Marine Biology

, Volume 158, Issue 7, pp 1473–1482

Distinctive types of leaf tissue damage influence nutrient supply to growing tissues within seagrass shoots

Authors

    • Institut de Reçerca i Tecnologia AgroalimentàriesSant Carles de la Ràpita
  • Catherine J. Collier
    • School of Marine and Tropical BiologyJames Cook University
  • Javier Romero
    • Departamento de Ecología, Facultad de BiologíaUniversidad de Barcelona
  • Teresa Alcoverro
    • Centre d’Estudis Avançats de BlanesCSIC
Original Paper

DOI: 10.1007/s00227-011-1664-0

Cite this article as:
Prado, P., Collier, C.J., Romero, J. et al. Mar Biol (2011) 158: 1473. doi:10.1007/s00227-011-1664-0

Abstract

Herbivory is now recognized as an important structuring agent in seagrass meadows but the attack pattern and tissue damage of consumers are highly variable. Nutritional preferences of herbivores and/or easy access to resources may cause differences in biomass loss among tissues that damage the plant in functionally distinctive ways. The two main Mediterranean herbivores, the fish Sarpa salpa (L.) and the sea urchin Paracentrotus lividus (Lmk.), remove higher amounts of intermediate and external shoot leaves, respectively. To test whether this selective feeding can have different consequences on the allocation patterns of nutrient within plants, we simulated the effect of both herbivores by clipping external and intermediate leaves (plus unclipped controls) of Posidonia oceanica (L.) and we measured plant tolerance in terms of shoot growth and leaf nutrient supply to new tissue using isotopic markers. As expected, control treatments displayed high carbon and nutrient supply from external leaves (83% of the total 15N and 84% of the total 13C incorporated by the shoot). When subjected to clipping, the remaining leaves enhanced carbon and nitrogen supply compared with the control by 16% of N and 36% of C—in the intermediate clipping—and by over 100% of N and 200% of C—in the external clipping—to compensate for the nutrient lost. However, only in the case of fish herbivory (intermediate clipping), enhanced supply alone was able to fully compensate for the nutrient losses. In contrast, this mechanism is not completely effective when external leaves are clipped (urchin herbivory). Yet, the consequences of this nutrient loss under sea urchin herbivory are not apparent from the nutrient content of the new tissue, suggesting that there are other sources of nitrogen income (uptake or reallocation from rhizomes). Our study does not only confirm the tolerance of P. oceanica to herbivory, but also constitutes the first evidence of leaf-specific, compensatory nutrient supply in seagrasses.

Introduction

Removal of tissue by grazers causes structural damage and nutrient and carbohydrate losses, and may ultimately reduce resource acquisition, growth, and reproduction (Marquis 1992; Crawley 1997). These potential deleterious effects can be alleviated through diverse mechanisms including adaptive plant structure and growth habit (Liu et al. 2007), reallocation of assimilates from remaining leaves (Sosebee and Wiebe 1971; Ryle and Powell 1975) and rhizomes (Harnett 1989; Vergés et al. 2008), increased photosynthetic rates in the remaining tissues (Detling et al. 1979; Caldwell et al. 1981), modifications of the hormonal balance (Avery and Briggs 1968; Avery and Lacey 1968), or increased nutrient uptake rates from roots (Caldwell et al. 1981). The mechanisms in place are, in part, dependent on the intensity of the grazing pressure. However, nutritional preferences and behavioral aspects of herbivore feeding (e.g. resource accessibility and/or aggregative feeding behavior) may also damage the plant in functionally distinctive ways that can influence plant fitness as much as the overall quantities of tissue lost (review by Kotanen and Rosenthal 2000). Damage by vertebrates is often spatially and temporally stochastic (Varnamkhasti et al. 1995) and can be sudden and severe, a reflection of their large size and basal metabolic requirements (Demment and Soest 1985). In contrast, invertebrate attacks often involve prolonged removal of small amounts of tissue, which causes a continuous drain of resources (e.g. Fay et al. 1996; Zimmerman et al. 2001). Tissue specificity is generally greater for species of invertebrates than for vertebrates (Strong et al. 1984; Crawley 1989) and, depending on damage type, often causes indirect impacts on plant physiology.

