A Review of the Role of the Green Turtle in Tropical Seagrass Ecosystems

Lemnuel V. Aragones

Abstract

This chapter reviews the role of the green turtle Chelonia mydas in tropical seagrass ecosystems by considering the results of investigations into the effects of herbivory on community structure, productivity, recovery and nutritional attributes of tropical sea- grasses. These results are based on simulated turtle cropping experiments held in the seagrass beds in Queensland, Australia. The models upon which these simulation experiments were based on the 'grazing plots' observed in the Caribbean.

Green turtles, by cropping play a role in community structure, dynamics, and nutritional attributes of tropical seagrass communities. The nature of the cropping effects depends on species composition, season, location, and abundance of other large marine herbivores and other forms of disturbance. Cropping produces mosaics of seagrass patches that have younger leaves with a higher nutrient contents and better nutritional quality on a local scale.

As a result of changes in the community structure, dynamics, and nutritional attributes, the function of these habitats also changes, and seagrass communities cropped by green turtles become preferred feeding areas. A major, long-term reduction in the number of green turtles may lead to irreversible degradation of these preferred feeding areas, as younger leafs with better nutritional quality are replaced by older and lower nutritional quality seagrass patches.

In addition to sea turtles, other forms of natural disturbance and environmental constraints also affect seagrass ecosystems, and require consideration in conservation programs.

Introduction

Green turtles, together with sirenians such as the dugong and the manatee, are marine megaherbivores (Aragones 1996); i.e. they have a large body size, which may weigh from 200 to more than 1000 kg, and which requires a considerable amount of food in the form of seagrasses (Bjorndal 1982, 1985, Thayer et al. 1984, Preen 1995, Aragones 1996).

Among the most significant contributions in the study of green turtles with regard to their ecological role in foraging grounds are studies on the feeding and nutritional ecology of the green turtle by Bjorndal (1980, 1985) and Mortimer (1981, 1982). Aspects of the green turtle-seagrass interactions have also been studied (Bjorndal 1980, 1982, Mortimer 1981, Ziemen et al. 1984, Williams 1988, Kuiper-Linley 1994). However, these studies were not able to elucidate the complete role of green turtles in seagrass ecosystems. Only Thayer et al. (1984) presented a general review of the role of large herbivores, which included the green turtle, in seagrass communities, but the cropping effects of the green turtle have not been comprehensively examined. To further our understanding of the role of green turtles in the tropical seagrass ecosystem, it is necessary to study their effects on seagrass community structure, productivity, recovery, and nutritional attributes. This chapter reviews the role of the green turtle in tropical seagrass ecosystems, by considering the effects of green turtle cropping on the factors mentioned above, and by posing the question: Will a major and long-term reduction in the numbers of green turtles in an area lead to an irreversible degradation of their habitats?

Materials and Methods

This review is based primarily on data resulting from a study which examined the ecological roles of the large marine vertebrates, such as the dugong and the green turtle, in tropical seagrass ecosystems through simulated herbivory experiments (Aragones 1996, Aragones & Marsh 1999, Aragones et al. ms). The simulated herbivory experiments used open plots (without exclosures) and four treatments: intensive turtle cropping, intensive dugong grazing, light dugong grazing and controls. These experiments demonstrated that these two large marine herbivores have specific and additive herbivory effects, and it would have been ideal to discuss them together. However, here, I will consider only the role of the green turtle based on these empirical data. In some instances in this review I refer to both of them together to emphasise certain points.

