- Email: [email protected]
Wind effects and regeneration in broadleaf and pine stands after hurricane Felix (2007) in Northern Nicaragua
Forest Ecology and Management 400 (2017) 199–207
Contents lists available at ScienceDirect
Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco
Wind effects and regeneration in broadleaf and pine stands after hurricane Felix (2007) in Northern Nicaragua E. Rossi a,b,⇑, I. Granzow-de la Cerda c, C.D. Oliver d, D. Kulakowski b,e a
Instituto de Bioética, Pontificia Universidad Javeriana, Bogotá, Colombia Graduate School of Geography, Clark University, Worcester, United States c CREAF and Univ. Autónoma de Barcelona, Cerdanyola del Vallés 08193, Spain d Yale School of Forestry and Environmental Studies, United States e WSL Institute for Snow and Avalanche Research, SLF, Davos, Switzerland b
a r t i c l e
i n f o
Article history: Received 27 October 2016 Received in revised form 18 May 2017 Accepted 21 May 2017
Keywords: Stand dynamics Wind disturbances Tropical forestry Forest management Región autónoma del Caribe Norte – RACN
a b s t r a c t Large-scale wind disturbances shape forest structure and composition and leave long lasting legacies. Adequate understanding of the role of disturbances in tropical stand dynamics is necessary to guide management efforts. In this study, we used field data to characterize the effects of a major hurricane in broadleaf and pine stands in Northern Nicaragua. First, we described tree and stand attributes associated with observed structural effects: branch loss, snapping and uprooting. Secondly, to assess the potential influence of hurricane Felix on stand composition, we characterized two key life-history traits: regeneration through resprouting and shade tolerance. Findings indicated that tree attributes such as diameter at breast height (dbh) and height to diameter ratio (hdr), were strongly associated with the type and magnitude of wind effects. All trees >70 cm dbh exhibited visible effects and trees taller than 14 m were more likely to be uprooted (7.3% vs. 0.8% of total). Results confirmed that Felix caused significant structural effects in broadleaf stands and mild effects in pine stands. Abundant post-hurricane resprouting was observed in both shade-tolerant and shade-intolerant species but was absent in pines. Among canopy trees we found eleven shade-intolerant species that exhibited abundant resprouting. These species could become dominant in the next decade. Our findings illustrate the role of wind disturbances on tropical stand dynamics at different spatial and temporal scales. Ó 2017 Elsevier B.V. All rights reserved.
1. Introduction Disturbance events and competitive interactions among trees drive forest dynamics (Sousa, 1984; Foster et al., 1998; West et al., 2009; Turner, 2010). Disturbances change the physical environment, reallocate resources, and create biological legacies that influence ecosystem processes (Norberg, 1988; Foster and Boose, 1992; Pacala et al., 1993; Kulakowski and Veblen, 2003; Boose et al., 2004). Disturbances may also be important in maintaining species diversity, particularly in species-rich tropical ecosystems (Huston, 1979; Warner and Chesson, 1985; Denslow, 1995; Granzow-de la Cerda et al., 1997). Although the factors that determine forest susceptibility to wind disturbances have been characterized in temperate ecosystems, they have received limited attention in tropical forests (e.g., Foster and Boose, 1992; ⇑ Corresponding author at: Instituto de Bioética, Pontificia Universidad Javeriana, Bogotá, Colombia E-mail address: [email protected] (E. Rossi). http://dx.doi.org/10.1016/j.foreco.2017.05.034 0378-1127/Ó 2017 Elsevier B.V. All rights reserved.
Boucher et al., 1990; Lugo and Scatena, 1996). Specifically, it is not well understood how tree-and-stand attributes influence susceptibility to wind and subsequent stand dynamics (Lugo, 2002). A few general relationships between wind effects and tree-andstand attributes have been identified. First, some studies have found a positive relationship between increasing tree height and increased uprooting (Everham and Brokaw, 1996). Second, stand susceptibility to wind is linked to the height to diameter ratio (hdr) of dominant trees. This ratio depends on light availability and stand density: trees growing in full sunlight tend to have lower values, while trees growing in shaded conditions – e.g., in dense stands-tend to have greater values (Mitchell, 1995). For this reason foresters use it as a quantitative indicator to guide thinning operations (Wilson and Oliver, 2000). The importance of forest type (i.e., broadleaf vs. pine) on the severity of wind effects has been examined in numerous locations with different results (Everham and Brokaw, 1996). Although conifers are sometimes more susceptible than angiosperms, given the greater diversity in wood density and tree architecture of the
200
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
latter, wind effects seem to be more dependent on stand characteristics such as tree density and canopy height, rather than species’ traits (Boucher et al., 1990; Mitchell, 1998). By snapping and uprooting trees differentially, windstorms also shape stand structure (Zimmerman et al., 1994; Vandermeer et al., 2000). For example, in Puerto Rico windstorms occur every decade and thus favor the development of dense and homogenous stands with low regeneration and high competition (Boose et al., 2004; Van Bloem et al., 2007). In contrast, in Nicaragua, windstorms occur approximately once every century and create space for the establishment of surviving and new seedlings (Vandermeer et al., 2001). Most studies of tropical forest dynamics have focused on shortterm and small-scale disturbances, i.e., gap dynamics (Hubbell, 1979; Denslow, 1987; Denslow et al., 1998; Condit et al., 2004), in part because of the difficulties in determining tree age using growth rings and identifying sterile individuals (Rejmanek and Brewer, 2001; Chambers et al., 2013). Consequently, the role of infrequent hurricanes in shaping forest composition and diversity is less understood (Nelson et al., 1994; Zimmerman et al., 1994; Baker et al., 2005). By opening space over large areas (>1000 km2), hurricanes create landscape heterogeneity, change the ecological space available to organisms and favor tree recruitment (Huston, 1979; Denslow, 1995; Negrón-Juárez et al., 2010). For example, Vandermeer et al. (2000) documented an increase in tree species richness after the passage of hurricane Joan in 1988, which was attributed to the growth of the seedling bank coupled with resprouting of affected trees. Their 10-year data series indicated that following geographically large disturbances (5000 km2), dispersal limitations prevented pioneer species (e.g., Cecropia spp.; Ochroma spp.) from rapidly colonizing available space, in contrast to small gaps (0.5 km2). Lastly, studies in Puerto Rico have showed that hurricanes can also shape the composition of the seedling layer (Comita et al., 2010). It has been proposed that tree response to wind disturbances depends on three main life-history traits: resprouting capacity, dispersal ability and shade tolerance (Canham et al., 1990; Oliver and Larson, 1996; Montagnini and Jordan, 2005). While post-hurricane resprouting and dispersal ability have been characterized in some Caribbean forests (Boucher and Mallona, 1997; Van Bloem et al., 2005), it is unknown if tropical canopies are dominated by shade tolerant or shade intolerant species, or a combination of the two. This information is important to understand historical stand dynamics and develop adequate management strategies (see Baker et al., 2005). By compiling empirical data from numerous studies, Gunter et al. (2011) defined three shade classes: tolerant, intolerant, and partially intolerant, which were used in this study to classify canopy species. The landfall of hurricane Felix provided us with the opportunity to characterize the factors that control wind effects and regeneration, specifically: (i) how do tree and stand attributes determine the type and magnitude of wind effects; (ii) what is the relationship between forest type (broadleaf vs. pine) and wind effects; (iii) which canopy tree species resprout following wind disturbances; and (iv) what is the proportion of canopy species that are shade tolerant. By using field data to characterize wind effects at the stand scale, this study complements previous efforts that used aerial surveys and satellite imagery to characterize wind effects at the landscape scale (Inafor, 2007; Rossi et al., 2013).