Seagrass beds have been shown to withstand intense rates of herbivore activity (see review by Heck and Valentine 2006). Several types of herbivores feed on seagrasses (e.g. Sirenians, green turtles, fishes, and sea urchins), and they often coexist but exhibit very distinctive ways of removing plant material. Feeding by Sirenians (i.e. dugongs and manatees) often involves the removal of the entire plant, including roots and rhizomes, and favors the dominance of early- or mid-successional species in frequently grazed meadows (Aragones and Marsh 2000; Aragones et al. 2006). Green turtles have a preference for young, actively growing tissues but, by cropping the seagrass just a few centimeters above the surface of the substrate, large blade areas above the cropped region are also lost (Bjorndal 1980; Thayer et al. 1984). Since large seagrass vertebrates remove plant tissue in bulk, tolerance to these herbivores may often be linked to more general genotypic traits such as clonal morphology, high rates of leaf turnover, and flexible allocation of resources (Nakaoka and Aioi 1999; Aragones et al. 2006). In contrast, patterns of grazing by fishes are more diverse and either show relatively homogeneous leaf mowing (Tomas et al. 2005; Ferrari et al. 2007) or are intensely patchy, causing the formation of “halos” (Randall 1965; McAfee and Morgan 1996), while selective consumption of epiphytised blade tips has also been reported (Lobel and Ogden 1981). All these can depend on factors such as seagrass and herbivore abundance, habitat heterogeneity and complexity and patterns of movement of the species (Macià and Robbinson 2005; Unsworth et al. 2007; Prado et al. 2008a). Marine invertebrates, such as sea urchins that are in permanent contact with the sea bottom, tend to feed on the more down leaning accessible sections of the plant such as external leaves (Shepherd 1987) or basal meristems if they are exposed (Alcoverro and Mariani 2002). The feeding behaviors of fish and urchins damage plants in functionally distinct ways, and as a result, the response of the plant, including nutritional compensatory mechanisms, may also differ. Growth-compensatory responses like enhanced levels of N-metabolizing enzymes and/or post-damage reallocation of internal resources may be common mechanisms to tolerate more idiosyncratic types of seagrass damage (Zimmerman et al. 2001; Valentine et al. 2004; Vergés et al. 2008) but effects under tissue-specific herbivore attack are still unknown.

Posidonia oceanica (L.) is the dominant seagrass species in oligotrophic Mediterranean waters. It features an exceptionally long leaf life span among seagrasses (up to 300 days, Romero 1989), which may account for some of the highest contributions of nutrient resorption to meet growth requirements ever reported (up to 40%, Alcoverro et al. 2000). In addition, plant losses to herbivores may be alleviated by enhanced reallocation from stored reserves, but not by increasing the photosynthetic rate (Vergés et al. 2008). In these ecosystems, two important types of leaf macrograzers coexist, the Sparid fish Sarpa salpa (L.) and the sea urchin Paracentrotus lividus (Lmk.). Regionally, these herbivores may remove, on average, 57% of the annual leaf primary production of the plant in shallow (5 m) meadows (Prado et al. 2007). However, grazing pressure and herbivore abundance are very heterogeneous in both time and space. Some locations have very high sea urchin abundances (up to ca. 20 ind. m−2), and during the period of higher herbivory (i.e. August–September), they can consume up to 139% of monthly leaf primary production. In other locations, fishes are the dominant herbivores with consumption rates that in summer may exceed local leaf production by 250% (Prado et al. 2007). The fish S. salpa is a mid-water forager and thus, usually removes the older sections of intermediate, upright leaves whereas sea urchins more commonly consume the older, down leaning sections of external leaves (Shepherd 1987; Cebrián et al. 1996a; Pinna et al. 2009), each together with differential epiphytic communities developed with leaf age (Prado et al. 2010). Hence, these distinctive grazing behaviors may interact with the dynamic balance between the loss of old leaves and the production of new ones (Sand-Jensen et al. 1994). The objectives of this study were to evaluate, by herbivory simulation experiments, the effects of two types of herbivore damage (fish loss of the upper part of the intermediate leaves and sea urchin loss of the upper part of the external leaves) on seagrass performance. More specifically, we used isotopic markers to (1) identify possible changes in leaf nutrient supply to young emergent leaves and (2) to assess tissue growth and the nutritional status of young leaves.