Impacts of cropping on seagrass community structure

Green turtle cropping was found to considerably alter the seagrass community structure, and in particularly the leaf age structure. Bjorndal (1980) discovered the existence of 'grazing plots' in the Caribbean. These plots were a result of individual green turtle's cutting of seagrass blades (of Thalassia testudinum) in the basal portion, and subsequent regular recropping of the same seagrass patches. This resulted in a patch of young leaves with a higher nutritional content. In a more recent study, Aragones & Marsh (1999) demonstrated that the aboveground biomass of seagrass plots with intensive green turtle cropping were significantly different from uncropped plots after one to two months (Fig. 1 and Table I). Furthermore, we demonstrated that the biomass harvested from plots cropped after one month, in terms of leaves, roots/rhizomes, whole plants and the ratio of roots and rhizomes to leaves was considerably different from uncropped plots. The leaf biomass was the only significantly different component from uncropped plots after two months (see Figs. 1 & 2, Table I). These results also indicated that the leaves, roots/rhizomes and whole plants in the Location x Treatment interaction were also significantly different after a month of cropping. However, the ratio of the roots and rhizomes to leaves was the only significant factor in the Location x Treatment interaction with cropping after two months (see Table I). These notable differences, apart from providing evidence of the significance of green turtle cropping to structuring tropical seagrass communities, also serve as important cues for the green turtles to detect or relocate their preferred sites, or 'grazing plots'. These results may possibly exist across various sites, i.e. not only in the experimental sites, but also in the rest of the tropical seagrass communities. However, because of the declining numbers of green turtles, and given that green turtles not only eat seagrasses but also algae (Bjorndal 1980, Bjorndal 1985), these effects are often limited at a local scale.

It is important to emphasise that green turtle cropping alone may not be sufficient to change the species composition of tropical seagrass communities. However, along with other effective marine megaherbivores such as the dugong, their collective efforts and accumulative effects may be able to change the species composition (see Aragones 1996, Aragones & Marsh1999).

Table I: Results of the ANO VAs for a short-term (1-2 months) cropping experiment in a monospecific meadow at Cardwell based on repeated measures of aboveground biomass of H. uninervis using video. Significant P values are in bold (adapted from Aragones & Marsh 1999)

¹ with Greenhouse-Geiser correction for correlation through time in repeated measures
² all factors fixed except location, which was random
³ tested against within subjects residual
⁴ tested against within location x treatment
⁵ transformed log (ratio+1)
⁶ tested against within location x species x treatment
⁷ index of net aboveground biomass production

Impacts of green turtle cropping on the recovery of tropical seagrass communities

Recovery from green turtle cropping is often relatively rapid. The recovery of the leaf biomass of seagrass under intensive green turtle cropping varies according to species (Figs 1 & 2; see Aragones 1996, Aragones and Marsh 1999). Recovery is referred to as when the leaf biomass of the experimental (treatment) plots is similar to the control plots. Halophila ovalis, an opportunistic, fragile and fast growth species (Hillman & McComb 1988), had the most rapid recovery from cropping (Fig 1). Zostera capricorni and Cymodocea sp., more robust and slow growth species (Kirkman et al. 1982, Brouns 1987), recovered after two months. For Halodule uninervis, an intermediate species between H. ovalis and Zostera / Cymodocea, recovery was after three months. Considering that these experiments were held during the autumn-spring season in Australia, which is the period of slow growth, it is likely that seagrass recovery from cropping would be even faster during other times of the year (Aragones 1996, Aragones & Marsh 1999).

The rate of recovery from turtle cropping is dependent on seagrass species composition and the timing of the cropping, which could be critical in maintaining sea turtles' favorite grazing plots. During certain seasons, they must recrop more often during the period of fast growth, before the grazing plots recover and become undetectable. However, they may not go back to recrop fast growing and opportunistic species such as H. ovalis, as this species is relatively superior in terms of nutritional contents than the other more robust, slow growing species (Aragones 1996). It is possible that remembering the locations of patches of higher nutritional quality may be more important for green turtles than regular recropping of poorer quality areas. In Australian waters, juvenile green turtles seem to prefer to feed on H. ovalis and H. uninervis (Read 1991, Brand 1995), supporting this notion.

Impacts of green turtle cropping on the productivity of tropical sea- grass communities

Although green turtle cropping does influence seagrass productivity, only limited studies (Williams 1988, Aragones 1996, Aragones & Marsh 1999) have directly related green turtle cropping to productivity. Williams (1988) found that intensive green turtle cropping at St. John, Virgin Islands, low- ered the leaf productivity of the seagrass beds of T. testudinum. This was contradictory to that reported by Aragones (1996) and Aragones & Marsh (1999), where there was an increase in the index of aboveground biomass production of H. ovalis, Zostera / Cymodocea and H. uninervis in Queensland after intensive cropping (Table I and Fig. 1). This difference was apparently due to the fact that Queensland seagrass plots with intensive green turtle cropping were not recropped again, unlike in the case of in St. John, and for other reasons inherent to the Caribbean region (see below). This is evident from the higher leaf productivity in the beds in enclosures (not re- cropped by the turtles) than those cropped regularly (see Williams 1988). The increase in seagrass leaf production is an indication of a compensatory growth after cropping. Compensatory growth from cropping and other similar disturbances, such as clipping of leaves, has been extensively reported (e.g. McNaughton 1979, Hilbert et al. 1981, Dyer et al. 1982, Jefferies 1988, Kuiper-Linley 1994, Aragones 1996).

