Читать книгу Tropical Marine Ecology - Daniel M. Alongi - Страница 34

The shores of many tropical estuaries are inhabited by mangrove forests. Their presence results in unique circulation patterns, which lead to distinct chemical and biophysical characteristics that are quite different from those in temperate estuaries (Mazda et al. 2007). Water flows in mangrove forests are strongly influenced by the presence of the trees and their aboveground roots as well as by the geomorphology of the tidal creeks. These unique characteristics have important ecological consequences. Tidal and wave energy in any estuary constitutes an auxiliary energy subsidy as tides allow mangrove forests to store and pass on new fixed carbon and benefits animals adapted to make use of subsidised energy. Tides thus do the work of bringing nutrients, food, and sediments to mangroves and their food webs as well as exporting their waste products. This subsidy is an advantage in that organisms do not have to expend energy on these processes and can shunt more energy into growth and reproduction.

A few tropical estuaries are driven by macro‐tides, although these are the exception rather than the rule as most tropical estuaries are micro‐ or meso‐tidal. Along the Brazilian coast south of the Amazon, there are four different types of macro‐tidal estuary: (i) ‘typical’ macro‐tidal, (ii) estuaries with large fluvial discharge, (iii) shallow, frictionally dominated macro‐tidal estuaries, and (iv) estuaries with structural control (Asp et al. 2013). In the ‘typical’ macro‐tidal estuary, ebb tides are longer than flood tides, and this condition is like that found in macro‐tidal estuaries in northern Australia (Wolanski et al. 2006). In the Daly estuary in northern Australia, freshwater becomes dominant up to the mouth and tides can be suppressed during the wet season (Wolanski et al. 2006). The Daly estuary is a ‘leaky’ sediment trap with its trapping efficiency varying with season and between years. In contrast, Darwin Harbour, a wider and much larger macro‐tidal estuary NE of the Daly, is poorly flushed, especially in the dry season; much sediment remains trapped on intertidal flats and in mangroves with little loss to the sea (Wolanski et al. 2006). Within the harbour, complex bathymetry results in the generation of jets, eddies, and stagnation zones that can trap sediments inshore. There may be a feedback between tidal circulation and bathymetry as tidally averaged circulation appears to control the formation and movement of sand banks.

In micro‐tidal estuaries, circulation is similarly complex and greatly influenced by the presence or absence of discharging rivers and the width of the connection between the estuary and the adjacent coastal ocean. Along the western Gulf of Mexico, tropical micro‐tidal estuaries share many characteristics, including a narrow connection between the estuary and the adjacent continental shelf (Salas‐Monreal et al. 2020). In the Jamapa River estuary, for example, surface horizontal displacements of the salinity and temperature fronts during the dry season occur, while during the wet season, the salinity and temperature gradients are observed in the vertical at about 1 m depth. A cyclonic recirculation at the mouth of the estuary occurs when the ratio between the mouth and the estuary width is below 0.4. This should hold true for all tropical micro‐tidal estuaries in the western Gulf of Mexico (Salas‐Monreal et al. 2020).

Not all tropical estuaries are driven solely by tides year round. In some northern Australian estuaries, a salinity maximum zone develops during in the dry season that is driven by high rates of evaporation (Wolanski 1986). This zone occurs near the mouth of each estuary where downwelling occurs, and a classical and an inverse estuarine circulation prevails upstream and downstream of the salinity maximum. This zone acts as a ‘high salinity plug’ inhibiting the mixing of estuarine and open ocean water to the extent that, in some cases, freshwater does not leave the estuary. Similar conditions have been found in estuaries along the SW coast of Ghana (Dzakpasu and Yankson 2015), in the Konkouré estuary of Guinea (Capo et al. 2009) and in the Gulf of Fonseca estuary on the Pacific coast of Central America (Valle‐Levinson and Bosley 2003).

Some tropical estuaries are so complex as to defy simple classification. Good examples of such complexity can be found along the north coast of Brazil (Medeiros et al. 2001; Schettini et al. 2013). These estuaries have multiple riverine systems feeding into a larger lagoon which is ordinarily fronted by coral reefs or coral‐fringed barrier islands. In the Itamaracá estuarine system, there are several estuarine waterways than feed into an ‘inner sea’ via a series of inlets, each considerably different from the other (Medeiros et al. 2001). Most of the freshwater enters the northern branch of the Santa Cruz Channel through the Catuama, Carrapicho, do Congo, Arataca, Botafogo, and Igarassu Rivers, the last three being the main source of freshwater. During the dry season, hypersaline conditions exist at both entrances in the “inner sea” due to evaporation, evapotranspiration by mangroves, and reduced exchange between the channels and reef shelf waters; a series of coral reefs fringe the outer edge of the estuary.