annual temperature is 25 °C and annual rainfall 2500 mm, with a marked dry season from February to early May. The RACN harbors 1.4 million hectares of forest (43% of Nicaragua’s forests), which include the Bosawás UNESCO’s Biosphere Reserve (Inafor, 2009). Two forest ecosystems cover most the region: dense, closed tropical broadleaf forests located towards the interior and pine savannas on coastal areas. Broadleaf (hardwood) forests are found mostly in crumbstructured humic clays in areas of undulating topography (Parsons, 1955). For centuries these forests have attracted foreign logging companies in search of mahogany (Switenia macrophylla) and Spanish cedar (Cedrela odorata) which were selectively harvested along rivers and roads. Yet, up until hurricane Felix’ impact, a significant portion of these forests remained inaccessible and poorly studied. Presently, numerous valuable timber species can be found in broadleaf stands, including Calophyllum brasiliense, Carapa guianensis, Dialium guianense, Guaiacum sanctum, Hyeronima alchorneoides, Hymenaea courbaril and Terminalia amazonia (Smith, 2003; Inafor, 2009). Over the last three decades, land-use practices consisted on selective logging and fuelwood collection, although a large portion of the forests of the region remains unmanaged (Salazar, 2005; Diaz, F. personal communication, 2010). Because of the high density and moisture, these broadleaf stands rarely burn (Inafor, 2009). Pine savannas dominated by Pinus caribaea comprise approximately 10,000 km2 along the Atlantic coast of Nicaragua. The soils of the pine savannas are mostly composed of nutrient poor sands, highly weathered quartz-gravels and clays in poorly drained areas (Parsons, 1955; Myers et al., 2006). This ecosystem harbors numerous herbaceous species and depends on recurrent fires that create adequate conditions for pine regeneration and prevent hardwood establishment (O’Brien et al., 2008; Ratnam et al., 2011; Hoffmann et al., 2012). Due to the abundance and accessibility of pines across the coastal plains of the region, logging and fuelwood collection have been common for over a century (Castilleja, 1993; McSweeney, 2004). In addition, near communities, pine stands are often burned to promote grass growth for cattle and deer (Alvarado, 2010). On 4 September 2007, hurricane Felix made landfall near the community of Awastara in Northern Nicaragua (14°180 44.0700 N, 8 3°120 12.0400 W), as a Saffir-Simpson category five hurricane (NOAA, 2013). The cyclone travelled across the region and became a low pressure system over northern Honduras early on September 5th (Brennan et al., 2009; Fig. 1). Felix caused substantial damage to infrastructure in coastal communities and affected more than 3000 km2 of broadleaf and pine stands. Estimates based on aerial surveys and satellite imagery suggest that one million m3 of tropical hardwoods were blown down by this hurricane (Inafor, 2007). Weather records indicate that no major storms had affected Northern Nicaragua since the late 1890s (Beven et al., 2008; Kar, 2010; NOAA, 2013). Similarly, historical data indicates that although numerous tropical storms regularly pass over Northern Nicaragua, most Atlantic hurricanes have not made landfall in this region (NOAA, 2016). Taken together, historical and weather records suggest that broadleaf stands located in the center of the RACN had remained undisturbed for at least a hundred years before Felix’s impact.
1.1. Study area
2.1. Data collection
The Autonomous Region of the Northern Caribbean of Nicaragua (hereafter RACN) comprises 32,000 km2 and hosts a population of 315,000 (Nicaraguan Census Bureau – INIDE, 2005). Average
Four field sites were defined using land cover maps based on proximity to the hurricane path and availability of large forest patches of broadleaf and pine stands to allow for comparisons
2. Methods
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
201
Fig. 1. (A) Land cover of the autonomous region of the northern Caribbean depicting the path of hurricane Felix and the distribution of broadleaf stands (dark green) and pine savannas (light blue). While pine savannas are accessible, broadleaf stands are mostly located in road less areas. (B) Study area and field site locations around which sampling plots were established. Site names from top to bottom: Miguel Bikan, Santa Clara, Awas Tigni and Moss. Timber extraction areas appear in orange, locations that experienced significant structural effects appear in grey (75% trees uprooted or snapped). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
between the two forest types (Vreugdenhil et al., 2002; Inafor, 2007). The four sites were named after the nearest indigenous community: Moss, Awas tigni, Miguel Bikan and Santa Clara. Sampling was conducted between June and August of 2011 in the Moss site, during 2012 in Awas Tigni and Santa Clara and Miguel Bikan in 2013. Distance between sites ranged from 8 to 21 km (Fig. 1). In each site, we randomly established 6–8 sampling plots separated by over 200 m and located at least 100 m away from the edge of the patch to avoid border effects. Combining the four sites; fifteen plots were established in broadleaf stands and fifteen plots in pine stands for a total of 30 plots (Fig. 1b). Each plot had a size of one hectare (100 100 m). Large plot sizes were selected to control for felled trees rooted outside the plot. Large trees falling unto small plots can affect all trees in the plot (e.g., 60% of trees uprooted, 40% snapped) causing an overestimation of wind effects at the stand scale. All plots were located within 30 km from the hurricane path and were representative of site conditions (elevation, slope, and aspect). Height, diameter at breast height (dbh) and taxonomic identity were recorded for all standing and fallen trees >10 cm dbh. Tree height was measured using a telescopic pole for individuals less than 15 m high and a hypsometer (Nikon 550) for all individuals taller than15 m. Trees were identified initially in the field and verified in the herbarium using the taxonomic criteria of Flora de Nicaragua (Stevens, 2001). For all trees >10 cm dbh, wind effects were recorded as: uprooted, snapped, lost branches, or no visible effects. Tree condition after the hurricane (alive or dead), resprout number, height and position on the stem were also recorded. A tree presented ‘‘branch loss” when branches larger than 10 cm in diameter were broken, or when the trunk exhibited scars or broken branches. Branch loss was defined in this way to capture changes that were visible years after the event, but did not capture the loss of branches smaller than 10 cm in diameter. Observations regarding the presence of tree mounds, hardwood stumps and fire scars were recorded for each stand. Information on the type and intensity of land use i.e., the presence of logging, grazing and fuel-
wood collection was gathered by interviewing community members. 2.2. Data analyses Data analyses are organized following the research questions. First, to determine how tree and stand attributes influenced the type and magnitude of wind effects we first compared the number of canopy (overstory) trees that experienced structural effects (branch loss, snapped, and uprooted). Mean overstory height was calculated for each stand following Oliver et al. (2005); to avoid confounding size effects, comparisons were conducted separately for trees 25 cm dbh and >25 cm dbh, which roughly segregates understory from overstory trees in broadleaf stands. Previous studies and field observations indicated that most broadleaf trees with diameters >25 cm were part of the overstory (Pérez-Flores, 2000). Whereas in pine stands trees 10 cm dbh were regarded as understory and trees >10 cm dbh were considered as part of the overstory. Subsequently, chi-square analysis of 2 2 contingency tables was used to test for differences in the frequencies of different wind effects (e.g., uprooted vs. snapped). Tests were not conducted if the expected frequency in any cell of the table was less than five. Unequal variance t-tests were conducted to compare the means of trees by size class. The influence of height on susceptibility to wind effects was examined by comparing median tree height within each wind effect class. Statistical significance of differences in the median heights among classes was evaluated with one-way analysis of variance; pairwise comparisons were made with Dunn’s method (Zar, 1999). Hurricane induced mortality was defined as the sum of uprooted and snapped non-resprouting trees. Second, to characterize the relationship between forest type (broadleaf vs. pine) and wind effects cover types were distinguished based on the abundance and dominance of Pinus caribaea and Dialium guianensis, in conifer and broadleaf stands, respectively (Pérez-Flores, 2001). The following four cover types were defined: (1) open pine forest (<400 trees/ha), (2) dense pine forest
202
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
(>400 trees/ha), (3) stands dominated by D. guianensis (>10% of stand’s basal area), and (4) mixed broadleaf. Differences in susceptibility to wind among cover types were examined by comparing mean percentage of affected trees in each class with a Kruskal– Wallis one-way analysis of variance; pairwise comparisons were made with Dunn’s method (Zar, 1999). Best-subset regression was used to identify stand attributes that could predict wind effects using the SigmaStat package (Jandel Corp. 1995). The dependent variable was the percentage of trees that were uprooted, snapped or lost branches; independent variables included stand density, mean height of the largest 30 trees, total basal area and distance from the hurricane path. All variables passed tests of normality and homogeneity of variance. Classification tree analysis (CTA) was used to examine the relationship between tree-and-stand attributes and wind effects. This method is well suited for the analysis of ecological data due to its ability to examine both numerical and categorical variables and non-linear relationships (De’ath and Fabricius, 2000). Classification trees are fitted using a binary recursive partitioning whereby data are successively split into more homogeneous groups using combinations of variables. We obtained CTs for broadleaf and pine trees and separately for each forest type using the CRT growing method and the Gini splitting rule in the SPSS statistical package. Species richness and similarity indices were calculated using EstimateS software (Version 9.1, Colwell, 2013). For every species the Importance Value Index (IVI) was calculated to identify dominant canopy species. The IVI is the sum of relative density, relative dominance, and relative frequency and ranges from 0 to 300, where 300 is the maximum value (Curtis and Cottam, 1962). Lastly, data on tree condition after the hurricane and resprouting were compared by species. Specifically, we aimed to determine which species exhibited resprouting capacity, whether resprouting was influenced by snapping or uprooting, and which species were tolerant to shade, intolerant or partially intolerant. This information can be useful to determine if hurricane Felix will cause changes in canopy composition in the next few decades.