Materials and methods

Study site and experimental design

We first aimed to confirm whether sea urchin and fish displayed distinctive feeding, potentially causing differential leaf damage. To do this, we reanalyzed data of leaf consumption by Prado et al. (2007) obtained in 10 different shallow sites selected according to regional availability and accessibility and at four different times during the year (i.e. one per season), using tethered shoots (n = 20 per site and season). Mean averages of external and intermediate leaf loss to each grazer per site and season (n = 40) were tested with paired t tests and confirmed higher feeding rates by the sea urchin P. lividus in external leaves (t = 6.01, df = 40, P < 0.001) and higher rates in intermediate leaves by the fish S. salpa (t = 5.41, df = 40, P < 0.001; Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-011-1664-0/MediaObjects/227_2011_1664_Fig1_HTML.gif
Fig. 1

Defoliation rate (cm2 shoot−1 day−1) of external and intermediate leaves by the fish S. salpa and the sea urchin P. lividus (mean ± SE). Values were obtained from seasonal values in Prado et al. (2007) across the study region (n = 40). Significant differences in defoliation rates among leaf positions for each herbivore are indicated with asterisks (***P < 0.001)

Plant responses (shoot elongation and nutrient supply) of P. oceanica shoots subjected to these two types of herbivore damage (sea urchin and fish) were then evaluated by a clipping experiment in a seagrass meadow in Cala Giverola (41°44′ N, 002°57′ E, NW Mediterranean) during late summer (early to mid-September)—the period of highest herbivory pressure (Prado et al. 2007). We worked at 5 m depth, in a patchy, semi-exposed seagrass meadow. Nutrient conditions were oligotrophic with concentrations of about 0.07 μM NH4, 0.96 μM NO3, and 0.29 μM PO4 (annual averages), and seawater temperature ranged between 22 and 24°C (Cebrián et al. 1996b).

A total of 60 seagrass shoots were carefully selected underwater making sure that each had five leaves (i.e. two external, two intermediate, and one internal), over an area of 20 m2. Treatments consisted of the factorial combination of two labeling positions (external leaves labeled and intermediate leaves labeled) and three clipping levels (no clipping, clipping in intermediate leaves, clipping in external leaves), thus resulting in six experimental conditions (see Fig. 2), each with ten replicated shoots. The treatments were assigned at random over the 20 m2 area. Clipping consisted of the removal of up to ca. 45 cm of leaf (either intermediate or external, generally less for intermediate leaves), which is roughly equivalent to the average value removed per month during the August–September herbivory peak (Prado et al. 2007). The remaining portion of the clipped leaf was around 15 cm, to ensure similar surface area for uptake during incubation. All experimental shoots were hole-punched with a needle to measure leaf production according to Romero (1989) and then marked with a peg inserted in the sediment for later retrieval and treatment identification.
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-011-1664-0/MediaObjects/227_2011_1664_Fig2_HTML.gif
Fig. 2

Illustration of the six different experimental treatments accordingly to the position of the clipped and the labeled leaves. a Unclipped, external labeled, b unclipped, intermediate labeled, c external clipped, external labeled, d external clipped, intermediate labeled, e intermediate clipped, external labeled, f intermediate clipped, intermediate labeled. An arrow indicates the position of the internal leaf

Isotopic incubation

N and C stable isotopes were used as traces to elucidate the effect of clipping of leaves on the uptake and reallocation of these elements toward growing tissues. Natural levels of these isotopes are low and quite consistent under uniform environmental conditions, therefore, when supplied, they can be easily detected and traced above background levels. To this end, clipped or unclipped external and intermediate leaves, depending on the treatment (see Fig. 2 for details), were incubated within an isotopic media. The incubation chamber was made of a plastic bag fitted with a 50-mm syringe filter holder and a one-way plastic stopcock (both from Cole–Parmer) that allowed closure after isotopic injection. Bags were pulled over the incubated leaves and then sealed at the base with a padded clip. The efficiency of the seal was assessed by previous in situ trials using colorants and showed no visual leakage (Prado et al. 2008b).