Tidal circulation within most mangrove waterways is characterised by a pronounced asymmetry between ebb and flood tides, with the ebb tide being shorter, but with stronger current velocity than the flood tide (Cavalcante et al. 2013). Current velocities in tidal creeks can often exceed 1 m s⁻¹ but only rarely approach 0.1 m s⁻¹ within the forest proper (Cavalcante et al. 2013). This asymmetry results in self‐scouring of the tidal waterways to the extent that the bottom of most channels is composed of bedrock, gravel, and sand, with little or no accumulation of fine sediment. The ecological implication of this asymmetry is that there tends to be export of particulate nutrients, rather than net import.

The velocity of tidal circulation ultimately depends on the geometry of the waterway, that is, the ratio of the forest area to the waterway area as well as the slope of the forest. The ratio appears to be on the order to 2–10 (although few such measurements have been made) with a very small forest slope (Wolanski 2007). The tidal prism of a mangrove estuary thus increases greatly with an increase in the ratio between forest area to waterway area.

Numerical modelling has determined the importance of the interaction between tidal creek geometry and mangrove forests in causing asymmetry of tides. The dominance of the ebb tide is due to friction in the mangrove forest which is in turn controlled by the density of the forest (Mazda et al. 1995). Inside the forest, the water level and the current velocity are strongly controlled by drag force due to the vegetation (Norris et al. 2019). The denser the forest, the greater the drag resulting in slower current velocity and greater tidal asymmetry in the creek. This relationship, however, is not straightforward. The peak velocity in the waterway decreases at flood tide and increases at ebb tide for increasing levels of drag force. But when the drag force is excessive, the ebb flow is reduced allowing the waterway to silt. There is a natural feedback relationship among the vegetation, water, and sediments. This phenomenon is unique to mangrove estuaries.

An additional asymmetry of the currents in mangroves is the direction of the currents in relation to forest position (Li et al. 2012). At rising tide, the currents flow perpendicular into the forest, while at falling tide they are oriented at an angle, typically 30–60° to the bank. This lengthens the pathways of water at falling tide reducing the chance that materials such as mangrove detritus can escape the forest.

Another characteristic feature of mangrove hydrodynamics is the mixing and lateral trapping of water (Wolanski 2007; Mazda and Wolanski 2009). Lateral trapping of water within the forest is a dominant process controlling longitudinal mixing in mangrove waterways. The trapping phenomenon occurs when some of the water flowing in and out of the estuary is temporarily retained in the mangrove forest to be returned to the main water channel later. Trapping of water is enhanced in the dry season when there is little, if any, freshwater to cause buoyancy effects on water circulation. In the wet season, the buoyancy effect is important as freshwater is trapped in the forest at high tide, and as a floating lens or boundary layer hugging the riverbank at low tide. This effect means that the forests control the runoff of freshwater, especially at the end of a flood tide. The evaporation of water and the build‐up of salt generated by the physiological activities of the trees helps to generate gradients of salt and other materials, both laterally and longitudinally, especially during a long dry season.

A significant lateral gradient within mangrove creeks during the dry season can be attributed to high evapotranspiration (Le Minor et al. 2021). Weak stratification usually prevails at the headwaters of mangrove waterways in the wet season as the result of freshwater input from rivers. Such gradients have been observed in Gazi Bay in Kenya (Kitheka 1997) and in the Konkoure River delta in Guinea (Wolanski and Cassagne 2000). In arid‐zone estuaries, the salinity structure is inverse due to the lack of freshwater input and the high evaporation rate, especially in relation to salt flats and mangrove bordering the estuary; salinities can be >50 (Kitheka 1997).

The behaviour of tidal water is also longitudinally complex. Longitudinal diffusion is proportional to the square of the water velocity, which means that mixing rates are very small at the headwaters of mangrove creeks where currents are also very small. Water speed decreases from the mouth to the headwaters along the length of a waterway. The longitudinal and cross‐sectional gradients in current speed are partly the result of shear dispersion processes that are magnified by the presence of the forest. This diffusion process drives the intensity of mixing and trapping. These complex processes translate into long residence times for water near the head of the waterway, especially in the dry season.