Fig. 2. Percentage of wind affected trees sorted by dbh size class in both forest types. Trees >50 cm dbh were more likely to experience branch loss (p < 0.001; unpaired t-test with unequal variances). All trees >70 cm dbh exhibited wind effects.
of trees in the understory was too small for statistical analyses (Supplementary Fig. 1). 3.2. Stand attributes and wind effects
3. Results 3.1. Tree attributes and wind effects Among the 10,461 trees recorded, 652 had been uprooted, 341 snapped, 468 lost branches, and 9000 did not exhibit visible effects. Defoliation was not reported because sampling was carried out years after the event. Tree size was positively correlated with wind effects: most trees (90%) larger than 70 cm (dbh) exhibited visible effects, while trees larger than 80 cm dbh were more likely to be uprooted (p < 0.001; unpaired t-test with unequal variances, Fig. 2). Wind effects were also positively correlated to tree height: the number of uprooted trees increased abruptly for individuals taller than 16 m (p < 0.01; unpaired t-test with unequal variances, Fig. 3). Similarly, differences in median height were found among all class pairs (P < 0.001) except uprooted vs. branch loss. In contrast, the number of snapped trees was not correlated to dbh or height. Susceptibility to wind was also influenced by tree position in the canopy (See supplementary Fig. 1). In broadleaf stands, the overstory experienced significantly larger effects than the understory (X2 = 872, p < 0.05): overstory trees exhibited 18.9% branch loss, 11.1% snapping and 20.9% uprooting; while the figures for understory trees were 2%, 4% and 7% respectively. Wind effects were less severe in pine stands: among overstory trees 1.29% lost branches, 0.19% snapped and 0.93% were uprooted; the number
There were contrasting differences in stand density and tree size between broadleaf and pine stands that shaped wind effects. Although pine stands 28, 29 and 30 presented over 700 trees/ha, overall, pine stands were less dense, shorter and exhibited lower hdr (Table 1). The low density and ‘‘open structure” of pine stands reflects a history of logging, coupled with grazing and fire (Table 1). The presence of stumps and fire scars in numerous stands indicated that sampled stands had been managed. Although field data were insufficient to precisely describe land use, field observations and local people indicated that management intensity was higher in pine stands. Regression analyses indicated that maximum canopy height (height of the largest 30 trees in the stand) predicted wind effects in both forest types (p < 0.001; adjusted r2 = 0.52) (Table 1). Cover types also exhibited significant differences in wind effects (Table 2). While mixed broadleaf forests exhibited the highest percentage of snapping and uprooting (7.1% and 13.3% respectively) stands dominated by D. guianense showed higher percentage of branch loss (9.2%). In contrast, pine stands were less affected than broadleaf stands: 2.5% branch loss; 0.5% snapped; 0.2% uprooted (X2 = 1434; p < 0.001). Classification trees, on the other hand, indicated that forest type and the height to diameter ratio (hdr) explained most of the differences among sampled trees; higher hdr indicated larger effects (See Table 1; Supp. Fig. 2). The CT model obtained accurately classified 85.8% of the observations.
203
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
3.3. Stand diversity and composition In total, 53 species were identified. One species could not be identified beyond family level and five could not be identified beyond genus and were thus assigned to distinct morphotypes. Species richness varied widely between forest types: while broadleaf stands contained up to 65 species in a 1-ha plot, only 11 species occurred in the most species-rich 1-ha plot in pine stands (Table 1). Broadleaf stands were dominated by Dialium guianense, Spondias mombin, Terminalia amazonia and Ficus aff. yoponensis, (ranked by BA) and harbored numerous commercially valuable species including Hyeronima alchorneoides, Calophyllum brasiliense, Carapa guianensis, Swietenia macrophylla and Dipteryx oleifera (Supp. Table 1). Uprooting was frequent for Dialium guianense (40%, 72 individuals) as for Calophyllum brasiliense (33%, 4 individuals). Spondias mombin sustained the highest rate of snapping (22%, 37 individuals), and Ficus aff. yoponensis the highest rate of branch loss (33%, 16 individuals and also high mortality 45.8%). Overall, tree mortality varied across species, ranging from 4.3% in P. caribaea, to 39.1% in Terminalia amazonia (Supp. Table 1). 4. Discussion 4.1. Wind effects
Fig. 3. Percentage of wind affected trees sorted by height class (labels correspond to the upper limit of 2 m classes). Branch loss and uprooting was more frequent among trees >14 m (p < 0.01; unpaired t-test with unequal variances). The number of unaffected trees decreased after 14 m (p < 0.001; unpaired t-test with unequal variances). Median height differed significantly among all classes except uprooted vs. branch loss (P < 0.001, Kruskal-Wallis one way ANOVA).
Findings indicated that wind effects were influenced by tree size: larger DBH and height increased susceptibility to uprooting and branch loss. Whereas subcanopy light-wooded pioneers such as Cecropia insignis, Ochroma aff. lagopus and Hibiscus tiliaceous exhibited low uprooting and mortality (<5% trees), large hardwood species including Dialium guianense, Spondias mombin and Terminalia amazonia experienced higher levels of uprooting, snapping and branch loss. These results suggest that wind effects depend more on tree size and canopy position than species attributes (>12%; Supp. Table 1). The relationship between size and susceptibility
Table 1 Characteristics of sampled stands in Northern Nicaragua.
a
Plot #
Forest type
Land use
Stems/(ha)
Mean tree height (30 trees)
HDR Over.
HDR Under.
BA/ha (m2)
Species richness/ha
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl P P P P P P P P P P P P P P P
SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW L, G L, G L, G L, G L, G L, G L, G L L L L L, F L, F L, F L, F
402 391 287 333 276 283 264 364 283 292 311 236 623 311 355 45 93 83 45 63 34 39 411 607 589 472 731 932 827 783
20 20 19 23 22 24 32 28 27 38 45 32 21 22 22 11.2a 18.5a 21.1a 17.1a 16.8a 14.6a 16.4a 17 16 15 14 15 15 16 14
55.1 56.2 63.7 55.6 51.2 54.0 59.9 65.4 66.2 66.3 71.7 67.8 40.1 41.0 46.0 47.9 49.1 52.3 41.5 47.7 36.3 39.5 17.5 17.7 17.3 17.5 17.0 17.0 17.0 16.9
85.5 94.4 86.6 101.1 87.6 93.6 98.1 88.4 92.3 94.8 109.8 95.3 93.4 96.0 101.4 60.0 62.8 61.3 63.5 71.6 59.7 58.4 17.0 26.2 25.5 25.6 24.1 75.7 21.0 19.7
13.3 13.7 8.1 16.5 15.3 17.2 15.6 17.3 13.0 31.4 28.9 15.7 32.2 30.1 24.7 0.7 4.3 4.2 2.5 2.3 1.7 2.2 159.1 228.1 171.6 130.3 217.2 257.0 246.0 175.6
54 57 51 42 42 30 42 40 34 44 38 37 65 47 53 1 11 1 1 1 5 6 2 3 2 2 4 5 4 3
The 10 tallest trees were sampled in open stands. Bl-broadleaf, P-pine. (L-logging; SL-selective logging; FW-fuelwood; G-grazing; F-fire. HDR height to diameter ratio).
204
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
Table 2 Percent basal area and wind effects for the most common species in the forest cover types. Selected wind effects are: BL – branch loss; SP – Snapped; UP – Uprooted. Values indicate the percentage of the total basal area covered by each species in the stands where it occurs, 100% corresponds to the total basal area of the stand. Wind effects correspond to the percent of trees in those stands that exhibited BL, SP or UP. Stand basal area and number of trees can be found in Table 1. Cover type
(1) (2) (3) (4)
Mixed broadleafa Dominated by Dialium Open pine Dense pine
No. of stands
9 6 7 8
Range of basal area (%)
Wind effects (%)
Dialium
Pinus
Byrsonima
Terminalia
Ficus
Spondias
BL
SP
UR
0–8.2 11.3–22 0 0
0 0 90.5–100 95.1–97.3
0.2–12.3 0–0.1 0 1.9–4. 9
0.9–24.4 0.5–7.7 0 0
0–30.6 0–10.2 0 0
0–26.5 0–5.2 0 0
7.3% 9.2% 2.5% 1.1%
7.1% 5.5% 0.5% 0.2%
13.3% 9.5% 0.2% 0.9%
to wind has been attributed to a ‘‘sail effect”, as large canopies present a greater surface to the wind (Yih et al., 1991; Foster and Boose, 1992). Overall, the height to diameter ratio best captured this ‘‘sail effect” and was perhaps the single most important variable for predicting wind effects (Fig. 4). This ratio depends on light availability because growing trees allocate photosynthates to primary growth (height), before secondary growth (diameter) (Mitchell, 1998; Wilson and Oliver, 2000). Light availability, in turn, is heavily dependent on stand density. Because of this, broadleaf trees in dense stands exhibited higher hdr than pines in open stands. On the other hand, tree susceptibility to snapping versus uprooting, could not be explained by measured attributes, as it might have been determined by a combination of microsite conditions and stochastic factors (Veblen et al., 2001). Although the role of tree architecture on wind effects was beyond the scope of this study, our data suggests that tree architecture was less important for two reasons. First, observations were consistent at the stand level, i.e., canopy trees in the same stand experienced similar effects. Second, crown size was more dependent on growing conditions than on species architecture (Supp. Fig. 2). For example, emergent trees such as Ficus aff. yoponensis were probably uprooted because of their large canopy size rather than crown architecture. The large size (dbh) and dominant canopy position of these trees can be a consequence of selective logging in the area. In the past, loggers probably avoided them because the abundant latex made them difficult to cut. In contrast, very large individuals from other species were extremely rare, suggesting that they were harvested for various purposes including fuelwood. Lastly, the relationship between tree size and wind susceptibility is consistent with observations in other locations (Mitchell, 1995; Foster and Boose, 1992). It is likely that the time since the last disturbance also shaped observed wind effects. Since broadleaf stands had not experienced
Fig. 4. Height to diameter ratios (hdr) among trees affected by hurricane Felix (2007). (1) Broadleaf; (2) Pines. Affected trees were snapped, uprooted or exhibited branch loss.
major hurricanes for at least a hundred years (Kar, 2010; Brennan et al., 2009; NOAA, 2016) and access difficulties prevented logging, especially in the central and northwestern part of the region, it was expected that broadleaf stands would experience larger wind effects (See timber extraction areas in Fig. 1). Pine savannas, in contrast, experienced a different disturbance regime characterized by frequent fires and complemented by logging and grazing (Myers et al., 2006); as a consequence, wind effects were less severe.
