Sphagnum magellanicum growth and productivity in Chilean anthropogenic peatlands Crecimiento y productividad de Sphagnum magellanicum en turberas antropogénicas de Chile

Sphagnum peatlands are threatened at a global scale, not only by peat extraction, but also by Sphagnum harvesting. In Chile, dry Sphagnum moss is mainly exported for use as substrate for horticulture and orchids. Although the use of Sphagnum within Chile is limited, there are no data about its productivity and growth. These peatlands have a special microtopography with hummocks, hollows and lawns, which vary the distance of moss to the water table level. In these ecosystems, the water table is almost all year near the surface. We measured cumulative and relative growth rates and productivity during approximately one annual cycle in private Sphagnum peatlands that are being yearly harvested for commercial purposes. We evaluated the relationship between Sphagnum magellanicum growth and productivity with microtopography and water table depth. Productivity, cumulative and relative growths were higher in lawns than in hummocks. Overall and relative growth of S. magellanicum showed a negative relationship with depth of the water table. There were also differences between sites, some of them showed high growth rates, but low productivity. Sphagnum extraction in Chile, is now at low scale, but the growing international market demands constitute a real threat to the resource.


INTRODUCTION
Peatlands are extensive types of wetlands characterized by the accumulation of dead plant matter (peat) that may form layers up to 20 m thick.This ecosystem has important consequences for climate change, human welfare and biodiversity conser vation (Parish et al. 2007).In par ticular, peatlands play a critical role in water storage, regulation and filtration, thus protecting the environment against fl oods and providing clean water; they are able to sequester even higher volumes of carbon dioxide than the world forests, which might potentially transform them in important sources of CO 2 if destroyed, and fi nally they harbor a unique biodiversity, which includes many endangered species.In spite of these, peatlands are being degraded and threatened in many regions as a result of plantations, land use change, drainage, fi re and its use as energy fuel (Sala et al. 2000, Parish et al. 2007).In southern Chile, the use of fi re and clearcutting since the middle of the 19 th century in places with low drainage has created areas of wetlands dominated by species of the genus Sphagnum (i.e.anthropogenic peatlands, Díaz et al. 2008).A similar situation has happened in New Zealand (Whinam & Buxton 1997) and now peatlands in both countries are one of the major sites for commercial harvesting of Sphagnum.
Sphagnum peatlands are characterized by a continuous layer of moss with sparse shrub or tree cover, and whose main water source is rainfall or snowmelt.They have a special microtopography with hummocks, hollows and lawns that var y the distance of moss to the water table (Van Breemen 1995).In Chile, Sphagnum peatlands are located in the south of the country (from 39º S to 55º S) and not only are threatened by peat extraction, but also by Sphagnum har vesting (i.e. the extraction of the fi rst -active-layer, the acrotelm), mainly for its use as a substrate in horticulture.Once har vested, Sphagnum re-growth depends on different factors such as water and light availability, distance to water table and air temperature, among others (Gerdol 1995, Buxton et al. 1996, Gunnarsson et al. 2004).Although the use of Sphagnum magellanicum within Chile seems limited and there are no statistics about its production, the exported volumes have increased exponentially in just 10 years: from 360 tons in 1997 to 2675 tons in 2007.This situation coupled to the lack of research, slow regeneration, absence of management plans, and little or no enforcement of har vesting regulations, seriously threat the persistence of this unique and limited resource.Here, we measured growth rates and pr oductivity during appr oximately one annual cycle in nine private Sphagnum peatlands that are being yearly harvested for commercial purposes.Thus, the overall aim of this study was to evaluate the relationship between Sphagnum magellanicum's growth and productivity with microtopography and water table depth.

Study area
Nine study sites (below 100 mas l) were chosen in private peatlands in the vicinity of Puerto Montt (41° S 72° W).With the exception of one site, which has not been harvested for at least 30 years (H.Aburto C), all peatlands are harvested yearly (i.e.fi rst layer of moss) for commercial purposes.Sphagnum magellanicum is the dominant species in all sites and the one that farmers export.Measurements of growth and productivity were done on this specie.Annual mean precipitation is 2110 mm with a dry period during summer months (January and February) and annual mean relative humidity is 85 %.Annual mean air temperature is 10 °C ranging from 4 to 19 °C (annual mean minimum and maximum temperatures, respectively) (Carmona et al. 2010).

Growth rates and productivity measurements
We have experimentally measured growth rates and productivity and recorded the depth of the water table (hereafter DWT) over approximately one annual cycle (August 2006 to July 2007).At each site, we installed three plots (2 x 1 m), each one with a hummock and lawn area (hummocks with at least 10 cm of difference in height of lawns).Within each plot we measured growth rates and productivity in at least fi ve individual plants.We measured growth every two months with the cranked wire method propose by Clymo (1970).It basically consists in a wire (shaped like a car starling handle) placed over the Sphagnum carpet with the horizontal section leveled with the capitula.The growth is measured from the amount of free wire above the surface.We estimated overall cumulative growth over the entire year, and relative growth as the growth rate per month.Productivity was estimated using 10 stems of each treatment in each plot.We cut the fi rst cm under the capitulum, dried and weighed (B stem ).We used the equation proposed by Gehrke (1998) as ∆h * B stem * ρcap where ∆h is the increase in shoot length measured with the cranked wire method propose by Clymo, B stem is biomass for unit length of the shoot and ρcap is the spatial density of capitulum, measured counting the number of capitulum 10 times per treatment in each plot with a square of 4 cm 2 .

