 2006 or 2007. In the verification model runs 0 seeds m−2 were produced in 2006; however, in 2007 the model produced maximum viable seed-bank densities of 302 seeds m−2 .

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45 100

-2

Above Ground Carbon (g C m )

90

A Model Goodwin Island

80 70 60

Table 4 Minimum sensitivity simulation (±5, 10, 20%) for model parameters which resulted in significant variation (≥10%) of state variables relative to base model concentrations. Non-significant values are denoted with (–). State variable

Parameter

Min % change

Epiphytes

PRepi Pmax Kgepi MRepi BMRepi KtBepi

−5 ±5 +5 – ±5 ±20

Z. marina shoots

PRzm Pmax Tczms MRzms Rzms Td

±5 ±5 – – ±20 ±5

Z. marina Root/Rhizome

Td Tczmsr MRzmr RRzmr Rzmr

−10 – ±20 – –

Seed-bank

VegD:C Fsden MRseeds PRseeds VRseeds

−10 ±10 ±5 +5 ±5

Seed Germination

Vseeds Msd

50 40 30 20 10 0 100

-2

Below Ground Carbon (g C m )

90

B

80 70 60 50 40 30

39

20

±5 ±5

10

Ap r M 06 ay 0 Ju 6 n 0 Ju 6 l Au 06 g 0 Se 6 p O 06 ct N 06 ov D 06 ec 0 Ja 6 n 0 Fe 7 b M 07 ar Ap 07 r M 07 ay 0 Ju 7 n 0 Ju 7 l Au 07 g 0 Se 7 p 0 O 7 ct N 07 ov D 07 ec 0 Ja 7 n 08

0

Date (month) Fig. 5. Calibration of Zostera marina above and below ground biomass model (black line) with observed Goodwin Island data (triangles). Observed data are given in monthly means ± SE.

Table 5 Minimum sensitivity simulation (±5, 10, 20%) for model parameters which resulted in significant variation (≥10%) of forcing functions relative to base model concentrations. Non significant values are denoted with (–). Forcing function

Parameter

Temperature

Epi Zm Shoots Zm Roots Seed-bank Seed Germination

±5 ±5 ±5 ±5 −5

PAR1

Epi Zm Shoots Zm Roots Seed-bank Seed Germination

−5 +5 – −20 –

PAR2

Epi Zm Shoots Zm Roots Seed-bank Seed Germination

−5 +5 – – –

PAR3

Epi Zm Shoots Zm Roots Seed-bank Seed Germination

−5 −5 ±20 +5 +5

Chlorophyll a

Epi Zm Shoots Zm Roots Seed-bank Seed Germination

+5 – – –

TSS

Epi Zm Shoots Zm Roots Seed-bank Seed Germination

+5 – – – –

3.3. Sensitivity analysis 3.3.1. Parameter effects Epiphyte biomass was most sensitive to changes in production and grazing and least responsive to mortality and respiration (Table 4). Z. marina above ground biomass was also most sensitive to changes in production while both above and below ground biomass were sensitive to shoot to root translocation. Overall, biomass (epiphyte, above and below ground) was least sensitive to mortality and respiration rates (Table 4). Seed-bank densities were more sensitive to factors that influenced seed density (i.e. predation, mortality, and viability) rather than seed production (total shoot carbon to density ratio, reproductive shoot densities). Once in the seed-bank, seed germination was highly sensitive to the number of viable seeds and seedlings. Overall seed germination was more sensitive to increasing than decreasing seed viability while the effects of seed mortality were similar across analyses (Table 4). 3.3.2. Forcing functions All state variables were sensitive to changes in temperature (Table 5). Z. marina state variables were more sensitive to decreases compared to increases in water temperature. Seed-bank density was the most sensitive to increased water temperature as seedbank density increased 460% in response to a 5% decrease in water temperature. Only epiphyte biomass and Z. marina above ground biomass were sensitive to changes in total available light as it entered the water column (PAR1 ) and after light attenuation in the water column (PAR2 ). All biomass pools (epiphyte, above and

% Change

40

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

A Model Gloucester Point

50

-2

Above Ground Carbon (g C m )

60

40

30

20

10

0

B

50

-2

Below Ground Carbon (g C m )

60

40

30

20

10

Ap r M 06 ay 0 Ju 6 n 0 Ju 6 l Au 06 g 0 Se 6 p 0 O 6 ct N 06 ov D 06 ec 0 Ja 6 n 0 Fe 7 b M 07 ar Ap 07 r M 07 ay 0 Ju 7 n 0 Ju 7 l Au 07 g 0 Se 7 p 0 O 7 ct N 07 ov D 07 ec 0 Ja 7 n 08

0

Date (month)

Fig. 6. Verification of Z. marina above and below ground biomass model (black line) with observed Gloucester Point data (circles). Observed data are given in monthly means ± SE.