4.2. Regeneration and shade tolerance Characterizing life history traits such as tree regeneration and shade tolerance is relevant to understand how hurricane Felix could shape stand composition in the future. As documented in Caribbean forests, we found that tree regeneration through resprouting was common among broadleaf species and absent among pines (Boucher et al., 1990; Vandermer et al., 2000; Van Bloem et al., 2005). Vigorous resprouting was observed in 38 broadleaf species out of 54 and in 264 individuals, but was absent in pine stands. Snapped broadleaf trees were more likely to resprout than uprooted trees (52% vs. 16%) (Fig. 5).
Fig. 5. Observed resprouting in snapped and uprooted trees. Top: number of resprouting trees (n) by structural effect class. Bottom: Percentage of trees exhibiting resprouting by structural effect class. Note that the percentage of trees snapped and uprooted was calculated from the total number of affected trees (n = 1768). In contrast, the % of uprooted resprouting and % snapped resprouting was calculated from the number of trees per class (i.e. snapped: 341, snapped resprouting: 178), (uprooted: 652, uprooted resprouting: 107). The bottom part of the figure illustrates that snapped trees are much more likely to resprout than uprooted trees.
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
Snapped trees were much more likely to resprout than uprooted trees, presumably because the former maintain an intact root system; yet resprouting also occurred among some fallen trees. Since species with resprouting capacity can rapidly occupy available growing space it is possible that these species will become dominant in the next few decades (See Oliver and Larson, 1996). In addition regeneration strategies, tree responses to wind also depend on shade tolerance and maximum potential height growth (Clark and Clark, 1992; Oliver et al., 2005; Gunter et al., 2011). While shade intolerant species grow rapidly under full sun, shade tolerant species can grow slowly in the understory (Pacala et al., 1996). Our classification of shade tolerance indicated that 17 species could be regarded as shade intolerant, 12 as shade tolerant and 14 as intermediate (Clark and Clark, 1992; Gunter et al., 2011). Among shade intolerant species, Spondias mombin, Terminalia ama-
205
zonia and Hyeronima alchorneoides, exhibited abundant resprouting (Table 3). Based on our observations of regeneration and shade tolerance among canopy trees we propose two hypotheses regarding the role of hurricanes in the dynamics of these stands. Although these hypotheses remain speculative they were included to inform future research. First, species with resprouting capacity and fast growth are likely to become dominant after large blowdowns and subsequently exclude other tree species. In this way, hurricanes could favor the establishment of shade intolerant species. This has been observed in Southern Mexico, where the recruitment of Swietenia macrophylla seems to be favored by large windstorms and subsequent fires (Negreros-Castillo et al., 2003; Snook, 2003). On the other hand, in our study area six stands out of fifteen were dominated by shade tolerant species, particularly Dialium
Table 3 Autoecology of dominant tree species. Canopy position: emergent (E), canopy (C), understory (U).
b
#
Species name
Resprouting (# trees)a
Canopy position
Autoecology (shade tolerance)b
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
Pinus caribaea Byrsonima crassifolia Cecropia insignis Dialium guianense Spondias mombin Inga densiflora Terminalia amazonia Homalium racemosum Muntingia calabura Castilla elastica Ochroma lagopus Bursera simaruba Hibiscus tiliaceus Miconia dodecandra Hirtella triandra Brosimum alicastrum Vochysia ferruginea Rheedia sp. Tabernaemontana sp. Tetragastris panamensis Luehea seemannii Virola koschnyi Zuelania guidonia Albizia adinocephala Hyeronima alchorneoides Ficus aff. yoponensis Apeiba tibourbou Lacmellea panamensis Psychotria chiapensis Schizolobium parahybum Morinda panamensis Lysilóma auritum Coccoloba belizensis Symphonia globulifera Calophyllum brasiliense Cupania cinerea Malvaceae sp1. Carapa guianensis Dendropanax arboreus Grias cauliflora Vochysia guatemalensis Oreopanax capitatus Pouteria sapota Lonchocarpus phaseolifolius Guazuma ulmifolia Pouteria campechiana Ehretia austin-smithii Ceiba pentandra Croton panamensis Ochroma sp. Swietenia macrophylla Didymopanax morototoni Dipteryx panamensis Cordia bicolor
0 3 2 32 32 11 27 9 1 16 0 13 0 1 18 4 0 1 0 16 15 2 9 6 5 0 1 1 1 2 1 1 5 10 2 3 0 2 2 0 0 0 0 1 2 0 0 0 0 1 1 0 3 2
C U U E, C C C C U C, U C, U U C, U C, U U C, U C, U C, U C, U C, U C, U C, U C, U C, U C, U E, C E, C C, U C, U U C C, U C, U C, U C, U E, C C, U C, U C C, U C, U C, U C, U C, U C, U C, U C, U U E, C C, U U C, U U E, C C, U
Intolerant Intolerant Intolerant (pioneer) Tolerant Intolerant Shade intermediate Intolerant Shade intermediate Unknown Unknown Intolerant Intolerant Shade intermediate Toleranta Toleranta Shade intermediate Shade intermediate Unknown Unknown Intolerant Unknown Shade intermediate Unknown Unknown Intolerant Tolerant Unknown Unknown Tolerant Intolerant Unknown Intolerant Tolerant Tolerant Tolerant Tolerant Unknown Shade intermediate Tolerant Tolerant Shade intermediate Tolerant Shade intermediate Intolerant Intolerant Shade intermediate Intolerant Intolerant (long lived pioneer) Shade intermediate Intolerant Shade intermediate Intolerant (pioneer) Intolerant Intolerant
Clark and Clark, 1992; Gunter et al., 2011.
206
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
guianense (Table 2) as documented by previous studies in the area (Pérez-Flores, 2001). The dominance of shade tolerant canopy trees can be facilitated by hurricanes that cause partial removal of the overstory and release understory species. Since tree species living in the understory are mostly shade tolerant, hurricanes can allow these species to reach the canopy and occupy available space. Therefore, wind disturbances can shape stand composition in two contrasting ways: while large blowdowns favor the regeneration of shade intolerant species, partial canopy removal favors shade tolerant species present in the understory. It is likely that due to differences in stand exposure both of these processes take place at the landscape level (Guariguata and Ostertag, 2001; Baker et al., 2005; Lugo, 2008). These hypotheses could be tested experimentally (Huston and Smith, 1987; Kraft et al., 2008; Comita et al., 2010) and by reconstructing stand history (Baker et al., 2005).
4.3. Management implications Previous studies using aerial photographs and satellite imagery showed that hurricane effects were more severe in broadleaf stands than in pine stands (Inafor, 2007; Rossi et al., 2013). Our data confirmed these observations but indicated that these studies overestimated hurricane effects in managed stands (Table 2). Pine resistance to wind was caused by a long history of land use that included logging, frequent surface fires, and grazing. The combined effects of these factors lowered stand density and eliminated young seedlings. However, because of its fast-growth and wide distribution, pine stands still offer various opportunities for management. Since seedlings of Pinus caribaea need three or four years to become resistant to surface fires (O’Brien et al., 2008), regeneration depends on seedlings escaping fire mortality (Myers et al., 2006; Veldman et al., 2015). Consequently, in remote locations or in stands specifically destined for timber production, recruitment can be favored by reducing fire frequency and grazing intensity. Since the land belonging to these indigenous communities comprises hundreds to thousands of hectares, the conflict between alternative land uses should be minimal. Grazing, timber production and agriculture could coexist in this landscape. Taken together, observed wind effects can be explained by marked differences in stand structure. While the dense and tall structure of broadleaf stands increased susceptibility to wind, the open and short structure of pines minimized it. These structural features were best captured by the height to diameter ratio which is perhaps the best indicator of tree and stand susceptibility to wind (Fig. 4; Supp. Fig. 2). Therefore, managers can increase wind resistance through early thinning in young stands or by removing trees with high ratios where appropriate (Mitchell, 1998; Wilson and Oliver, 2000). In sum, our results describe how strong wind coupled with selective logging and fires have shaped stand structure and illustrate potential effects of hurricane Felix on stand composition. Lastly, the findings of this stand-scale study highlight the importance of examining large disturbances across a range of spatial scales.