For incubation of unclipped leaves, the chamber was 0.5 l, whereas for clipped leaves (ca. 15 cm), the chamber was 0.25 l (Fig. 2). This ensured a more consistent leaf area to volume ratio between the treatments and allowed a more adequate fastening of the plastic bag to the base of the leaf. The chambers were filled by injecting seawater with a syringe to the desired volume, and then, the isotopic solution was injected. For all treatments, the injected solution contained NaH13CO3 and 15NH4Cl to obtain a final concentration of 300 μM NaH13CO3 and 40 μM 15NH4Cl within the chamber (Marbà et al. 2002; Prado et al. 2008b). Nitrogen was supplied as NH4+, for which seagrasses have higher uptake affinity than for NO3, in order to maximize enrichment (Touchette and Burkholder 2000). Leaves were left to incubate for 4 h, and then, the chambers were removed by taking away the bag while pinching at the base to retain the fluid within the bag. Throughout removal, seawater was rapidly moved across the shoot to prevent possible external contamination of unlabeled leaves. Isotope-marked shoots were left at the study site for 15 days to allow for transport of nutrients from the labeled leaves to the rest, including very young leaves (primords) produced during the experimental period. Then, all shoots, including leaves and a section of the associated rhizome, were retrieved and carefully placed into plastic bags for further processing in the laboratory. Non-labeled shoots of P. oceanica were also collected within the area in order to determine ambient values of δ15N and δ13C in internal leaves. Additionally, ten shoots, which were immediately adjacent to the labeled shoots, but not connected by any rhizome, were collected in order to further verify whether there were possible leakages from the incubation chambers (see Table 1).
Table 1

C:N ratio, δ15N and δ13C in the internal leaf and primords of P. oceanica (mean ± SE) by experimental treatment (Unclipped, External clipped and Intermediate clipped) (n = 10)

Treatment

Unclipped

External clipped

Intermediate clipped

EL

IL

EL

IL

EL

IL

C:N

16.23 ± 0.45

19.85 ± 1.03

17.50 ± 0.71

19.15 ± 1.29

16.27 ± 0.83

18.02 ± 0.50

δ15N

1528.8 ± 326

787.1 ± 130

110.2 ± 52

1797.7 ± 707

2267.8 ± 493

88.02 ± 41

δ13C

29.73 ± 8.5

0.59 ± 2.9

−11.2 ± 0.5

29.91 ± 13.5

29.59 ± 12.6

−10.88 ± 0.8

EL External labeling, IL intermediate labeling

Sample processing

In the laboratory, epiphytes were gently removed from the leaves with a razor blade. Leaves and rhizomes from each shoot were separated and their length and width measured. Leaf production was assessed according to Romero (1989) for all leaves within a shoot. Newly grown sections were cut out, measured, dried, and weighed for each leaf (mg day−1). Internal leaves and new growing primords (usually a single translucent leaf of 0.5–1 cm length) were dried and ground together with the original internal leaf to a fine powder in a Retsch mixer mill (MM 200) and the combined material used for isotopic determination.