below ground), seed bank density, and germination were sensitive to changes in light after it was reduced by both water column light attenuation factors and by epiphytic growth on Z. marina blades (PAR3 ; Table 5). 3.4. Model scenarios 3.4.1. Reproduction Under ambient conditions, the model scenarios with the mixed (vegetative and sexual) mode of reproduction resulted in greater predicted above and below ground biomass than the scenarios with vegetative growth only (Fig. 7). Recovery after the fall 2005 die-off was only possible with sexual reproduction of seeds and germination of seedlings included in the model; without this mode of reproduction simulated above ground biomass was too low to initiate new growth in 2006. Below ground biomass was also lost when run in the absence of sexual reproduction, although at a slower rate (Fig. 7). By year three below ground biomass was completely absent from either site in the vegetative scenarios while the mixed mode scenarios retained <10 g C m2 (Fig. 7). 3.4.2. Water temperature The resilience provided by sexual reproduction was limited, however, with predicted loss of all biomass at both sites by year three once water temperature was increased 1 ◦ C (Fig. 8). For models running in vegetative mode an increase of 1 ◦ C resulted

in complete above ground biomass loss in year one (results not shown). Without seed input, above ground biomass did not recover in year two or three regardless of site. Above ground biomass at both sites recovered in year two when seed production was included. Adding the mixed mode of reproduction to the model allowed for limited recovery; however, multiple years of stressful water temperatures resulted in complete loss of Z. marina with or without the inclusion of sexual reproduction. For years one and two, loss of below ground biomass under elevated temperature was similar between mixed (Fig. 8) and vegetative (not shown) reproductive modes with percent change between the two modes ranging from 0 to 33%. By the end of year three below ground biomass was absent from all runs where temperature was increased by ≥1 ◦ C. As with above and below ground biomass, total seed production, seed-bank density, and maximum viable seed density all decreased with increasing water temperature. Seed production was similar between sites with rates decreasing 9% at Goodwin Island and 13% at Gloucester Point with a 1 ◦ C increase in water temperature in 2005. Seed production decreased by 69.1% and 70.6% at Goodwin Island and Gloucester Point respectively when water temperature was increased by 5 ◦ C. Due to a lack of flowering shoots, seeds were no longer produced in the model when water temperatures were increased to ambient + 1 ◦ C. Seed-bank and viable seed densities followed similar patterns to total seed production. 3.4.3. Interactive effects of temperature and light stress Overall model runs projected greater above ground biomass, below ground biomass, seed production, and seed germination at Goodwin Island than at Gloucester Point (Figs. 7 and 8). While differences in water temperature between sites was <10%, Gloucester Point waters were more turbid with daily TSS and chlorophyll a values that were up to 43% and 52% greater, respectively, than values at Goodwin Island depending upon the year (Fig. 3). When the Goodwin Island model was run with the higher water column light attenuation values from Gloucester Point, the model projected a substantial decrease in above ground biomass (32.74% in 2006 and 91.13% in 2007) under ambient water temperatures (Fig. 9). This decrease under ambient conditions highlights the interactive effects of temperature and light stress on Z. marina survival. However, there were no differences in predicted above ground biomass when using the higher attenuation coefficients from Gloucester Point at Goodwin Island when temperatures were increased above ambient conditions. Lowering the light levels observed at Goodwin Island to those reflective of the slightly higher turbidity conditions at the Gloucester Point did not appear to compound the effects of the temperature stress and Z. marina populations continued to decline at rates similar to those in simulations with temperature stress only (Fig. 9). 4. Discussion The model presented here reproduced the general observed trends in above and below ground Z. marina biomass in the York River following the 2005 heat induced decline only after sexual reproduction was included in the model formulation. These results reflect the observed recovery in the York River in 2006 where newly germinated seedlings constituted >80% of total shoot density at both Goodwin Island and Gloucester Point (Jarvis and Moore, 2010). As sexual reproduction has been shown to play a critical role in the recovery of perennial Z. marina beds following large scale declines (Plus et al., 2003; Greve et al., 2005; Jarvis and Moore, 2010) the inclusion of flowering shoot production, seeds production, seed bank viability, and seed germination should be included in future Z. marina production models to accurately describe the resilience and response of these beds to stressful environmental conditions.

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

Above Ground Biomass (g C m-2)

40

A

Goodwin Island

41

Gloucester Point

30

20

10

0

Below Ground Biomass (g C m-2)

40

B

30

20

10

Date

07 D ec 07

A ug

D ec 06 A pr 07

5 A pr 06 A ug 06

D ec 0

A pr 05 A ug 05

7 D ec 0

6 A pr 07 A ug 07

D ec 0

5 A pr 06 A ug 06

D ec 0

A pr 05 A ug 05

0

Date

Fig. 7. Goodwin Island and Gloucester Point (A) above ground biomass and (B) below ground biomass model projections with sexual reproduction (solid line) and without sexual reproduction (dashed line) for 2005–2007 under ambient temperatures.

4.1. Modeling sexual reproduction On average, perennial Z. marina seed production ranges between 50–25,500 seeds m−2 (Silberhorn et al., 1983; Harwell and Rhode, 2007; Lee et al., 2007). In the mixed model, flowering shoot/seed production was absent during the spring of 2006 due to the prevalence of seedlings which do not flower during their first year of growth in the Chesapeake Bay (Silberhorn et al., 1983; Orth and Moore, 1986; Jarvis and Moore, 2010). Without flowering shoots in year two during the elevated temperature scenarios, seeds were not produced, and the seed-bank was not replenished. In the York River, Z. marina seed-banks are transient and need to be replenished on a yearly basis because seeds are unable to maintain viability for periods longer than 6–12 months (Orth et al., 2000; Harwell and Orth, 2002; Jarvis, 2009). After this period any non-germinating seeds remaining in the seed bank are no longer capable of germination; therefore, the resilience provided by the sediment seed bank is eliminated. As shown in the elevated temperature model scenarios, the lack of a persistent seed bank increases the vulnerability of Chesapeake Bay Z. marina beds to consecutive years of temperature and light stress which may be problematic for long term bed persistence (Jarvis and Moore, 2010). In addition, seed-bank density is not a direct reflection of yearly seed production (Baskin and Baskin, 1998). Z. marina seeds are lost to dispersal (Källström et al., 2008), predation (Fishman and Orth, 1996), and mortality (Morita et al., 2007) reducing the potential resilience provided by the local sediment seed bank. In Ago Bay, Japan up to 72% of total seeds produced are lost from the system annually (Morita et al., 2007). In our model seeds were subject to mortality and predation after they were produced. As a result, simulated seed-bank densities were 69.1% less than total seed production. Currently, the model does not account for any deposition