Acknowledgements This research was supported by a Prusser Fellowship from Clark University and by the Finfor project at CATIE, Costa Rica. The following people provided valuable support: Y. Ordóñez, W. Lau, and Canales J. Special thanks to D. Salazar, E. Cordon, W. Watler, F. Bascope and the late Rodolfo Cruz ‘‘Chaparro”, from Moss. Two anonymous reviewers provided thoughtful suggestions.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.foreco.2017.05. 034. References Alvarado, C., 2010. Manejo forestal: los primeros pasos de una comunidad indígena enfrentada a grandes dificultades. En Sabogal, C y Casaza, J. Casos ejemplares de manejo forestal sostenible en America latina y el Caribe. Junta de Castilla y Leon. FAO. ISBN: 978-92-5-306651-3. Baker, P.J., Bunyavejchewin, S., Oliver, C.D., Ashton, P.S., 2005. Disturbance history and historical stand dynamics of a seasonal tropical forest in Western Thailand. Ecol. Monog. 75 (3), 317–343. http://dx.doi.org/10.1890/04-0488. Beven, J.L., Avila, L., Blake, E.S., Brown, D.P., Franklin, J.L., Knabb, R.D., Pasch, R.J., Rhome, J.R., Stewart, S.R., 2008. Atlantic hurricane season of 2005. Mon. Weather Rev. 136 (3), 1109–1173. Boose, E.R., Serrano, M.I., Foster, D.R., 2004. Landscape and regional impacts of hurricanes in Puerto Rico. Ecol. Monog. 74 (2), 335–352. http://dx.doi.org/ 10.1890/02-4057. Boucher, D.H., Vandermeer, J., Yih, K., Zamora, N., 1990. Contrasting hurricane damage in tropical rain forest and pine forest. Ecology 71 (5), 2022–2024. http://dx.doi.org/10.2307/1937611. Boucher, D.H., Mallona, M.A., 1997. Recovery of the rain forest tree Vochysia ferruginea over 5 years following Hurricane Joan in Nicaragua: a preliminary population projection matrix. For. Ecol. Manag. 91, 195–204. Brennan, M.J., Knabb, R.D., Mainelli, M., Kimberlain, T.B., 2009. Atlantic Hurricane Seasonof 2007. Mon. Weather Rev. 137 (12), 4061–4088. Canham, C.D., Denslow, J., Platt, W.J., Runkle, J.R., Spies, T.A., White, P.S., 1990. Light regimes beneath closed canopies and tree-falls gaps in temperate and tropical forests. Can. J. For. Res. 20, 620–630. http://dx.doi.org/10.1139/x90-084. Castilleja, G., 1993. Changing trends in forest policy in Latin America: Chile, Nicaragua and Mexico. Unasylva 43(4), 175. (accessed 15 February 2014). Chambers, J.Q., Negron-Juarez, R.I., Marra, D.M., Di Vittorio, A., Tews, J., Roberts, D., Ribeiro, G., et al., 2013. The steady-state mosaic of disturbance and succession across an old-growth Central Amazon forest landscape. P. Natl. Acad. Sci. USA 110 (10), 3949–3954. http://dx.doi.org/10.1073/pnas.1202894110. Clark, D.A., Clark, D.B., 1992. Life history diversity of canopy and emergent trees in a Neotropical rain forest. Ecol. Monogr. 62 (3), 315–344 . Colwell, R.K., 2013. EstimateS: Statistical Estimation of Species Richness and Shared Species from Samples. Version 9 (accessed March 10-2015). Comita, L.S., Thompson, J., Uriarte, M., Jonckeheere, I., Canham, C.D., Zimmerman, J. K., 2010. Interactive effects of land use history and natural disturbance on seedling dynamics in a subtropical forest. Ecol. Appl. 20 (5), 1270–1284. Condit, R., Aguilar, S., Hernandez, A., Perez, R., Lao, S., Angehr, G., Hubbell, S.P., et al., 2004. Tropical forest dynamics across a rainfall gradient and the impact of an El Niño dry season. J. Trop. Ecol. 20, 51–72. http://dx.doi.org/10.1017/ S0266467403001081. Curtis, J.T., Cottam, G., 1962. Plant ecology workbook. Burgess, Minneapolis. De’ath, G., Fabricius, K.E., 2000. Classification and regression trees: a powerful yet simple technique for ecological data analysis. Ecology 81 (11), 3178–3192 . Denslow, J.S., 1987. Tropical rainforest gaps and tree species diversity. Ann. Rev. Ecol. Syst. 18, 431–451. http://dx.doi.org/10.1146/annurev. es.18.110187.002243. Denslow, J.S., 1995. Disturbance and diversity in tropical rain forests: the density effect. Ecol. App. 5 (4), 962–968 . Denslow, J.S., Ellison, A.E., Sanford, R.E., 1998. Tree-fall gap size effects on aboveand below-ground processes in tropical wet forests. J. Ecol. 86, 597–606 . Everham, E.M., Brokaw, V.L., 1996. Forest damage and recovery from catastrophic wind. Bot. Rev. 62, 113–185. http://dx.doi.org/10.1007/BF02857920. Foster, D.R., Boose, E.R., 1992. Patterns of forest damage resulting from catastrophic wind in Central New England, USA. J. Ecol. 80, 79–98 . Foster, D.R., Motzkin, G., Slater, B., 1998. Land use history as a long-term broad-scale disturbance: regional forest dynamics in central New England. Ecosystems 1, 96–119. http://dx.doi.org/10.1007/s100219900008. Granzow-de la Cerda, I., Zamora, N., Vandermeer, J., Boucher, D., 1997. Diversidad de especies arbóreas en el bosque tropical húmedo del Caribe Nicaragüense siete años después del huracán Juana. Revista Biología Tropical 45, 1409–1419 . Guariguata, M., Ostertag, R., 2001. Neotropical secondary forest succession: changes in structural and functional characteristics. For. Ecol. Manag. 148, 185–206. Gunter, S., Weber, M., Stimm, B., Mosandl, R., 2011. Silviculture in the Tropics. Springer, p. 558. ISBN 978-3-642-19986-8. Hoffmann, W.A., Geiger, E.L., Gotsch, S.G., et al., 2012. Ecological thresholds at the savanna-forest boundary: how plant traits, resources and fire govern the
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207 distribution of tropical biomes. Ecol. Lett. 15, 759–768. http://dx.doi.org/ 10.1111/j.1461-0248.2012.01789.x. Hubbell, S.P., 1979. Tree dispersion, abundance and diversity in a tropical dry forest. Science 203, 1299–1309. http://dx.doi.org/10.1126/science.203.4387.1299. Huston, M., 1979. A general hypothesis of species diversity. Am. Nat. 113, 81–101. Huston, M., Smith, T., 1987. Plant succession: life history and competition. Am. Nat. Am. Nat. 130 (2), 168–198 . Instituto Nacional Forestal – INAFOR, 2007. Evaluación de Daños al ecosistema forestal ocasionados por el huracán Félix. Gobierno Región Autónoma Atlántico Norte – RAAN Nicaragua. (accessed 22 April 2011). Instituto Nacional Forestal – INAFOR, 2009. Resultados del inventario nacional forestal (2007–2008) . Lugo, A.E., 2002. Can we manage tropical landscapes? An answer from the Caribbean perspective. Lan. Ecol. 17 (7), 601–615. http://dx.doi.org/10.1023/ A:1021419815480. Lugo, A.E., 2008. Visible and invisible effects of hurricanes on forest ecosystems: an international review. Austral Ecol. 33, 368–398. http://dx.doi.org/10.1111/ j.1442-9993.2008.01894.x. McSweeney, K., 2004. The dugout canoe trade in Central America’s Mosquitia: approaching rural livelihoods through systems of exchange. Ann. Assoc. Am. Geogr. 94 (3), 638–661. Mitchell, S.J., 1995. The windthrow triangle: a relative windthrow hazard assessment procedure for forest managers. Forest. Chron. 71 (4), 446–450. http://dx.doi.org/10.5558/tfc71446-4. Mitchell, S.J., 1998. A diagnostic framework for windthrow risk estimation. Forest. Chron. 74 (1), 100–105. http://dx.doi.org/10.5558/tfc74100-1. Montagnini, F., Jordan, C., 2005. Tropical Forest Ecology: The Basis for Conservation and Management. Springer, p. 306. ISBN: 3-540-23797-6. Myers, R., O’Brien, J., Morrison, S., 2006. Fire management overview of the Caribbean pine (Pinus caribaea) Savannas of the Mosquitia, Honduras. Global Fire Initiative (GFI) Technical Report 2006–1b. The Nature Conservancy, Arlington, Virginia, 30p. National Oceanic and Atmospheric Administration – NOAA, 2013. Historical Hurricane Tracks. . National Oceanic and Atmospheric Administration – NOAA, 2016. Historical Hurricane Tracks. (accessed, February 2011). Norberg, R.A., 1988. Theory of growth geometry of plants and self-thinning of plant populations: geometric similarity, elastic similarity, and different growth modes of plant parts. Am. Natur. 131 (2), 220–256 . O’Brien, J.J., Hiers, J.K., Callaham Jr., M.A., Mitchell, R.J., Jack, S.B., 2008. Interactions among overstorey structure, seedling life history traits, and fire in frequently burned neotropical pine. Forest. Ambio. 37 (7–8), 542–547. Oliver, C.D., Larson, B., 1996. Forest Stand Dynamics. Wiley, New York. ISBN13:978-0471138334. Oliver, C.D., Burkhardt, E.C., Skojac, D., 2005. The increasing scarcity of red oaks in the Mississippi River floodplain forests: influence of the residual overstory. For. Ecol. Manag. 210, 393–414. http://dx.doi.org/10.1016/j.foreco.2005.02.036. Parsons, J.J., 1955. The Miskito pine savanna of Nicaragua and Honduras. Ann. Assoc. Am. Geogr. 45 (1), 36–63.