 
Fig. 1: The response of the multi-species seagrass meadow at Ellie Point t treatments in the longer-term grazing experiments; (a-d) temporal response (g dw/m²) monitored using video of the aboveground biomass of (a) sea- grass community, (b) Halophila ovalis (Ho), (c) Zostera / Cymodocea (Zc+Cr), (d) Halodule uninervis (Hu); (e-h) response (g dw/m²) at harvest after 10 months for (e) leaf biomass, (f) ratio of root/rhizome to leaf biomass, (g) detrital matter, (h) net aboveground biomass production. Bars represent the Least Significant Difference (LSD) between means showing differences among: (a-c) 1. Treatments within time and 2. Times within treatment, (g) treatments. May (1) = before and (2) = after, simulation measurements were carried out (from Aragones & Marsh 1999, with permission).

Another factor related to the dynamics of seagrass productivity and which has been influenced by green turtle cropping, is the amount of detrital matter produced. Thayer et al. (1982) presented evidence that green turtle cropping short-circuited the detrital cycle. This was because green turtles consumed the seagrass leaves, reducing the potential biomass for decomposition and in turn the decomposition time of the remaining matter. This was opposite to the findings in Queensland (see Fig. 1 and see Aragones 1996 for more details), where plots were not recropped, resulting in trends of leaf biomass accumulation due to compensatory growth. This growth could subsequently increase the detrital matter, especially if left untouched (un- cropped) after 10 months. It is presumed that this is the scenario in tropical seagrass beds where drastic reductions in the numbers of green turtles have occurred. However, if plots were recropped at least once a month or even once every two months, depending on the species composition of the sea- grass beds, there may have been a reduction in the detrital matter in the same 10 months.

Impacts of green turtle cropping on the nutritional attributes of tropical seagrasses

The nutritional attributes of seagrasses improve after green turtle cropping (see Tables II & III)(Bjorndal 1980, Kuiper-Linley 1994, Aragones 1996, Aragones et al. ms). Nutritional attributes include nutrient contents (e.g. nitrogen, water-soluble carbohydrates) and nutritional quality (in vitro digestibility). Most studies that examined the effects of cropping on nutritional aspects of seagrasses have looked only at a handful of nutrient components, such as nitrogen and organic matter. Only Bjorndal (1980) indirectly measured the nutritional quality of seagrasses after cropping, in addition to N, OM, neutral detergent fibre, acid detergent fibre, and acid lignin measurements. Bjorndal calculated the apparent digestibility coefficient (ADC) using the standard lignin ratio technique. More recently, we attempted to measure nutritional quality more directly by using in vitro dry matter digestibility (IVDMD) together with nine other nutrient components (i.e. N, OM, WSC, ADF, NDF, acid lignin, and total starch). In vitro dry matter digestibility measures the proportion of total dry matter that disappears from the initial sample after incubation with pepsin, HC1 and fungal cellulase (following Choo et al. 1982 and Van Soest 1994, for more details see Aragones 1996). Measuring IVDMD is an appealing alternative technique to feeding experiments in order to measure an animal's performance on a particular food plant because in vivo techniques are troublesome to perform, especially since most wild herbivores are extremely difficult to keep in captive conditions (Aragones 1996). 