of seeds produced from outside the local population or the loss of seeds through dispersal of flowering shoots. The potential of seed loss and gain through the dispersal away from and deposition into the local seed bank is not yet well quantified (Kendrick et al., 2012). However, as Z. marina seeds can disperse up to 20–300 km away from their source bed (Harwell and Orth, 2002) this is a potential source of seeds which may significantly impact the level of resilience provided by the seed bank and requires further investigation. Literature values for Z. marina sediment seed-bank densities range from 0–25,746 seeds m−2 (Harwell and Orth, 2002; Morita et al., 2007; Lee et al., 2007). Maximum model seed-bank densities were found in 2005 and were well within the literature values (maximum of 13,034 seeds m−2 ). Modeled seed-bank densities in 2006 were similar to York River observations; however, modeled seed-bank densities were significantly greater in 2007 (maximum of 9677 seeds m−2 ) compared to observed York River values (0 ± 0 seeds m−2 , Jarvis and Moore, 2010). In addition to the potential overestimation of seed bank density due to the lack of a dispersal loss term, the discrepancy between model predicted seed-bank values and observed values may be explained by the non-homogeneous development of reproductive shoots (Harwell and Rhode, 2007) and the patchy distribution of local seeds within established Z. marina beds (Harwell and Orth, 2002). Prior to deposition in the seed bank, seeds may still be attached to reproductive shoots that have detached from the main vegetative shoot and distributed in clumps rather than as individual seeds (Terrados, 1993; Harwell and Orth, 2002; Kendrick et al., 2012). Faunal interactions including direct consumption of shoots and seeds (Ganter, 2000; Heck et al., 2008; Hughes et al., 2009; Sumoski and Orth, 2012) and the movement of sediment and seeds via bioturbation (Valdemarsen et al., 2011; Delefosse and Kristensen, 2012) can

42

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

Above Ground Carbon (g C m -2)

20

10

40

Aug 05

Dec 05

Apr 06

Aug 06

Dec 06

Apr 07

Aug 07

Ambient Amb + 1 Amb + 2 Amb + 3 Amb + 4 Amb + 5

30

20

10

0 Apr 05

Dec 07

Aug 05

Dec 05

Apr 06

Aug 05

Dec 05

Apr 06

Aug 06

Dec 06

Apr 07

Aug 07

Dec 07

Aug 06

Dec 06

Apr 07

Aug 07

Dec 07

40

B

30

20

10

0 Apr

Gloucester Point

40

30

0 Apr 05

Below Ground Carbon (g C m -2)

Goodwin Island

A

Below Ground Carbon (g C m -2)

Above Ground Carbon (g C m -2)

40

Aug

Dec

Apr

Aug

Dec

Apr

Aug

Dec

Date (month)

30

20

10

0 Apr 05

Date (month)

Fig. 8. Goodwin Island and Gloucester Point model projections with sexual reproduction for 2005–2007 under ambient temperatures (solid line), ambient temperatures + 1◦ C (dotted line), ambient temperatures + 2 ◦ C (short dashed line), ambient temperatures + 3 ◦ C (short dashed and dotted line), ambient temperatures + 4 ◦ C (long dashed line), and ambient temperature + 5 ◦ C (long dashed and dotted line).

concentrate or remove seeds from the sediment seed bank, resulting in uneven seed dispersal and extremely patchy distributions making it hard to quantify overall seed bank densities (Harwell and Orth, 2002). The model described here does not have a spatial component, therefore the patchy distribution of seeds was not taken into account and all seeds were easily accounted for, possibly resulting in the greater predicted seed-bank densities. 4.2. Temperature and light Temperature within the Chesapeake Bay is predicted to increase by 2–5 ◦ C over the next century (Najjar et al., 2010). In addition, seasonal warming patterns (Hayhoe et al., 2007) and episodic heat waves (Najjar et al., 2010) are also expected to increase over time causing short term period of extreme temperature stress. The resilience provided by sexual reproduction to temperature stress within the modeled scenarios was limited, as even a 1 ◦ C increase in temperature resulted in a reduction of biomass to zero after two consecutive years of stressful temperatures. Although a temperature increase in the winter might help Z. marina production (Short and Neckles, 1999), it did not compensate for the negative effects during warmer summer months which reached a maximum of 32 ◦ C during the 2005 heat induced decline (Moore and Jarvis, 2008). This response was primarily driven by a decline in photosynthetic rate and increase in respiration when temperatures were greater than 25 ◦ C. The model output matches literature values which indicate that as water temperatures increase above 20 ◦ C Z. marina respiration increases at a greater rate than photosynthesis