207
Pacala, S.W., Canham, C.D., Silander, J.A., 1993. Forest models defined by field measurements: I. The design of a northeastern forest simulator. Can. J. For. Res. 23, 1980–1988. http://dx.doi.org/10.1139/x93-249. Pacala, S.W., Canham, C.D., Saponara, J., Silander, J.A., Kobe, R.K., Ribbens, E., 1996. Forest models defined by field measurements: II. Estimation, error analysis and dynamics. Ecol. Monogr. 66 (1), 1–43. Pérez-Flores, M.A., 2000. Composición y diversidad de los bosques de la región autónoma del Atlántico Norte nicaragüense: una base para el manejo sostenible. ‘‘Composition and diversity of the forests of The Autonomous Region of the Nicaraguan North Atlantic: a base for the sustainable management”. Master’s thesis. CATIE, Turrialba, Costa Rica. 155 p. Pérez-Flores, M.A., Finegan, B., Delgado, D., Louman, B., 2001. Composición y diversidad de los bosques de la región autónoma del Atlántico norte de Nicaragua: una base para el manejo sostenible. Revista forestal centroamericana 34 (Abr–Jun), 66–72.. Rejmanek, M., Brewer, S.W., 2001. Vegetative identification of tropical woody plants: state of the art and annotated bibliography. Biotropica 33 (2), 214–228 . Ratnam, J., Bond, W.J., Fensham, R.J., et al., 2011. When is a ‘forest’ a savanna, and why does it matter? Glob. Ecol. Biogeogr. 20 (5), 653–660. http://dx.doi.org/ 10.1111/j.1466-8238.2010.00634.x. Rossi, E., Rogan, J., Schneider, L., 2013. Mapping forest damage in northern Nicaragua after Hurricane Felix (2007) using MODIS enhanced vegetation index data. Gisci. Rem. Sens. 50, 385–399. Salazar, M.E., 2005. Propuesta para el desarrollo y manejo sostenible de los bosques naturales latifoliados en las regiones del Atlantico Norte, Atlantico sur y Rio san Juan. Ministerio agropecuario y forestal de Nicaragua (MAGFOR). 113 p. Snook, L.K., 2003. Regeneration, growth, and sustainability of mahogany in México’s Yucatán forests. In: Lugo, A., Figueroa Colón, J.C., Alayón, M. (Eds.), Big-Leaf Mahogany: Genetics, Ecology, and Management. Ecological Studies, vol. 159, pp. 169–192. ISBN 978-0-387-21778-9. Smith, J.H., 2003. Land-cover assessment of conservation and buffer zones in the BOSAWAS natural resource reserve of Nicaragua. Environ. Manage. 31 (2), 252– 262. Sousa, W.P., 1984. The role of disturbance in natural communities. Ann. Rev. Ecol. Sys. 15, 353–391. http://dx.doi.org/10.1146/annurev.es.15.110184.002033. Stevens, W.D., 2001. Flora de Nicaragua, vol. 1–2. Missouri Botanical Garden Press. 966 p. ISBN 0915279959, 9780915279951. . Turner, M.G., 2010. Disturbances and landscape dynamics in a changing world. Ecology 91 (10), 2833–2849. http://dx.doi.org/10.1890/10-0097.1. Van Bloem, S.J., Murphy, P.G., Lugo, A.E., Ostertag, R., Costa, R., Bernard, I.R., 2005. The influence of hurricane winds on Caribbean dry forest structure and nutrient pools. Biotropica 37 (4), 571–583 . Van Bloem, S.J., Murphy, P.G., Lugo, A.E., 2007. A link between hurricane-induced tree sprouting, high stem density and short canopy in tropical dry forest. Tree Phys. 27 (3), 475–480. http://dx.doi.org/10.1093/treephys/27.3.475. Vandermeer, J., Granzow-de la Cerda, I., Boucher, D., Perfecto, I., Ruiz, J., 2000. Hurricane disturbance and tropical tree species diversity. Science 788 (290), 788–790. http://dx.doi.org/10.1126/science.290.5492.788. Vandermeer, J.H., Boucher, D.H., Granzow-de la Cerda, I., Perfecto, I., 2001. Growth and development of the thinning canopy in a post-hurricane tropical rain forest in Nicaragua. For. Ecol. Manag. 148 (1–3), 221–242. http://dx.doi.org/10.1016/ S0378-1127(00)00538-7. Veblen, T., Kulakowski, D., Eisenhart, K.S., Baker, W.L., 2001. Subalpine forest damage from a severe windstorm in northern Colorado. Can. J. For. Res. 31, 2089–2097. Veldman, J.W., Buisson, E., Durigan, G., et al., 2015. Toward an old-growth concept for grasslands, savannas, and woodlands. Front. Ecol. Environ. 13 (3), 154–162. http://dx.doi.org/10.1890/140270. Vreugdenhil, D., Meerman, A., Meyrat, L., Gomez, D., Graham, D.J., 2002. Map of the ecosystems of Central America: final report. In World Bank. Washington, DC. Accessed September, 2012, 65 p. (accessed 15 November 2014). Warner, R.R., Chesson, P.L., 1985. Coexistence mediated by recruitment fluctuations: a field guide to the storage effect. Amer. Nat. 125, 768–787 . West, G.B., Enquist, B.J., Brown, J.H., 2009. A general quantitative theory of forest structure and dynamics. Proc. Nat. Acad. Sci. USA 106 (17), 7040–7045. http:// dx.doi.org/10.1073/pnas.0812294106. Wilson, J.S., Oliver, C.D., 2000. Stability and density management in Douglas-fir plantations. Can. J. For. Res. 30, 910–920. http://dx.doi.org/10.1139/x00-027. Yih, K., Boucher, D.H., Vandermeer, J.H., Zamora, N., 1991. Recovery of the rain forest of Southeastern Nicaragua after destruction by hurrican Joan. Biotropica 23 (2), 106–113 . Zar, J.H., 1999. Biostatistical analysis. Prentice-Hall, New Jersey. 663 p.. Zimmerman, J.K., Everham, E.D. III., Waide, R.B., Lodge, D.J., Taylor, C.M., Brokaw, N. V.L., 1994. Responses of tree species to hurricane winds in subtropical wet forest in Puerto Rico: implications for tropical tree life histories. J. Ecol. 82(4), 911–922. .
Contents lists available at ScienceDirect
Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco
Wind effects and regeneration in broadleaf and pine stands after hurricane Felix (2007) in Northern Nicaragua E. Rossi a,b,⇑, I. Granzow-de la Cerda c, C.D. Oliver d, D. Kulakowski b,e a
Instituto de Bioética, Pontificia Universidad Javeriana, Bogotá, Colombia Graduate School of Geography, Clark University, Worcester, United States c CREAF and Univ. Autónoma de Barcelona, Cerdanyola del Vallés 08193, Spain d Yale School of Forestry and Environmental Studies, United States e WSL Institute for Snow and Avalanche Research, SLF, Davos, Switzerland b
a r t i c l e
i n f o
Article history: Received 27 October 2016 Received in revised form 18 May 2017 Accepted 21 May 2017
Keywords: Stand dynamics Wind disturbances Tropical forestry Forest management Región autónoma del Caribe Norte – RACN
a b s t r a c t Large-scale wind disturbances shape forest structure and composition and leave long lasting legacies. Adequate understanding of the role of disturbances in tropical stand dynamics is necessary to guide management efforts. In this study, we used field data to characterize the effects of a major hurricane in broadleaf and pine stands in Northern Nicaragua. First, we described tree and stand attributes associated with observed structural effects: branch loss, snapping and uprooting. Secondly, to assess the potential influence of hurricane Felix on stand composition, we characterized two key life-history traits: regeneration through resprouting and shade tolerance. Findings indicated that tree attributes such as diameter at breast height (dbh) and height to diameter ratio (hdr), were strongly associated with the type and magnitude of wind effects. All trees >70 cm dbh exhibited visible effects and trees taller than 14 m were more likely to be uprooted (7.3% vs. 0.8% of total). Results confirmed that Felix caused significant structural effects in broadleaf stands and mild effects in pine stands. Abundant post-hurricane resprouting was observed in both shade-tolerant and shade-intolerant species but was absent in pines. Among canopy trees we found eleven shade-intolerant species that exhibited abundant resprouting. These species could become dominant in the next decade. Our findings illustrate the role of wind disturbances on tropical stand dynamics at different spatial and temporal scales. Ó 2017 Elsevier B.V. All rights reserved.