δ13C and δ15N isotope values of samples were determined using an ANCA-NT (Europa Scientific, Crewe, UK) interfaced with a 20–20-isotope ratio mass spectrometer (Europa Scientific, Crewe, UK). Isotope ratios in the samples are calculated from linear calibration curves constructed with standard reference materials of known composition and a blank correction. The average difference in isotopic composition between samples and reference material is determined by the equation:
$$ [(R_{\text{sample}} - R_{\text{standard}} )/(R_{\text{standard}})] \times 1000 = \delta_{\text{sample - standard}} $$
where Rsample is 13C/12C (or 15N/14N) in the sample; Rstandard is 13C/12C (or 15N/14N) in the working reference gas (PBD carbonate standard for δ13C and N2 for δ15N) which is calibrated against an internal standard (Atropina, IAEA, and/or UGS), and δsample-standard is the difference in isotopic composition of the sample relative to that of the reference, expressed in ‰.
Delta values for leaves and rhizome were first converted to atom %15N and atom %13C using the following equations (Gonfiantini et al. 1995):
$$ \begin{gathered} {\text{Atom}}\,\% {}^{15}{\text{N}}_{i} = 100/[272/[1 + (\delta {}^{15}{\text{N}}_{{i/{\text{air}}}} /1000)] + 1] \hfill \\ {\text{Atom}}\,\% {}^{13}{\text{C}}_{i} = 100/[89/[1 + (\delta {}^{13}{\text{C}}_{{i/{\text{standard}}}} /1000)] + 1] \hfill \\ \end{gathered} $$

Atom % excess of δ13C and δ15N in internal leaves and primords was calculated by subtracting Atom % values in reference leaves and rhizomes and then, transforming them in mg of 15N and 13C for each clipping and labeling treatment. In addition, for each clipping treatment (i.e. unclipped, external and intermediate leaves clipped), we estimated total nutrient contribution that each shoot makes to the growth of internal leaves and primords by adding the mg of isotopic material (15N and 13C) supplied by intermediate and external leaves.

Data analyses

The effect of each clipping treatment (fixed factor, 3 levels: unclipped, external leaves clipped, and intermediate leaves clipped) on the mass of 15N and 13C supplied to the internal leaf and new emerging primords from each leaf position (fixed factor, 2 levels: external leaves marked and intermediate leaves marked) was investigated with a two-way factorial ANOVA. The effect of the clipping treatment (fixed factor, 3 levels) on biomass production (mg DW day−1) and N content of internal leaves and primords and on the total production per shoot was investigated with one-way ANOVA.

For all ANOVAs, data were first tested for homogeneity of variances (Cochran’s test) and normality (Kolmogorov–Smirnov distribution-fitting test of the residuals) and transformed when necessary to satisfy these assumptions. The existence of significantly different groupings was investigated with Student–Newman–Keuls post hoc comparisons.

Results

Patterns of 15N and 13C supply to growing tissues

Basal isotopic signatures (obtained from shoots far apart from those labeled) in the internal leaf and primords were 6.67 ± 1.5 ppt for δ15N and −11.72 ± 0.28 ppt for δ13C. Procedural controls (shoots collected near the labeled ones) had signatures similar to these basal values (5.72 ± 0.36 ppt for δ15N and −11.93 ± 0.20 ppt for δ13C) and were substantially lower than those of labeled plants (see Table 1), indicating that leakage from the plastic bag during incubation was negligible.

The amounts of 15N and 13C that were supplied to the internal leaf and primords varied significantly among the six experimental combinations of clipping and labeling (i.e. clipping × labeling interaction, Table 2a). The highest contributions were observed in treatments with the intermediate leaves clipped and the external leaves labeled and in controls with the external leaves labeled; and the lowest in treatments in which intermediate and external leaves had been simultaneously clipped and labeled (see Fig. 3). Both nutrients followed the same trend although the supply of nitrogen from external leaves was not significantly different from that of unclipped leaves whereas the supply of carbon was increased when intermediate leaves were clipped (Fig. 3a, b).
Table 2

ANOVA results for (a) 15N and 13C in the internal leaf and primords per clipping treatment (unclipped, external clipped and intermediate clipped) and leaf labeling (external and intermediate); (b) total (i.e. isotopic mass from external and intermediate leaves summed) 15N and 13C in the internal leaf and primords per clipping treatment; (c) nitrogen content in the internal leaf and primords per clipping treatment and; (d) leaf production of the internal leaves and primords and of the whole shoot per clipping treatment

ANOVA

(a)

Isotope mass per treatment

df

15N

13C

MS

F ratio

P

MS

F ratio

P

Clipping = C

2

0.00

2.94

0.062

0.001

0.999

0.37

Labeling = L

1

0.002

12.08

0.01

0.003

5.060

0.029

C × L

2

0.002

10.05

0.000

0.003

5.784

0.005

Error

53

0.0002

  