causing stress and eventually mortality when water temperatures are greater than 25 ◦ C (Marsh et al., 1986; Nejrup et al., 2008). Increasing year round water temperatures by 1 ◦ C resulted in the loss of all above ground biomass from the model before cooler fall temperatures could release Z. marina populations from summer water temperature limitation. The response of temperature stressed Z. marina shoots near the southern limit of species distribution in the western Atlantic, to increased water temperature requires further research as global water temperatures are predicted to rise at an increasing rate over the next century (IPCC, 2007). Z. marina populations within the Chesapeake Bay are also stressed by low light conditions (Dennison et al., 1993; Moore et al., 1996, 2012). Under ambient temperatures, the combination of light and temperature stress present at Gloucester Point resulted in overall lower production and resilience to declines compared to Goodwin Island. This may be due to increased light requirements for Z. marina associated with increasing temperatures (Marsh et al., 1986; Moore et al., 2012). These results highlight possible mechanisms behind reported limitations in Z. marina restoration success and natural recovery of Z. marina beds in the York River which have been attributed to seasonal pulses in turbidity (Moore et al., 1996). When low light conditions are combined with increased water temperature the effects become lethal to Z. marina (Moore and Jarvis, 2008; Moore et al., 2012, 2013). However, while not modeled here, increasing light availability above that currently observed at Goodwin Island may help these populations persist under the stress of higher temperatures (Moore and Jarvis, 2008; Moore et al., 2012).

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

Above Ground Biomass (g C m )

50

Amb Temp

-2

45

43

Temp + 1 ºC

40

Ambient Elevated

35 30 25 20 15 10 5 0

Above Ground Biomass (g C m )

50 -2

45

Temp + 2 ºC

Temp + 3 ºC

Temp + 4 ºC

Temp + 5 ºC

40 35 30 25 20 15 10 5 0

Above Ground Biomass (g C m-2 )

50 45 40 35 30 25 20 15 10 5

05 ec 05 A pr 06 A ug 06 D ec 06 A pr 07 A ug 07 D ec A 07 pr 05 A ug 05 D ec 05 A pr 06 A ug 06 D ec 06 A pr 07 A ug 07 D ec 07 D

A ug

A

pr 0

5

0

Date

Date

Fig. 9. Model projections of Goodwin Island above ground biomass with both ambient (solid line) and elevated (dashed line) water column light attenuation conditions under all temperature scenarios.

4.3. Model limitations The model presented here reproduced the general observed trends in above and below ground Z. marina biomass in the York River following the 2005 heat induced decline; however, it does have several limitations. One of the greatest percent errors in base model calibration occurred due to a significant underestimate (up to 84 ± 5%) of fall Z. marina production which may be attributed to the use of constant rates for translocation of carbon from Z. marina above ground to below ground biomass. The lack of above ground production due to temperatures >25 ◦ C (Marsh et al., 1986; Nejrup and Pedersen, 2008; Hosokawa et al., 2009; Höffle et al., 2010) may inhibit carbon translocation to below ground biomass; however, as in Madden and Kemp (1996) and Cerco and Moore (2001), the exact relationship is unknown so translocation was held constant throughout all model runs. Defining the seasonality of the relationships between temperature, translocation, and leaf sloughing within Z. marina plants is necessary to increase the accuracy of the model. There were several limitations on the accuracy of sexual reproductive output in the model resulting in key components, such as the total number of seeds produced, to be overestimated. The areas that are primarily lacking include flowering shoot development, tracking of seeds into and out of the modeled population via

dispersal processes, and seed bank development over time. Both flowering shoot development and site can impact successful sexual reproduction by affecting seed quality (Carter and Blair, 2013). Seed source is affected by a variety of factors including temperature, light, and predation (Setchell, 1929; Phillips et al., 1983; Silberhorn et al., 1983) as these factors can impact flowering shoot and ultimately seed development. Not all impacts are lethal to the flowering shoot; however, they may result in reduced seed production or lower seed viability (Harper, 1977; Carter and Blair, 2013). As the relationship between environmental factors, seed development, and seed bank viability is not well defined, this remains a limitation of the model. Once the model was expanded to include data from 2005 to 2007, the above ground biomass predictions for Gloucester Point were up to three times greater in the model run compared to observed values. The over-prediction is related to larger initial above ground seedling biomass following the 2005 decline. Relationships between seedling growth and survival and surrounding environmental conditions are not well defined. There is some evidence that seedlings respond similarly to temperature limitations when compared to established Z. marina plants (Bintz and Nixon, 2001; Abe et al., 2008); however, there is little other information available on Z. marina seedlings or the effects of changes in habitat conditions on seedling growth and survival. Information on