All estuaries, including those inhabited by mangroves, exhibit secondary circulation patterns superimposed on the primary tidal circulation. This phenomenon is responsible for the often observed trapping of floating detritus in density‐driven convergence fronts during a rising tide (Stieglitz and Ridd 2001). These fronts occur in well‐mixed estuaries due to the interaction between the velocity of water across the estuary and the density gradient up the estuary. Due to friction, the velocity is slower near the riverbanks than in the centre of the estuary, thus causing on flood tides a greater density mid‐channel than at the banks. A two‐cell circulation pattern results from the sinking of water in the centre of the estuary. The existence of these cells has ecological consequences. A net upstream movement of floating debris occurs, on the order of several km per day; mangrove propagules are unlikely to enter the mangrove forest when these cells are present and will accumulate in large numbers in ‘traps’ upstream from the convergence and upstream from the mangrove fringe (Stieglitz and Ridd 2001). Trapping of propagules is not conducive to the natural strategy of maximizing dispersal of seeds.

Within the mangrove forest, trees, roots, animal burrows and mounds, timber, and other decaying vegetation exert a drag force on the movement of tidal waters (Le Minor et al. 2021). The drag force of the trees can be simplified to a balance between the slope of the surface water and the flow resistance due to the vegetation (Mullarney et al. 2017). Water flow in the forest depends on the volume of the trees relative to the total forest area. The momentum of tidal forces is greater than the shear stress induced by the presence of the trees, including friction with the soil surface. Even the presence of dense pneumatophore roots induces turbulent friction near the forest floor (Norris et al. 2019). The presence of mangrove seedlings results in alteration of tidal flow by modifying the vertical velocity and the magnitude of turbulent energy (Chang et al. 2020). The dynamics of tidal forces in mangrove forests changes in relation to different tree species, density of the vegetation, and state of the tides.

Currents in the forest itself are not negligible, and a secondary circulation pattern is usually present due to the vegetation density and the overflow of water into the forest at high tide (Mazda et al. 2007; Mullarney et al. 2017). This secondary circulation enhances the trapping effect of tides. The drag force has two main influences: (i) inundation of the forest is inhibited and this decrease in water volume results in smaller dispersion and (ii) the trapping of water in the forest is enhanced, favouring dispersion. Thus, the magnitude of tidal trapping depends on the drag force due to the vegetation, so the magnitude of dispersion depends ultimately on the vegetation density.

Animal structures also influence water circulation in mangrove forests (Le Minor et al. 2021). Benthic organisms, especially crabs, produce numerous burrows and tubes in the forest floor through which tidal water flows. Tidal water flows through a labyrinth of interconnected crab burrows in the same direction as the surface current (Ridd 1996; Stieglitz et al. 2013). The flow through the burrows is caused by a pressure difference between multiple burrow openings. The total quantity of water that flows through crab burrows can range from 1000 to 1 010 000 m³ representing from 0.3 to 3% of the total volume of water moving through the forest. This percentage is not large but is enough for the burrows to play an important role in flushing salt away from mangrove roots. The transport of salt derived from the tree roots results in variations in the density of water flowing through the burrows, having an impact on flushing time. Burrows, tubes, and other biogenic structures impart some significant delay in water flow, assisting in the trapping of water in the forest (Stieglitz et al. 2013).

Not all water entering an estuary leaves via the surface. Some water leaves via subterranean pathways such that mangrove estuaries often have significant groundwater flow. This flow can have significant biogeochemical and biological effects, such as removing excess salt from mangrove roots and removing high concentrations of respired carbon from the forest to the adjacent coastal zone (Gleeson et al. 2013). Crab burrows and other biogenic structures can facilitate groundwater flow. The flow of groundwater in mangroves usually has three components (Mazda and Ikeda 2006): (i) a near‐steady flow towards the open ocean due to the pressure gradient induced by the difference in height between water levels in the forest and the open sea, (ii) a reversing tidal flow with a damped amplitude and delayed phase towards the forest, and (iii) a residual flow towards the forest caused by the damped tidal flow. This residual flow reduces the outflow of water from the forest towards the sea.