Figure 2. The response of the monospecific seagrass meadow at Cardwell, Queensland to treatments in the longer- and shorter-term herbivory experiments (from Aragones 1996, Aragones and Marsh 1999). (a) temporal response (g dry weight/m²) monitored using video of the aboveground biomass of Halodule uninervis during the longer-term experiment, (b) leaf biomass (g dw/m²) removed naturally by dugong over time during the longer-term experiment (note. December is not shown in this figure because there was no monitoring for this month due to inaccessibility of experimental units); (c-d) response (g dw/m²) at harvest after 13 months for (c) seagrass plant parts, (d) net aboveground biomass production; (e) temporal response (g dw/m²) monitored using video of the aboveground biomass of H. uninervis during the shorter-term experiments (1-4 months); (f) response (g-dw/m²) at the harvest after 1-4 months exposure to treatments in the short-term experiments. Bars represent the Least Significant Difference (LSD) between means showing differences among: (a) 1. treatments within time and 2. time within treatments; (c) treatments for 1. Leaves, 2. roots/rhizomes, 3. whole plants, 4. ratio (corresponds with right axis); (d) treatments; (e) time within treatments of 1. light grazing/control, 2. 1 month after cropping/control, 3. 2 months after cropping/control. The LSDs are not necessary for (f) as comparisons are summarized in Table 1. June and August (1) = before and (2) after simulation measurements were carried out. For more discussion see Aragones & Marsh 1999.

The results from cropping experiments on nutritional attributes are summarised in Table II. The results indicate that the first two months appear to be the best time to recrop if nutritional benefits are to be maximised (Aragones 1996, Aragones et al. ms). The leaves were significantly more digestible in the first month of recovery after recropping (see Table II), while the concentrations of nitrogen and organic matter in the leaves in the second month remained significantly higher (see Table II). A significantly lower concentration of acid detergent fibre (ADF) is associated with younger plant tissues (Van Soest 1994). These results clearly provide a biochemical basis for the maintenance of 'grazing plots' by green turtles in the Caribbean, and show how seagrasses cope with such herbivory (see Aragones 1996, Aragones et al. ms) or elucidate the internal mechanism for compensatory compensatory growth for seagrasses. All of these validate the suggestions by Bjorndal (1980) and concur in all the aspects of turtle cropping on community structure and dynamics discussed above.

Table II: Summary of the proportional (%) changes (+ increased or - decreased relative to the control) in the nutritional composition of H. uninervis significantly (P < 0.05) affected by simulated cropping (in parentheses)(from Aragones 1996, Aragones et al. ms).

C¹ = Intensive turtle cropping harvested after 10 months;
C² = Turtle cropping harvested after 1 month
C³ = Turtle cropping harvested after 2 months

Green turtle cropping improves seagrass nutrients and nutritional quality, which should be beneficial not only for these large marine herbivores but also to other herbivores within the seagrass ecosystem. This nutrient improvement after cropping is further supported by Table III, which summarises the increases in the concentrations of the various nutritional components (i.e. N, OM, WSC, and starch) of seagrass leaves and roots/rhizomes after cropping. Nitrogen is considered a major plant nutrient because it is essential for protein synthesis for herbivores (Mattson 1980, Crawley 1983), while starch, from the plant perspective, is regarded as a reserve carbohydrate and often classified with soluble carbohydrates because of its partial solubility to hot water (Van Soest 1994)

Table III: Effects of cropping on some nutritional components of seagrass from Aragones 1996 and others. This was limited only to studies whose laboratory techniques for the nutritional analyses were reasonably comparable. For more details see Aragones 1996.

* data taken at several different seasons;
¹long-term experiments; 
²shortterm experiments 
³winter sampling;
⁴spring sampling; 
⁵fall sampling

The role of the green turtle

The impacts of green turtle herbivory on the structure and dynamics of tropical seagrass communities appear to be the most important contribution of these animals in the tropical seagrass ecosystem. The cropping effects discussed above serve as a means of disturbance and benefit the herbivore and localised patches of seagrasses (except probably in some areas, when intensively grazed, see discussion below). The green turtles appear to be an important component in tropical seagrass ecosystems in structuring communities through cropping. The resulting changes in the structure of the tropical seagrass communities have parallel implications in the functions of these seagrass beds. As a result of cropping impacts, some seagrass beds are preferred over others without major herbivores or other sources of periodic disturbance (Aragones 1996). In areas where green turtle populations have declined or are chronically declining, cropping effects may become insignificant, and green turtles in these areas would be considered 'ecologically extinct' (after Estes et al. 1989), since they can no longer perform their ecological role.