44

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45

seedling physiology would enable parameterization of a separate seedling sub-model to track seedlings in their first year of growth likely increasing the overall accuracy of the sexual reproduction component of the model. 4.4. Conclusions The results presented here highlight a need for the inclusion of sexual reproduction within Z. marina production models. This directly applies to models attempting to project loss and recovery processes within Z. marina communities, as sexual reproduction plays a large role in recovery from large scale declines. In addition, model projections indicate that current York River Z. marina populations are near the limit for temperature stress and increases in water temperatures by the predicted 1 ◦ C may have large impacts on Z. marina survival if there is no increase in light availability. The loss of Z. marina beds due to temperature stress or combined temperature and light stress can be ameliorated to some degree by sexual reproduction. However, this resilience is limited by seed production and seed-bank viability. Multiple concurrent years of stress may result in long term loss of Z. marina in the lower Chesapeake Bay. Further research into all aspects of sexual reproduction, seed dispersal, seedling establishment, growth, and survival in temperate Z. marina beds is required to advance the understanding of loss and recovery patterns in Chesapeake Bay Z. marina populations and similar systems. Acknowledgements The authors would like to thank the National Estuarine Research Reserve Graduate Research Fellowship Program and the Virginia Institute of Marine Science Graduate Research Assistantship Program for funding. We would also like to thank Erin Shields, Brittany Haywood, Betty Neikirk, and Brandon Jarvis for field and laboratory assistance. This is contribution number 3330 from the Virginia Institute of Marine Science, College of William and Mary. References Abe, M., Kurashima, A., Maegawa, M., 2008. Temperature requirements for seed germination and seedling growth of Zostera marina from central Japan. Fish. Sci. 74, 589–593, http://dx.doi.org/10.1111/j.1444-2906.2008.01562.x. Aveytua-Alcázar, L., Camacho-Ibar, V.F., Souza, A.J., Allen, J.I., Torres, R., 2008. Modeling Zostera marina and Ulva spp. in a coastal lagoon. Ecol. Model. 218, 354–366, http://dx.doi.org/10.1016/j.ecolmodel.2008.07.019. Bach, H.K., 1993. A dynamic model describing the seasonal variations in growth and distribution of Z. marina (Zostera marina L.) I. Model theory. Ecol. Model. 65, 31–50. Baskin, C.B., Baskin, J., 1998. Seeds: Ecology, Biogeography, and Evolution of Dormancy and Germination. Academic Press, London. Bintz, J.C., Nixon, S.W., 2001. Responses of eelgrass Zostera marina seedlings to reduced light. Mar. Ecol. Prog. Ser. 223, 133–141, http://dx.doi.org/ 10.3354/Meps223133. Burkholder, J.M., Tomasko, D.A., Touchette, B.W., 2007. Seagrasses and eutrophication. J. Exp. Mar. Biol. Ecol. 350, 46–72, http://dx.doi.org/10.1016/ j.jembe.2007.06.024. Buzzelli, C.P., Wetzel, R.L., Meyers, M.B., 1999. A linked physical and biological framework to assess biogeochemical dynamics in a shallow estuarine ecosystem. Estuar. Coast. Shelf Sci. 49, 829–851, http://dx.doi.org/10.1006/ecss.1999.0556. Carter, D.L., Blair, J.M., 2013. Seed source has variable effects on species, communities, and ecosystem properties in grassland restorations. Ecosphere 4, 1–16, http://dx.doi.org/10.1890/es13-00090.1. Cerco, C.F., Moore, K., 2001. System-wide submerged aquatic vegetation model for Chesapeake Bay. Estuaries 24, 522–534, http://dx.doi.org/10.2307/1353254. Combroux, I., Bornette, G., Willby, N.J., Amoros, C., 2001. Regenerative strategies of aquatic plants in disturbed habitats: the role of the propagule bank. Arch. Hydrobiol. 152, 215–235. Delefosse, M., Kristensen, E., 2012. Burial of Zostera marina seeds in sediment inhabited by three polychaetes: laboratory and field studies. J. Sea Res. 71, 41–49, http://dx.doi.org/10.1016/j.seares.2012.04.006. den Hartog, C., 1970. The Seagrasses of the World. North Holland Publishing, Amsterdam. Dennison, W.C., Orth, R.J., Moore, K.A., Stevenson, J.C., Carter, V., Kollar, S., Bergstrom, P.W., Batiuk, R.A., 1993. Assessing water quality with submersed aquatic