1. Introduction Disturbance events and competitive interactions among trees drive forest dynamics (Sousa, 1984; Foster et al., 1998; West et al., 2009; Turner, 2010). Disturbances change the physical environment, reallocate resources, and create biological legacies that influence ecosystem processes (Norberg, 1988; Foster and Boose, 1992; Pacala et al., 1993; Kulakowski and Veblen, 2003; Boose et al., 2004). Disturbances may also be important in maintaining species diversity, particularly in species-rich tropical ecosystems (Huston, 1979; Warner and Chesson, 1985; Denslow, 1995; Granzow-de la Cerda et al., 1997). Although the factors that determine forest susceptibility to wind disturbances have been characterized in temperate ecosystems, they have received limited attention in tropical forests (e.g., Foster and Boose, 1992; ⇑ Corresponding author at: Instituto de Bioética, Pontificia Universidad Javeriana, Bogotá, Colombia E-mail address: [email protected] (E. Rossi). http://dx.doi.org/10.1016/j.foreco.2017.05.034 0378-1127/Ó 2017 Elsevier B.V. All rights reserved.
Boucher et al., 1990; Lugo and Scatena, 1996). Specifically, it is not well understood how tree-and-stand attributes influence susceptibility to wind and subsequent stand dynamics (Lugo, 2002). A few general relationships between wind effects and tree-andstand attributes have been identified. First, some studies have found a positive relationship between increasing tree height and increased uprooting (Everham and Brokaw, 1996). Second, stand susceptibility to wind is linked to the height to diameter ratio (hdr) of dominant trees. This ratio depends on light availability and stand density: trees growing in full sunlight tend to have lower values, while trees growing in shaded conditions – e.g., in dense stands-tend to have greater values (Mitchell, 1995). For this reason foresters use it as a quantitative indicator to guide thinning operations (Wilson and Oliver, 2000). The importance of forest type (i.e., broadleaf vs. pine) on the severity of wind effects has been examined in numerous locations with different results (Everham and Brokaw, 1996). Although conifers are sometimes more susceptible than angiosperms, given the greater diversity in wood density and tree architecture of the
200
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
latter, wind effects seem to be more dependent on stand characteristics such as tree density and canopy height, rather than species’ traits (Boucher et al., 1990; Mitchell, 1998). By snapping and uprooting trees differentially, windstorms also shape stand structure (Zimmerman et al., 1994; Vandermeer et al., 2000). For example, in Puerto Rico windstorms occur every decade and thus favor the development of dense and homogenous stands with low regeneration and high competition (Boose et al., 2004; Van Bloem et al., 2007). In contrast, in Nicaragua, windstorms occur approximately once every century and create space for the establishment of surviving and new seedlings (Vandermeer et al., 2001). Most studies of tropical forest dynamics have focused on shortterm and small-scale disturbances, i.e., gap dynamics (Hubbell, 1979; Denslow, 1987; Denslow et al., 1998; Condit et al., 2004), in part because of the difficulties in determining tree age using growth rings and identifying sterile individuals (Rejmanek and Brewer, 2001; Chambers et al., 2013). Consequently, the role of infrequent hurricanes in shaping forest composition and diversity is less understood (Nelson et al., 1994; Zimmerman et al., 1994; Baker et al., 2005). By opening space over large areas (>1000 km2), hurricanes create landscape heterogeneity, change the ecological space available to organisms and favor tree recruitment (Huston, 1979; Denslow, 1995; Negrón-Juárez et al., 2010). For example, Vandermeer et al. (2000) documented an increase in tree species richness after the passage of hurricane Joan in 1988, which was attributed to the growth of the seedling bank coupled with resprouting of affected trees. Their 10-year data series indicated that following geographically large disturbances (5000 km2), dispersal limitations prevented pioneer species (e.g., Cecropia spp.; Ochroma spp.) from rapidly colonizing available space, in contrast to small gaps (0.5 km2). Lastly, studies in Puerto Rico have showed that hurricanes can also shape the composition of the seedling layer (Comita et al., 2010). It has been proposed that tree response to wind disturbances depends on three main life-history traits: resprouting capacity, dispersal ability and shade tolerance (Canham et al., 1990; Oliver and Larson, 1996; Montagnini and Jordan, 2005). While post-hurricane resprouting and dispersal ability have been characterized in some Caribbean forests (Boucher and Mallona, 1997; Van Bloem et al., 2005), it is unknown if tropical canopies are dominated by shade tolerant or shade intolerant species, or a combination of the two. This information is important to understand historical stand dynamics and develop adequate management strategies (see Baker et al., 2005). By compiling empirical data from numerous studies, Gunter et al. (2011) defined three shade classes: tolerant, intolerant, and partially intolerant, which were used in this study to classify canopy species. The landfall of hurricane Felix provided us with the opportunity to characterize the factors that control wind effects and regeneration, specifically: (i) how do tree and stand attributes determine the type and magnitude of wind effects; (ii) what is the relationship between forest type (broadleaf vs. pine) and wind effects; (iii) which canopy tree species resprout following wind disturbances; and (iv) what is the proportion of canopy species that are shade tolerant. By using field data to characterize wind effects at the stand scale, this study complements previous efforts that used aerial surveys and satellite imagery to characterize wind effects at the landscape scale (Inafor, 2007; Rossi et al., 2013).
annual temperature is 25 °C and annual rainfall 2500 mm, with a marked dry season from February to early May. The RACN harbors 1.4 million hectares of forest (43% of Nicaragua’s forests), which include the Bosawás UNESCO’s Biosphere Reserve (Inafor, 2009). Two forest ecosystems cover most the region: dense, closed tropical broadleaf forests located towards the interior and pine savannas on coastal areas. Broadleaf (hardwood) forests are found mostly in crumbstructured humic clays in areas of undulating topography (Parsons, 1955). For centuries these forests have attracted foreign logging companies in search of mahogany (Switenia macrophylla) and Spanish cedar (Cedrela odorata) which were selectively harvested along rivers and roads. Yet, up until hurricane Felix’ impact, a significant portion of these forests remained inaccessible and poorly studied. Presently, numerous valuable timber species can be found in broadleaf stands, including Calophyllum brasiliense, Carapa guianensis, Dialium guianense, Guaiacum sanctum, Hyeronima alchorneoides, Hymenaea courbaril and Terminalia amazonia (Smith, 2003; Inafor, 2009). Over the last three decades, land-use practices consisted on selective logging and fuelwood collection, although a large portion of the forests of the region remains unmanaged (Salazar, 2005; Diaz, F. personal communication, 2010). Because of the high density and moisture, these broadleaf stands rarely burn (Inafor, 2009). Pine savannas dominated by Pinus caribaea comprise approximately 10,000 km2 along the Atlantic coast of Nicaragua. The soils of the pine savannas are mostly composed of nutrient poor sands, highly weathered quartz-gravels and clays in poorly drained areas (Parsons, 1955; Myers et al., 2006). This ecosystem harbors numerous herbaceous species and depends on recurrent fires that create adequate conditions for pine regeneration and prevent hardwood establishment (O’Brien et al., 2008; Ratnam et al., 2011; Hoffmann et al., 2012). Due to the abundance and accessibility of pines across the coastal plains of the region, logging and fuelwood collection have been common for over a century (Castilleja, 1993; McSweeney, 2004). In addition, near communities, pine stands are often burned to promote grass growth for cattle and deer (Alvarado, 2010). On 4 September 2007, hurricane Felix made landfall near the community of Awastara in Northern Nicaragua (14°180 44.0700 N, 8 3°120 12.0400 W), as a Saffir-Simpson category five hurricane (NOAA, 2013). The cyclone travelled across the region and became a low pressure system over northern Honduras early on September 5th (Brennan et al., 2009; Fig. 1). Felix caused substantial damage to infrastructure in coastal communities and affected more than 3000 km2 of broadleaf and pine stands. Estimates based on aerial surveys and satellite imagery suggest that one million m3 of tropical hardwoods were blown down by this hurricane (Inafor, 2007). Weather records indicate that no major storms had affected Northern Nicaragua since the late 1890s (Beven et al., 2008; Kar, 2010; NOAA, 2013). Similarly, historical data indicates that although numerous tropical storms regularly pass over Northern Nicaragua, most Atlantic hurricanes have not made landfall in this region (NOAA, 2016). Taken together, historical and weather records suggest that broadleaf stands located in the center of the RACN had remained undisturbed for at least a hundred years before Felix’s impact.