Statistical Analyses
We used a linear mixed modeling approach to evaluate the effect of the microtopography (hummock and lawn) and DWT on our variables, while taking into account the spatial pseudoreplication of our design and the variance heterogeneity across sites.Productivity was analyzed using microtopography as a fixed effect.Growth-related variables were analyzed using DWT and microtopography as fixed effects.Hypothesis testing for fixed and random effects was carried out using Likelihood Ratio Tests of nested models based on Maximum Likelihood (ML) and Restricted ML estimation, respectively (West et al. 2007).The asymptotic null distribution of the test is a χ 2 with degrees of freedom equal to the difference in the number of parameters between the two models (West et al. 2007).All variables were log 10 transf ormed to meet normality assumptions.We performed all the statistical analyses using R (R Development Core Team 2009).Results are presented as mean ± 1 SD.
Although cumulative growth was measured over a different number of days in the different sites (range 226-354, see Fig. 2 for details), number of days did not have an infl uence on cumulative growth (χ 2 1 = 0.02, P = 0.888).

DISCUSSION
Water availability, one of the factors that deter mine growth of Sphagnum species, depends on the distribution of rainfalls, evaporation and the mean annual water table level (Gignac & Vitt 1990, Grosvernier et al. 1997).As expected, overall and relative growth of S. magellanicum showed a negative relationship with DWT.Weltzin et al. (2001) suggested that DWT infl uences the growth of S. magellanicum only when it is between 20 and 25 cm under the moss surface.On the other hand, Grosvernier et al. (1997) proposed that when water table is below 40 cm under the surface (drainage conditions), physicochemical properties of peat change and these in turn affect negatively Sphagnum's growth.It has been suggested that Sphagnum species tend to built tight hummocks in extremely dr y conditions in order to improve the capillar y network and thus water transport (Rydin 1985, Li et al. 1992).This phenomenon could explain Sphagnum's growth with little or no water availability, even in summer when water table is lower than in other seasons.Sphagnum's growth rate fl uctuated extensively (range: 0.1-17.3cm year -1 ), with a mean cumulative growth of 3.38 ± 2.87 cm year -1 (mean ± SD).This is similar to reported values for temperate alpine ecosystems (i.e.Italy, Gerdol, 1995) and higher than the mean growth for higher latitudes (e.g., Sweden, Aerts et al., 1992, Wallen et al. 1988), probably because of climatic conditions (high temperatures, water availability and light intensity) during the growing season in Southern Chile.Sphagnum's productivity also showed a wide range, from 67 to 1590 g m -2 year -1 , with an overall mean of 458 ± 287 g m -2 year -1 .This is twice the value recorded for S. magellanicum in Tierra del Fuego National Park, Chile (54°51' S) (Robson et al. 2003).A likely explanation for such difference might be related to mean annual temperatures at both sites: 11.2 °C in Chiloé and 5.8 ºC in Tierra del Fuego.Thus, our results are in agreement to what was proposed by Gunnarsson (2005), that mean annual temperature was the most important factor explaining global productivity in Sphagnum dominated wetlands.
As expected, growth (overall and relative) and productivity were higher in lawns than in hummocks (see panel (A) in Figs.1-3).However, this trend was not (statistically) consistent across sites, as we detected an interaction between sites and microtopography.That is, the incidence of microtopography on growth and productivity was more evident in some sites.More importantly, sites with the highest productivity were not necessarily the sites with the highest growth in length (see panel (B) in Figs.1-3), and vice versa.Our results suggest that it is extremely important to include biomass production and not only height increment or growth when considering the right time to har vest.If only growth in length is used as a proxy for the decision to harvest or not, overexploitation of the resource will be imminent.Moss need to growth not only in length but also in biomass, because when farmers sell it, they sell it by weight.Furthermore, some sites are even harvested on a yearly basis.
In the past decades the increase of moss extraction and export in Chile has generated the need to analyze the cur rent state of conser vation of the resource.In 2010 Chile exported 4214 tons of Sphagnum.Our results indicate an average annual productivity for Sphagnum as 4.6 tons ha -1 year -1 , then our productivity estimates suggest that ever y year 920 ha are needed to keep those export volumes.Fur thermore, most of Sphagnum moss in the Chilean Lake District is located in private properties of small landowners (ca.11500 ha) (ODEPA, 2007) and its exploitation has provided a signifi cant income to thousands of families in the Region.However, moss extraction at the current rate and without a proper management, would mean the extinction of the resource in less than 12 years.In fact, we are already witnessing the evident symptoms of overexploitation and deterioration because low tender prices (US$ 0.3 to 1.3 per Kg) restrict the revenues to moss collectors stimulating the extraction of increased volumes to increase revenues.This situation imposes the urgent need to promote sustainable management practices for the resource, and ensure the economic viability of the product in the long term.

Fig. 1 :
Fig. 1: Overall productivity of Sphagnum magellanicum according to microtopography (hummocks and lawns) (A), and in each study site (B).Data are presented as mean ± 1 SD.