vegetation: habitat requirements as barometers of Chesapeake Bay health. Bioscience 43, 86–94, http://dx.doi.org/10.2307/1311969. Duarte, C.M., 2002. The future of seagrass meadows. Environ. Conserv. 29, 192–206, http://dx.doi.org/10.1017/S0376892902000127. Evans, A.S., Webb, K.L., Penhale, P.A., 1986. Photosynthetic temperature acclimation in two coexisting seagrasses, Zostera marina L. and Ruppia maritima L. Aquat. Bot. 24, 185–197, http://dx.doi.org/10.1016/0304-3770(86)90095-1. Fishman, J.R., Orth, R.J., 1996. Effects of predation on Zostera marina L. seed abundance. J. Exp. Mar. Biol. Ecol. 198, 11–26, http://dx.doi.org/ 10.1016/0022-0981(95)00176-X. Ganter, B., 2000. Seagrass (Zostera spp.) as food for brent geese (Branta bernicla): an overview. Helgol. Mar. Res. 54, 63–70, http://dx.doi.org/10.1007/s101520050003. Gibson, J.R., Najjar, R.G., 2000. The response of Chesapeake Bay salinity to climate-induced changes in stream flow. Limnol. Oceanogr. 45, 1764–1772. Greve, T.M., Krause-Jensen, D., Rasmussen, M.B., Christensen, P.B., 2005. Means of rapid eelgrass (Zostera marina L.) recolonisation in former dieback areas. Aquat. Bot. 82, 143–156, http://dx.doi.org/10.1016/j.aquabot.2005.03.004. Harper, J.L., 1977. Population Biology of Plants. Academic Press, London, UK. Harwell, M.C., 2000. Ecological dispersal mechanisms, reproductive ecology and the importance of scale in Zostera marina in Chesapeake Bay. PhD Dissertation. Virginia Institute of Marine Science, Gloucester Point, VA. Harwell, M.C., Orth, R.J., 2002. Seed-bank patterns in Chesapeake Bay eelgrass (Zostera marina L.): a bay-wide perspective. Estuaries 25, 1196–1204, http://dx.doi.org/10.1007/BF02692216. Harwell, M.C., Rhode, J.M., 2007. Effects of edge/interior and patch structure on reproduction in Zostera marina L. in Chesapeake Bay, USA. Aquat. Bot. 87, 147–154, http://dx.doi.org/10.1016/j.aquabot.2007.04.007. Hayhoe, K., Wake, C.P., Huntington, T.G., Luo, L.F., Schwartz, M.D., Shefield, J., Wood, E., Anderson, B., Bradbury, J., DeGaetano, A., Troy, T.J., Wolfe, D., 2007. Past and future changes in climate and hydrological indicators in the US Northeast. Clim. Dynam. 28, 381–407, http://dx.doi.org/10.1007/s00382-006-0187-8. Heck Jr., K.L., Carruthers, T.J.B., Duarte, C.M., Hughes, A.R., Kendrick, G., Orth, R.J., Wiliams, S.W., 2008. Trophic transfers from seagrass meadows subsidize diverse marine and terrestrial consumers. Ecosystems 11, 1198–1210, http://dx.doi.org/10.1007/s10021-008-9155-y. Höffle, H., Thomsen, M.S., Holmer, M., 2010. High mortality of Zostera marina under high temperature regimes but minor effects of the invasive macroalgae Gracilaria vermiculophylla. Estuar. Coast. Shelf Sci. 92, 35–46, http://dx.doi.org/10.1016/j.ecss.2010.12.017. Hootsmans, M.J.M., Vermaat, J.E., Van Vierssen, W., 1987. Seed-bank development, germination, and early seedlings survival of two seagrass species from the Netherlands: Zostera marina L. and Zostera noltii Hornem. Aquat. Bot. 28, 275–285, http://dx.doi.org/10.1016/0304-3770(87)90005-2. Hosokawa, S., Nakamura, Y., Kuwae, T., 2009. Increasing temperature induces shorter leaf life span in an aquatic plant. Oikos 118, 1158–1163, http://dx.doi.org/10.1111/j.1600-0706.2009.17288.x. Hughes, A.R., Wiliams, S.L., Duarte, C.M., Heck Jr., K.L., Waycott, M., 2009. Associations of concern: declining seagrasses and threatened dependent species. Front. Ecol. Environ. 7, 242–246, http://dx.doi.org/10.1890/080041. Intergovernmental Panel on Climate Change (IPCC), 2007. Climate change 2007: the physical science basis. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom/New York, NY, USA. Jarvis, J.C., 2009. Function of seed bank ecology in Mid-Atlantic semi-annual and perennial Zostera marina beds. PhD Dissertation. School of Marine Science, Virginia Institute of Marine Science, College of William and Mary. Jarvis, J.C., Moore, K.A., 2010. The role of seedlings and seed bank viability in the recovery of Chesapeake Bay, USA, Zostera marina populations following a large-scale decline. Hydrobiologia 649, 55–68, http://dx.doi.org/10.1007/s10750-010-0258-z. Jarvis, J.C., Moore, K.A., Kenworthy, W.J., 2012. Characterization and ecological implication of eelgrass life history strategies near the species’ southern limit in the western North Atlantic. Mar. Ecol. Prog. Ser. 444, 43–56, http://dx.doi.org/10.3354/meps09428. Källström, B., Nyqvist, A., Åberg, P., Bodin, M., André, C., 2008. Seed rafting as a dispersal strategy for eelgrass (Zostera marina). Aquat. Bot. 88, 148–153, http://dx.doi.org/10.1016/j.aquabot.2007.09.005. Kendrick, G.A., Waycott, M., Carruthers, T.J.B., Cambridge, M.L., Hovey, R., Krauss, S.L., Lavery, P.S., Les, D.H., Lowe, R.J., Mascaró, O., Vidal, I., Ooi, J.L.S., Orth, R.J., Rivers, D.O., Ruiz-Montoya, L., Sinclair, E.A., Statton, J., Kornelis van Dijk, J., Verduin, J.J., 2012. The central role of dispersal in the maintenance and persistence of seagrass populations. Bioscience 62, 56–65, http://dx.doi.org/10.1525/bio.2012.62.1.10. Kirk, J.T.O., 1983. Light and Photosynthesis in Aquatic Ecosystems. Cambridge University Press, Cambridge, UK. Leck, M.A., Parker, V.T., Simpson, R.L., 1989. Ecology of Soil Seed-banks. Academic Press, Inc., Sand Diego, CA, USA. Lee, K.S., Park, S.R., Kim, Y.K., 2007. Effects of irradiance, temperature, and nutrients on growth dynamics of seagrasses: a review. J. Exp. Mar. Biol. Ecol. 350, 144–175, http://dx.doi.org/10.1016/j.jembe.2007.06.016. Madden, C.J., Kemp, W.M., 1996. Ecosystem model of an estuarine submersed plant community: calibration and simulation of eutrophication responses. Estuaries 19, 457–474, http://dx.doi.org/10.2307/1352463.