0.00052

  
  

Transformation:

Cochran C = 0.45

Transformation:

Cochran C = 0.53

(b)

Total isotope mass per clipping treatment

df

15N

13C

MS

F ratio

P

MS

F ratio

P

Clipping = C

2

7.695

4.045

0.029

18.02

3.52

0.044

Error

27

0.0038

  

0.0056

  
  

Transformation: √x

Cochran C = 0.43

Transformation: √x

Cochran C = 0.55

(c)

N content the internal leaf and primords

df

MS

F ratio

P

Clipping = C

2

0.11

0.73

0.48

Error

56

0.1573

  
  

Transformation:

Cochran C = 0.36

Leaf production

(d)

Internal leaf and primords

Shoot

df

MS

F ratio

P

MS

F ratio

P

Clipping = C

2

0.03

1.56

0.22

0.01

0.89

0.41

Error

56

0.0017

  

0.0056

  
  

Transformation: √x

Cochran C = 0.42

Transformation: √√x

Cochran C = 0.43

Significant P values are indicated in bold

https://static-content.springer.com/image/art%3A10.1007%2Fs00227-011-1664-0/MediaObjects/227_2011_1664_Fig3_HTML.gif
Fig. 3

a mg 15N and b mg 13C (mean ± SE), in the internal leaf and primords of P. oceanica for each clipping and labeling treatment. In SNK, significantly different groups are indicated with letters (n = 10). Detail as in Fig. 2

The effects of the clipping treatment alone (i.e. the supply of external and intermediate leaves together) showed that the capacity of intermediate and external leaves to contribute to internal leaves and primords was different (i.e. a significant clipping effect, Table 2b). The amount of nitrogen and carbon supplied to young leaves was higher in unclipped shoots (0.028 ± 0.007 mg of 15N and 0.030 ± 0.005 mg of 13C), than in shoots with external leaves clipped (0.010 ± 0.002 mg of 15N and 0.014 ± 0.004 mg of 13C) as increased supply from intermediate leaves (109% and 208% increase in N and C supply, respectively; see Fig. 3) could not compensate for the reduced supply from old leaves. In contrast, clipping of intermediate leaves did not change the overall amounts of isotope material supplied to newly growing leaves (0.028 ± 0.006 mg of 15N and 0.034 ± 0.015 mg of 13C), as they have modest contributions and the supply from external leaves increased by 16% of N and 36% of C (Fig. 3). Hence, modest increases in the nutrient supply from external leaves but not large increases in the supply from intermediate leaves are able to compensate nutrient contributions to the internal leaf and primords (Fig. 3). In fact, unclipped external leaves supplied ca. fivefold more 15N and 13C than unclipped intermediate ones, suggesting that old leaves might have a major role as a source of C and N of new growing leaves.

The reduced supply of 15N in shoots with external leaves clipped did not result, however, in any significant decline in the total nitrogen content in the internal leaf and primords (Table 2c) since similar values of nitrogen content were observed in all treatments (Unclipped: 2.26 ± 0.09%N; External Clipped: 2.26 ± 0.08%N; Intermediate clipped: 2.39 ± 0.09%N).

Leaf production

No significant effects of clipping were observed either in the biomass production of internal leaf and primords (Fig. 4; Table 2d) or in the total leaf biomass production (mg DW day−1 shoot−1) as the values of these parameters remained unaltered relative to control shoots (Fig. 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-011-1664-0/MediaObjects/227_2011_1664_Fig4_HTML.gif
Fig. 4

Primary production (mg DW shoot−1) of the internal leaf and primords (mean ± SE) and of the entire shoot (n = 10). U Unclipped, EC external clipped, IC intermediate clipped, N·S. not significant effect