The nature and extent of cropping effects are related to the nature of the seagrass community, including species composition and location. For example, in the Caribbean, extensive cropping in the same area, like 'grazing plots' led to ammonium reduction resulting in lower net primary productivity (see Ziemen et al. 1984), whereas in the Australasian region this may not be the case. Seagrass beds in the northeastern Australia and the rest of the Indo-Pacific are more diverse than those in the Caribbean, and it is probably a better strategy for the green turtles to sample a variety of food plants rather than maintain plots which are repeatedly cropped. This may be why patches resembling Caribbean 'grazing plots' have never been reported in the Australasian region. Secondly, most of the seagrass biomass in the Australasian region comprises pioneer or opportunistic and fast-growth species, such as Halophila and Halodule, rather than the robust and long-lived species such as Thalassia hemprichii and Enhalus acoroides (Coles et al. 1989; Lee Long et al. 1993). In addition, disturbance generated by dugong grazing in Australia (Preen 1995, Aragones 1996, Aragones & Marsh 1999) enhances the microbial activities in the sediment nutrients (Perry 1998) which enables the remaining plants to recover rapidly and increase productivity (Aragones 1996, Aragones et al. ms). This occurs as dugong grazing uproots whole seagrasses resulting in long serpentine feeding trails 3-5cm deep (Preen 1995), which enhances sediment mineralisation (Perry 1998). In the Caribbean the density of manatees is low (Marsh and Lefebvre 1995) and they do not generate similar effects.

Conclusions

The role of green turtles in structuring and modifying the dynamics and nutritional attributes of tropical seagrasses is important particularly in areas where there are no other sources of periodic natural disturbances on seagrass communities such as typhoons and cyclones, and other megaherbivores such as sirenians (dugongs and manatees). Green turtle cropping influences the community structure and dynamics of tropical seagrass communities in terms of recovery, productivity and detrital production, and determines if a particular seagrass community will have a detrital or herbivory-based cycle.

The influence of cropping on the nutritional content and quality of tropical seagrasses depends on species and abundance of other large marine herbivores such as the sirenians. In other areas, other forms of natural disturbance and environmental constraints maintain and structure seagrass communities.

There have been limited studies which integrated the combined effects of cropping on community structure, dynamics and nutritional attributes (Bjorndal 1980, 1985, Mortimer 1982, Aragones 1996).

Will a major and long-term reduction in the numbers of green turtles in an area lead to an irreversible degradation of their habitats? It is likely to be locality dependent.

Acknowledgements

I am grateful and thankful to the following organisations and individuals: AUSAID for my PhD scholarship and Helene Marsh for supporting me during my studies. Bill Foley and Tony Preen for sharing their knowledge on the ecology of dugong, green turtle, and nutritional ecology. Glen De'ath and Steve Delean for statistical support, Karen Bjorndal and Jeanne Mortimer for providing the initial researches on green turtle feeding ecology and being appreciative of my experimental approach. Nicolas Pilcher for the encouragement to write this review, and the organising committee of the  2^nd ASEAN Symposium and Workshop on Sea Turtle Biology and Conservation for travel assistance to enable me to attend and present the core of this review.

References

Aragones, LV., 1996. Dugongs and green turtles: grazers in the tropical seagrass ecosystem. PhD thesis, James Cook University, Townsville: 292pp.

Aragones, L.V. & H. Marsh, 1999. Impact of dugong grazing and green turtle cropping on tropical seagrass communities. Pacific Conservation Biology 3.

Aragones, LV., W. Foley, I. Lawler & H. Marsh, ms. Impact of dugong grazing and turtle cropping on the nutritional attributes of tropical seagrass communities. Unpublished report

Bjorndal, K.A., 1980. 'Nutrition and grazing behaviour of the green turtle, Chelonia mydas. Marine Biology 56: 147-154.

Bjorndal, K.A., 1982. The consequence of herbivory for the life history pattern of the Caribbean green turtle, Chelonia mydas. In: Biology and conservation of sea turtles. (Bjorndal, K.A. ed): 111-116.

Bjorndal, K.A., 1985. Nutritional ecology of sea turtles. Copeia 3: 736-751. 

Brand, S.J., 1995. Diet selection by juvenile green turtles, Chelonia mydas, in subtropical flathead gutter, Moreton Bay. Honours thesis, University of Queensland, Brisbane: 84pp.

Brouns, J.J.W.M., 1987. Aspects of production and biomass of four seagrass species (Cymodoceoideae) from Papua New Guinea. Aquatic Botany 27: 333-362.