J.C. Jarvis et al. / Aquatic Botany 113 (2014) 32–45 Marsh, J.A., Dennison, W.C., Alberte, R.S., 1986. Effects of temperature on photosynthesis and respiration in eelgrass (Zostera marina L.). J. Exp. Mar. Biol. Ecol. 101, 257–267, http://dx.doi.org/10.1016/0022-0981(86)90267-4. Meling-López, A.E., Ibarra-Obando, S.E., 1999. Annual life cycles of two Zostera marina L. populations in the Gulf of California: contrasts in seasonality and reproductive effort. Aquat. Bot. 65, 56–69, http://dx.doi.org/10.1016/s0304-3770(99)00031-5. Moore, K.A., Orth, R.J., Nowak, J.F., 1993. Environmental regulation of seed germination in Zostera marina L. (eelgrass) in Chesapeake Bay: effects of light, oxygen, and sediment burial. Aquat. Bot. 45, 79–91, http://dx.doi.org/10.1016/0304-3770(93)90054-Z. Moore, K.A., Neckles, H.A., Orth, R.J., 1996. Zostera marina (eelgrass) growth and survival along a gradient of nutrients and turbidity in the lower Chesapeake Bay. Mar. Ecol. Prog. Ser. 142, 247–259, http://dx.doi.org/10.3354/Meps142247. Moore, K.A., 2004. Influence of seagrasses on water quality in shallow regions of the lower Chesapeake Bay. J. Coast. Res. SI(45), 162–178, http://dx.doi.org/10.2112/SI45-162.1. Moore, K.A., Jarvis, J.C., 2008. Environmental factors affecting recent summertime eelgrass diebacks in the lower Chesapeake Bay: implications for long-term persistence. J. Coast. Res. SI(55), 135–147, http://dx.doi.org/10.2112/Si55-014. Moore, K.A., Shields, E.C., Parrish, D.B., Orth, R.J., 2012. Eelgrass survival in two contrasting systems: role of turbidity and summer water temperatures. Mar. Ecol. Prog. Ser. 448, 247–258, http://dx.doi.org/10.3354/Meps09578. Moore, K.A., Shields, E.C., Parrish, D.B., 2013. Impacts of varying estuarine temperature and light conditions on Zostera marina (eelgrass) and its interactions with Ruppia maritima (widgeon grass). Estuar. Coast., http://dx.doi.org/10.1007/s12237-013-9667-3. Morita, T., Okumura, H., Abe, M., Kurashima, A., Maegawa, M., 2007. Density and distribution of seeds in bottom sediments in Zostera marina beds in Ago Bay, central Japan. Aquat. Bot. 87, 38–42, http://dx.doi.org/10.1016/j.aquabot.2007.03.001. Najjar, R.G., 1999. The water balance of the Susquehanna River Basin and its response to climate change. J. Hydrol. 219, 7–19, http://dx.doi.org/10.1016/S0022-1694(99)00041-4. Najjar, R.G., Pyke, C.R., Adams, M.B., Breitburg, D., Hershner, C., Kemp, M., Howarth, R., Mulholland, M.R., Paolisso, M., Secor, D., Sellner, K., Wardrop, D., Wood, R., 2010. Potential climate-change impacts on the Chesapeake Bay. Estuar. Coast. Shelf Sci. 86, 1–20, http://dx.doi.org/10.1016/j.ecss.2009.09.026. Nejrup, L.B., Pedersen, M.F., 2008. Effects of salinity and water temperature on the ecological performance of Zostera marina. Aquat. Bot. 88, 239–246, http://dx.doi.org/10.1016/j.aquabot.2007.10.006. Newton, A., Mudge, S.M., 2003. Temperature and salinity regimes in a shallow, mesotidal lagoon, the Ria Formosa, Portugal. Estuar. Coast. Shelf Sci. 57, 73–85, http://dx.doi.org/10.1016/S0272-7714(02)00332-3. Orth, R.J., Moore, K.A., 1983. Chesapeake Bay: an unprecedented decline in submerged aquatic vegetation. Science 222, 51–53, http://dx.doi.org/10.1126/science.222.4619.51. Orth, R.J., Moore, K.A., 1986. Seasonal and year-to-year variations in the growth of Zostera marina L. (eelgrass) in the lower Chesapeake Bay. Aquat. Bot. 24, 335–341, http://dx.doi.org/10.1016/0304-3770(86)90100-2. Orth, R.J., Harwell, M.C., Bailey, E.M., Bartholomew, A., Jawad, J.T., Lombana, A.V., Moore, K.A., Rhode, J.M., Woods, H.E., 2000. A review of issues in seagrass seed dormancy and germination: implications for conservation and restoration. Mar. Ecol. Prog. Ser. 200, 277–288, http://dx.doi.org/10.3354/Meps200277. Orth, R.J., Carruthers, T.J.B., Dennison, W.C., Duarte, C.M., Fourqurean, J.W., Heck Jr., K.L., Hughes, A.R., Kendrick, G.A., Kenworthy, W.J., Olyarnik, S., Short, F.T., Waycott, M., Williams, S.L., 2006. A global crisis for seagrass ecosystems. Bioscience 56 (12), 987–996, http://dx.doi.org/10.1641/0006-3568(2006)56[987:Agcfse]2.0.Co;2. Orth, R.J., Marion, S.R., Moore, K.A., Wilcox, D.J., 2010. Eelgrass (Zostera marina L.) in the Chesapeake Bay region of mid-Atlantic Coast of the USA: challenges in conservation and restoration. Estuar. Coast. 33, 139–150, http://dx.doi.org/10.1007/s12237-009-9234-0. Phillips, R.C., Grant, W.S., McRoy, C.P., 1983. Reproductive strategies of eelgrass (Zostera marina L.). Aquat. Bot. 16, 1–20, http://dx.doi.org/ 10.1016/0304-3770(83)90047-5.