Mangroves often receive a significant amount of wave action, even in an estuary. Mangroves attenuate wave energy via two primary mechanisms: (i) multiple interactions of waves with mangrove trunks and roots and (ii) bottom friction. The latter is not well understood, but a significant amount of attention has focused on the effect of the presence of tree trunks and roots. Forces induced by waves on tree stems and roots are inertial and drag‐type forces, with drag force dominating for most mangroves (Hashim et al. 2013). The degree of wave attenuation increases with increasing tree diameter, although interactions between tree stems can influence the extent of drag. Waves within a mangrove forest are strongly dissipated by these interactions. Dissipation of wave energy is a function of total tree area which is in turn a function of both tree diameter and forest density. Water depth can also play a role in wave dissipation. For a very dense forest, wave energy is almost totally dissipated within 40–50 m from the mangrove‐sea boundary, but in less dense forests, about 35% of the incident wave energy is still extant behind the forest (Hashim et al. 2013). In mangrove forests that are small in area due to urban disturbance, such as in Singapore, the percentage of wave height reduction is higher under storm events compared to normal conditions, with vegetation drag being the main mechanism of wave dissipation; mangrove density and width were positively correlated to the percentage of wave height reduction during a storm (Lee et al. 2021). Mangrove roots contributed to a larger percentage of wave height reduction than trunks and canopies, although there were no significant differences in the extent of wave height reduction between forest types, incident wave heights, and water levels. Thus, even comparatively small, disturbed mangrove forests can offer some protection from wave energy.

Mangroves grow best where wave energy is low and tidal flow and subsequent attenuation within the forest result in the deposition of fine particles from the incoming tidal water. The transport and deposition of suspended sediment is controlled by several interrelated processes: tidal pumping; baroclinic circulation; trapping of small particles in the turbidity maximum zone; flocculation; the mangrove tidal prism; physicochemical reactions that destroy flocs of cohesive sediment; and microbial production of mucus (McLachlin et al. 2020). The relative importance of these processes is site‐specific. Along the river's edge, mud banks form not just as the result of baroclinic circulation but also as the result of tidal pumping and mixing, especially in the turbidity maximum zone. A turbidity maximum zone is formed within an estuary where the residual inward bottom flow meets the outflow river flow. This zone is usually at the most landward point reached by the saline water flow. The water is generally shallowest at this point because this is a convergence point where sediment accumulates, the bulk of which creeps along the bottom. Some suspended matter settles here as the net current velocities in the zone are low; the zone is not stationary but moves with tidal ebb and flood.

In relation to the turbidity maximum, flocculation of particles begins at salinities often <1; the largest flocs remain near the river bottom. The small flocs and unflocculated particles move further downstream with the currents where they aggregate with local particles (Gratiot and Anthony 2016). As floc size increases, they move towards the riverbed where they are entrained upstream by the baroclinic circulation. Due to tidal pumping, these flocs are carried further upstream at flood tide rather than downstream at ebb tide. The flocs are a loose matrix of clay and silt particles, typically a few micrometres in diameter, with their small size controlled somewhat by the strength of the tidal currents. Disaggregation starts when tidal velocities exceed 1 m s⁻¹. During spring tides, flocs are typically between 15 and 40 μm in diameter and are larger during neap tides. These flocs are colonised by bacteria, protists, and fungi and their extracellular mucus and threads, which help to cement the flocs and to maintain size when subjected to turbulence. Within the forest, flocs can remain in suspension owing to turbulence generated by flow around the trees. The settling of flocs occurs for a short period when the tides turn from rising to falling and the waters are quiescent. Settling is facilitated by the sticking of microbial mucus and by formation of invertebrate faecal pellets.

It is thus correct to state that mangroves actively help to settle fine particles and are not just passive importers. The size, shape, and distribution patterns of mangrove trees have a profound impact on sedimentation. Tree species with large above‐ground root systems, such as Rhizophora, facilitate the deposition of particles to a much greater degree than tree species without extensive roots. The flocculation of particles results in faster settling velocities; most flocs settle within 30 minutes just before high slack tide. Until slack water, turbulent wakes created by tree trunks, roots, and pneumatophores maintain particles in suspension. Once in the forest, however, conditions are unfavourable for them to be resuspended as the high vegetation density inhibits water motion.