Discussion

The seagrass P. oceanica is highly tolerant of herbivore damage, sustaining leaf growth and maintaining constant nutrient concentration in growing leaves when subjected to considerable tissue losses that mimic the action of either sea urchin or fish. Although tolerance to herbivory has been previously reported for this species in terms of leaf growth (Tomas et al. 2005; Vergés et al. 2008), here we show that plants are able to partially compensate for tissue losses by increasing nutrient supply from both intermediate and external leaves but that the importance of this process is largely specific to leaf age features. When the overall shoot supply could not be fully compensated, the total nutrient content of the new tissue grown and growth rates of young leaves were still unaltered during the experimental period (Fig. 5), thus indicating that simultaneous contribution from other sources also occurs (e.g. adjacent shoots or rhizomes).
https://static-content.springer.com/image/art%3A10.1007%2Fs00227-011-1664-0/MediaObjects/227_2011_1664_Fig5_HTML.gif
Fig. 5

Primary production (mg DW shoot−1) of the internal leaf and primords (mean ± SE) and of the entire shoot (n = 10). Detail as in Fig. 4. N.S. Not significant effect

Isotope tracing studies have long been conducted to assess the importance of resource sharing among adjacent seagrass shoots (e.g. Harrison 1978; Libes and Boudouresque 1987; Marbà et al. 2002; Prado et al. 2008b); however, to the best of our knowledge, never to address compensation following distinctive leaf damage by marine herbivores. Fish herbivory (simulated by clipping intermediate leaves) caused either equal (for N) or even slightly higher (for C) supply from external leaves to newly growing leaves than that of control plants. In contrast, when grazing by sea urchins occurs (simulated by clipping external leaves), supply from intermediate leaves was also stimulated, but was not effective enough to maintain the overall supply of C and N, which decreased by 62% and 50% (N and C, respectively). These patterns of nutrient supply can be explained by uneven basal nutrient contributions among seagrass leaves as the oldest; most external leaves are major sources of C and N whereas those of intermediate age are only minor contributors (83% vs. 17% of the C supply and 84% vs. 16% of the N supply from external and intermediate leaves, respectively). External, fully grown leaves may have similar rates of N uptake per unit weight compared with intermediate age leaves and appear as important organs providing a continuous supply of nutrients and photosynthates to growing tissues (Borum et al. 1989; Pedersen et al. 1997). External leaves also have slightly larger surface areas than intermediate leaves (ca. 10 cm2) and may achieve a higher incorporation of nutrients from the media. Yet, unclipped intermediate leaves provided 3.7 times less 15N and 4 times less 13C than expected by area differences alone. Therefore, leaf age features connected to the transition period from sink to source status (Chabot and Hicks 1982) and, to a lesser extent, higher rates of nutrient uptake induced by greater leaf area appear to influence the enhanced contribution of external leaves to the overall nutrient supply to growing tissues. In fact, P. oceanica features the longest leaf life span and the highest N resorption among seagrasses (ca. 40% of requirements; Alcoverro et al. 2000), which provides further indirect evidence that nutrient supply from external leaves is an important mechanism of compensation in this species.

The influence of leaf age features in physiological integration among plant parts and/or organs might also be affected by other structural and developmental factors such as plant anatomy and growth morphology as well as with specific adaptations to the physical environment. For instance, Honkanen et al. (1999) studied the effect of age-needle defoliation on Pinus sylvestris trees and found that the loss of 2-year-old needles of the branch leader shoot did not affect productivity whereas loss of 1-year-old needless reduced the mass and length of needles in the new shoots. In contrast, plants without true specialized organs and vascular connections, such as the alga Fucus vesiculosus, appear to lack any ability to respond to any type of tissue losses (Honkanen and Jormalainen 2002) and may persist through herbivore grazing by adopting faster growth strategies compared with higher plants. In seagrasses, previous tracing studies suggest that sharing of nutrients among shoots along the main rhizome is comparatively less important (up to 5 times lower) than within-shoot reallocation from external leaves (see Prado et al. 2008b) possibly, as a result of the large contribution of leaves to nutrient acquisition (up to 60–70% of their total N uptake; Hemminga et al. 1991). In our experiment, external leaves were only able to increase nutrient supply to a small amount under plant damage (16% for 15N and 36% for 13C when intermediate leaves were clipped) but this was enough to fully compensate the loss of intermediate leaves. In contrast, supply from intermediate seagrass leaves was unable to compensate for the loss of external leaves, despite an increase in the supply of over 100% 15N and over 200% 13C, possibly because they are not yet fully developed and have more active growth (ca. 11–17% of total shoot production). Valentine et al. (2004) also found that clipping of external leaves of Thalassia testudinum increased the level of activity of nitrogen-metabolizing enzymes, thus suggesting that they are fully active and may be important organs sustaining growing tissues.