45

Plus, M., Deslous-Paoli, J.M., Dagault, F., 2003. Seagrass (Zostera marina L.) bed recolonisation after anoxia-induced full mortality. Aquat. Bot. 77, 149–164, http://dx.doi.org/10.1016/S0304-3770(03)00089-5. Plus, M., Sébastien, D., Gilles, T., Auby, I., de Montaudouin, X., Emery, É., Claire, N., Christophe, V., 2010. Long-term evolution (1988–2008) of I spp. meadows in Arcachon Bay (Bay of Biscay). Estuar. Coast. Shelf Sci. 87, 357–366, http://dx.doi.org/10.1016/j.ecss.2010.01.016. Preston, B.L., 2004. Observed winter warming of the Chesapeake Bay Estuary (1949–2002): implications for ecosystem management. Environ. Manage. 34, 125–139, http://dx.doi.org/10.1007/s00267-004-0159-x. Probert, R.J., Brenchly, J.L., 1999. The effect of environmental factors on field and laboratory germination in a population of Zostera marina L. from southern England. Seed Sci. Res. 9, 331–339, http://dx.doi.org/10.1017/S0960258599000343. Robertson, A.I., Mann, K.J., 1984. Disturbance of ice and life-history adaptations of the seagrass Zostera marina. Mar. Biol. 80, 131–141, http://dx.doi.org/ 10.1007/Bf02180180. Santamaría-Gallegos, N.E., Sánchez-Lizaso, J.L., Félix-Pico, E.F., 2000. Phenology and growth cycle of annual subtidal eelgrass in a subtropical locality. Aquat. Bot. 66, 329–339, http://dx.doi.org/10.1016/s0304-3770(99)00082-0. Setchell, W.A., 1929. Morphological and phonological notes on Zostera marina L. Univ. Calif. Publs. Bot. 14, 389–452. Shafter, D., Bergstrom, P., 2008. Large scale submerged aquatic vegetation restoration in the Chesapeake Bay. Final Report to the U.S. Army Corps of Engineers, ERDC/EL TR-09-20., pp. 63. Short, F.T., Wyllie-Echeverria, S., 1996. Natural and human-induced disturbance of seagrasses. Environ. Conserv. 23, 17–27, http://dx.doi.org/ 10.1017/S0376892900038212. Short, F.T., Neckles, H.A., 1999. The effects of global climate change on seagrasses. Aquat. Bot. 63, 169–196, http://dx.doi.org/10.1016/S0304-3770(98)00117-X. Short, F.T., Moore, K.A., 2006. Zostera: biology, ecology, and management. In: Larkum, A.W.D., Orth, R.J., Duarte, C.M. (Eds.), Seagrasses: Biology, Ecology, and Conservation. , pp. 361–386. Short, F.T., Polidoro, B., Livingstone, S.R., Carpenter, K.E., Bandeira, S., Bujang, J.S., Calumpong, H.P., Carruthers, T.J.B., Coles, R.G., Dennison, W.C., Erftmeijer, P.L.A., Fortes, M.D., Freeman, A.S., Jagtap, T.G., Kamal, A.H.M., Kendrick, G.A., Kenworthy, W.J., La Nafie, Y.A., Nasution, I.M., Orth, R.J., Prathep, A., Sanciangco, J.C., Tussenbroek, B.V., Vergara, S.G., Waycott, M., Zieman, J.C., 2011. Extinction risk assessment of the world’s seagrass species. Biol. Conserv. 144, 1961–1971, http://dx.doi.org/10.1016/j.biocon.2011.04.010. Silberhorn, G.M., Orth, R.J., Moore, K.A., 1983. Anthesis and seed production in Zostera marina L. (eelgrass) from the Chesapeake Bay. Aquat. Bot. 15, 133–144, http://dx.doi.org/10.1016/0304-3770(83)90024-4. Sumoski, S.E., Orth, R.J., 2012. Biotic dispersal in eelgrass Zostera marina. Mar. Ecol. Prog. Ser. 471, 1–10, http://dx.doi.org/10.3354/Meps10145. Terrados, J., 1993. Sexual reproduction and seed-banks of Cymodocea nodosa (Ueria) Ascherson meadows on the southeast Mediterranean coast of Spain. Aquat. Bot. 46, 239–299, http://dx.doi.org/10.1016/0304-3770(93)90009-L. Valdemarsen, T., Wendelboe, K., Egelund, J.T., Kristensen, E., Flindt, M.R., 2011. Burial of seeds and seedlings by the lugworm Arenicola marina hampers eelgrass (Zostera marina) recovery. J. Exp. Mar. Biol. Ecol. 410, 45–52, http://dx.doi.org/10.1016/j.jembe.2011.10.006. van Lent, F., 1995. Intraspecific variability of Zostera marina L. in the southwestern Netherlands. PhD Dissertation. Laboratory of Aquatic Ecology, Catholic University of Nijmegen, Nijmegen, The Netherlands. Waycott, M., Duarte, C.M., Carruthers, T.J.B., Orth, R.J., Dennison, W.C., Olyarnik, S., Calladine, A., Forqurean, J.W., Heck Jr., K.L., Hughes, A.R., Kendrick, G.A., Kenworthy, W.J., Short, F.T., Williams, S.L., 2009. Accelerating loss of seagrasses across the globe threatens coastal ecosystems. Proc. Natl. Acad. Sci. U.S.A. 106, 12377–12381, http://dx.doi.org/10.1073/pnas.0905620106. Wetzel, R.L., Neckles, H.A., 1986. A model of Zostera marina L. photosynthesis and growth: simulated effects of selected physical-chemical variables and biological interactions. Aquat. Bot. 26, 307–323, http://dx.doi.org/10.1016/0304-3770(86)90029-X.