Choo, G.M., PG. Waterman, D.B Doyle & J.S.Gartlan, 1982. A simple enzyme assay for dry matter digestibility and its value in studying food selection by generalist herbivores. Oeceologia 49: 170-178.

Coles, R.G., I.R. Poiner & H. Kirkman, 1989. Regional studies- seagrasses of northeastern Queensland. In: Biology of seagrasses. (Larkum, A., A. McComb, & S. Shepherd, eds): 261-278.

Crawley, M.J., 1983. Herbivory: The dynamics of animal-plant interactions. University of California Press: 435pp.

Dawes, C.J. & J.M. Lawrence, 1979. Effects of blade removal on the proximate composition of the rhizome of the seagrass Thalassia testudinum. Aquatic Botany 7: 256-267.

Dawes, C.J., K. Bird, M. Durako, R. Goddard, W. Hoffman & R. McIntosh, 1979. Chemical fluctuations due to seasonal and cropping effects on an algal-seagrass community. Aquatic Botany 6: 79-86.

Dyer, M.I., J.K. Detling, D.C. Coleman & D.W. Hilbert, 1982. The role of herbivores in grasslands. In: Grasses and grasslands. (Estes, J.R., R.J. Tyrl, & J.N. Brunken, eds): 255-295.

Estes, J.A., Duggins, D.O. & Rathbun, GB., 1989. The ecology of extinctions in kelp forest communities. Conservation Biology 3(3): 252-264.

Hilbert, D.W., D.M. Swift, J.K. Detling & Dyer, M.I., 1981. Relative growth rates and grazing optiinisation hypothesis. Oecologia 51:14-18.

Hillman, K. & A.J. McComb, 1988. The primary production of the seagrass Halophila ovalis in the Swan/Canning estuary, Western Australia. II. Seasonal changes in productivity and morphology. In: Biology of seagrasses (Larkum, A., A. McComb, & S. Shepherd, eds): 635-685.

Jefferies, R.L., 1988. Vegetational mosaics, plant-animal interactions and resources for plant growth. In: Plant evolutionary biology (Gottlieb, L.D. & S.K. Jain, eds): 341-369.

Kuiper-Linley, M., 1994. Effects of simulated turtle grazing on seagrasses of Moreton Bay. Honours thesis, University of Queensland, Brisbane: 56 p.

Larkum, A.W.D., L.C. Collet & R.J. Williams, 1984. The standing crop, growth and shoot production of Zostera capricorni Achers in Botany Bay, New South Wales, Australia. Aquatic Botany 19: 307-327.

Lee Long, W., J.E. Mellors & R.G. Coles, 1993. Seagrasses between Cape York and Hervey Bays, Queensland, Australia. Australian J. Mar. Freshwater Res. 44 33-42.

Mattson, W.J. Jr., 1980. Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics 11: 119-161.

Marsh, H. & LW. Lefebvre, 1995. Sirenian status and conservation efforts. Aquatic mammals 20: 155-170.

McNaughton, S.J., 1979. Grazing as an optimisation process: grass-ungulate relationships in the Serengeti. American Naturalist 113(5): 69 1-703.

Mortimer, J.A., 1982. Feeding ecology of sea turtles. In: Biology and conservation of sea turtles (Bjorndal, K.A. ed): 103-109.

Mortimer, J.A., 1981. The feeding ecology of the West Caribbean green turtle (Chelonia mydas) in Nicaragua. Biotropica 13: 49-58.

Perry, C.J., 1998. Sediment microbial interactions. PhD thesis, University of Queensland, Brisbane: 200pp.

Preen, A., 1995. Impacts of dugong foraging on seagrass habitats: observational and experimental evidence for cultivation grazing. Marine Ecological Progress Series 124: 201-213.

Read, MA., 1991. Observations on the feeding ecology of immature green turtles, Chelonia mydas, in the Moreton Banks region of Moreton bay, south-east Queensland. Honours thesis, University of Queensland, Brisbane: 68 p.

Thayer, G.W., K.A. Bjorndal, J.C. Ogden, S.L. Williams & J.C. Zieinen, 1984. Role of larger herbivores in seagrass communities. Estuaries 7(4): 351-376.

Thayer, G.W., D.W. Engel & K.A. Bjorndal, 1982. Evidence for short-circuiting of the detrital cycle of seagr