At first, given the absence of negative effects in the nutrient budget when intermediate leaves were clipped, and the decrease in the amounts of nutrients supplied when external leaves were removed, it might appear that continuous sea urchin grazing could affect the plant nutrient budget much more than fish grazing. Yet, this assertion has several counterarguments. First, losing intermediate leaves in summer could have negative consequences on mid-term processes. In effect, after a few months, intermediate leaves become external and, if part of their biomass has been lost, their role as nutrient source could be curtailed. Second, and probably more importantly, the seagrass shoot-specific growth does not seem to be depressed by this tissue loss, or at least not in the short to mid-term (Vergés et al. 2008; this study), which confirms the capacity of P. oceanica for compensation and suggests that other sources of nitrogen income (leaf uptake, reallocation from rhizomes, or adjacent ramets) might also be enhanced. For instance, Vergés et al. (2008) conducted a clipping experiment in which treatments were applied independently of leaf age and found that reallocation of nutrients stored in belowground rhizomes is an important mechanism of compensation, at least in P. oceanica. Shoots may also boost their nutrient content by increasing leaf nutrient uptake, an important mechanism of nutrient acquisition in seagrasses, even at low nutrient concentrations (Pedersen et al. 1997; Lepoint et al. 2002; Romero et al. 2007). During the experiment, we measured the nutrient supply to internal, newly growing leaves, and therefore, the possible effects of clipping on nutrient uptake cannot be fully separated from that of transport to new growing leaves. Yet, the large differences in nutrient supply capacity between unclipped intermediate leaves and expected amounts for external leaves with the same area suggest that compensatory uptake was not the main responsible mechanism.

Little is still known about mechanisms driving distinctive leaf damage by macrograzers. A first hypothesis would be a simple matter of accessibility, with more intermediate, upright leaves being more reachable to fish while older, external leaning ones more reachable to sea urchins. However, very recent work has shown that there is also a preferential consumption, in part mediated by distinctive epiphyte communities (Prado et al. 2010; Vergés et al., submitted). Yet, the lack of other studies reporting leaf age-dependent herbivory and mechanisms of tolerance to different types of leaf damage by seagrass consumers makes it difficult to generalize our results. Traits causing herbivore preference and conferring tolerance to herbivory may vary among different groups of plants (Obeso 1993) and among types of herbivores (Kotanen and Rosehental 2000). Seagrasses, however, gather a number of features that have been associated with tolerance to both vertebrate (e.g. well-developed clonality, stored reserves, presence of dormant and basal meristems) and invertebrate herbivory (e.g. flexible branching and post-damage resource allocation, Kotanen and Rosehental 2000). Such multiplicity of traits may stem from seagrass evolution in systems with a rich diversity of grazers (e.g. Domning 1981; Ivany et al. 1990; De Muizon et al. 2004; Uhen 2007). In this case, compensation to herbivory, mostly by increased nutrient supply from external leaves, also supports this view and suggests that it may be a broad-scale mechanism to tolerate recurrent leaf damage by common macroherbivores such as fish and sea urchins.

Acknowledgments

This work was supported by a FI scholarship from the Departament d’Universitats, Recerca i Societat de la Informació (DURSI, Generalitat de Catalunya) to P. Prado, and the CGL2007-66771-C02 and CGL2009-12562 grants from the Spanish Ministry of Science and Innovation. C. Collier was supported by a short-term outgoing scholarship from the Australian Strategic Research Fund for the Marine Environment and a travel grant from the Centre for Ecosystem Management, Edith Cowan University, Perth, Western Australia.

Copyright information

© Springer-Verlag 2011