1.1. Study area
2.1. Data collection
The Autonomous Region of the Northern Caribbean of Nicaragua (hereafter RACN) comprises 32,000 km2 and hosts a population of 315,000 (Nicaraguan Census Bureau – INIDE, 2005). Average
Four field sites were defined using land cover maps based on proximity to the hurricane path and availability of large forest patches of broadleaf and pine stands to allow for comparisons
2. Methods
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
201
Fig. 1. (A) Land cover of the autonomous region of the northern Caribbean depicting the path of hurricane Felix and the distribution of broadleaf stands (dark green) and pine savannas (light blue). While pine savannas are accessible, broadleaf stands are mostly located in road less areas. (B) Study area and field site locations around which sampling plots were established. Site names from top to bottom: Miguel Bikan, Santa Clara, Awas Tigni and Moss. Timber extraction areas appear in orange, locations that experienced significant structural effects appear in grey (75% trees uprooted or snapped). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
between the two forest types (Vreugdenhil et al., 2002; Inafor, 2007). The four sites were named after the nearest indigenous community: Moss, Awas tigni, Miguel Bikan and Santa Clara. Sampling was conducted between June and August of 2011 in the Moss site, during 2012 in Awas Tigni and Santa Clara and Miguel Bikan in 2013. Distance between sites ranged from 8 to 21 km (Fig. 1). In each site, we randomly established 6–8 sampling plots separated by over 200 m and located at least 100 m away from the edge of the patch to avoid border effects. Combining the four sites; fifteen plots were established in broadleaf stands and fifteen plots in pine stands for a total of 30 plots (Fig. 1b). Each plot had a size of one hectare (100 100 m). Large plot sizes were selected to control for felled trees rooted outside the plot. Large trees falling unto small plots can affect all trees in the plot (e.g., 60% of trees uprooted, 40% snapped) causing an overestimation of wind effects at the stand scale. All plots were located within 30 km from the hurricane path and were representative of site conditions (elevation, slope, and aspect). Height, diameter at breast height (dbh) and taxonomic identity were recorded for all standing and fallen trees >10 cm dbh. Tree height was measured using a telescopic pole for individuals less than 15 m high and a hypsometer (Nikon 550) for all individuals taller than15 m. Trees were identified initially in the field and verified in the herbarium using the taxonomic criteria of Flora de Nicaragua (Stevens, 2001). For all trees >10 cm dbh, wind effects were recorded as: uprooted, snapped, lost branches, or no visible effects. Tree condition after the hurricane (alive or dead), resprout number, height and position on the stem were also recorded. A tree presented ‘‘branch loss” when branches larger than 10 cm in diameter were broken, or when the trunk exhibited scars or broken branches. Branch loss was defined in this way to capture changes that were visible years after the event, but did not capture the loss of branches smaller than 10 cm in diameter. Observations regarding the presence of tree mounds, hardwood stumps and fire scars were recorded for each stand. Information on the type and intensity of land use i.e., the presence of logging, grazing and fuel-
wood collection was gathered by interviewing community members. 2.2. Data analyses Data analyses are organized following the research questions. First, to determine how tree and stand attributes influenced the type and magnitude of wind effects we first compared the number of canopy (overstory) trees that experienced structural effects (branch loss, snapped, and uprooted). Mean overstory height was calculated for each stand following Oliver et al. (2005); to avoid confounding size effects, comparisons were conducted separately for trees 25 cm dbh and >25 cm dbh, which roughly segregates understory from overstory trees in broadleaf stands. Previous studies and field observations indicated that most broadleaf trees with diameters >25 cm were part of the overstory (Pérez-Flores, 2000). Whereas in pine stands trees 10 cm dbh were regarded as understory and trees >10 cm dbh were considered as part of the overstory. Subsequently, chi-square analysis of 2 2 contingency tables was used to test for differences in the frequencies of different wind effects (e.g., uprooted vs. snapped). Tests were not conducted if the expected frequency in any cell of the table was less than five. Unequal variance t-tests were conducted to compare the means of trees by size class. The influence of height on susceptibility to wind effects was examined by comparing median tree height within each wind effect class. Statistical significance of differences in the median heights among classes was evaluated with one-way analysis of variance; pairwise comparisons were made with Dunn’s method (Zar, 1999). Hurricane induced mortality was defined as the sum of uprooted and snapped non-resprouting trees. Second, to characterize the relationship between forest type (broadleaf vs. pine) and wind effects cover types were distinguished based on the abundance and dominance of Pinus caribaea and Dialium guianensis, in conifer and broadleaf stands, respectively (Pérez-Flores, 2001). The following four cover types were defined: (1) open pine forest (<400 trees/ha), (2) dense pine forest
202
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
(>400 trees/ha), (3) stands dominated by D. guianensis (>10% of stand’s basal area), and (4) mixed broadleaf. Differences in susceptibility to wind among cover types were examined by comparing mean percentage of affected trees in each class with a Kruskal– Wallis one-way analysis of variance; pairwise comparisons were made with Dunn’s method (Zar, 1999). Best-subset regression was used to identify stand attributes that could predict wind effects using the SigmaStat package (Jandel Corp. 1995). The dependent variable was the percentage of trees that were uprooted, snapped or lost branches; independent variables included stand density, mean height of the largest 30 trees, total basal area and distance from the hurricane path. All variables passed tests of normality and homogeneity of variance. Classification tree analysis (CTA) was used to examine the relationship between tree-and-stand attributes and wind effects. This method is well suited for the analysis of ecological data due to its ability to examine both numerical and categorical variables and non-linear relationships (De’ath and Fabricius, 2000). Classification trees are fitted using a binary recursive partitioning whereby data are successively split into more homogeneous groups using combinations of variables. We obtained CTs for broadleaf and pine trees and separately for each forest type using the CRT growing method and the Gini splitting rule in the SPSS statistical package. Species richness and similarity indices were calculated using EstimateS software (Version 9.1, Colwell, 2013). For every species the Importance Value Index (IVI) was calculated to identify dominant canopy species. The IVI is the sum of relative density, relative dominance, and relative frequency and ranges from 0 to 300, where 300 is the maximum value (Curtis and Cottam, 1962). Lastly, data on tree condition after the hurricane and resprouting were compared by species. Specifically, we aimed to determine which species exhibited resprouting capacity, whether resprouting was influenced by snapping or uprooting, and which species were tolerant to shade, intolerant or partially intolerant. This information can be useful to determine if hurricane Felix will cause changes in canopy composition in the next few decades.
Fig. 2. Percentage of wind affected trees sorted by dbh size class in both forest types. Trees >50 cm dbh were more likely to experience branch loss (p < 0.001; unpaired t-test with unequal variances). All trees >70 cm dbh exhibited wind effects.
of trees in the understory was too small for statistical analyses (Supplementary Fig. 1). 3.2. Stand attributes and wind effects
3. Results 3.1. Tree attributes and wind effects Among the 10,461 trees recorded, 652 had been uprooted, 341 snapped, 468 lost branches, and 9000 did not exhibit visible effects. Defoliation was not reported because sampling was carried out years after the event. Tree size was positively correlated with wind effects: most trees (90%) larger than 70 cm (dbh) exhibited visible effects, while trees larger than 80 cm dbh were more likely to be uprooted (p < 0.001; unpaired t-test with unequal variances, Fig. 2). Wind effects were also positively correlated to tree height: the number of uprooted trees increased abruptly for individuals taller than 16 m (p < 0.01; unpaired t-test with unequal variances, Fig. 3). Similarly, differences in median height were found among all class pairs (P < 0.001) except uprooted vs. branch loss. In contrast, the number of snapped trees was not correlated to dbh or height. Susceptibility to wind was also influenced by tree position in the canopy (See supplementary Fig. 1). In broadleaf stands, the overstory experienced significantly larger effects than the understory (X2 = 872, p < 0.05): overstory trees exhibited 18.9% branch loss, 11.1% snapping and 20.9% uprooting; while the figures for understory trees were 2%, 4% and 7% respectively. Wind effects were less severe in pine stands: among overstory trees 1.29% lost branches, 0.19% snapped and 0.93% were uprooted; the number
There were contrasting differences in stand density and tree size between broadleaf and pine stands that shaped wind effects. Although pine stands 28, 29 and 30 presented over 700 trees/ha, overall, pine stands were less dense, shorter and exhibited lower hdr (Table 1). The low density and ‘‘open structure” of pine stands reflects a history of logging, coupled with grazing and fire (Table 1). The presence of stumps and fire scars in numerous stands indicated that sampled stands had been managed. Although field data were insufficient to precisely describe land use, field observations and local people indicated that management intensity was higher in pine stands. Regression analyses indicated that maximum canopy height (height of the largest 30 trees in the stand) predicted wind effects in both forest types (p < 0.001; adjusted r2 = 0.52) (Table 1). Cover types also exhibited significant differences in wind effects (Table 2). While mixed broadleaf forests exhibited the highest percentage of snapping and uprooting (7.1% and 13.3% respectively) stands dominated by D. guianense showed higher percentage of branch loss (9.2%). In contrast, pine stands were less affected than broadleaf stands: 2.5% branch loss; 0.5% snapped; 0.2% uprooted (X2 = 1434; p < 0.001). Classification trees, on the other hand, indicated that forest type and the height to diameter ratio (hdr) explained most of the differences among sampled trees; higher hdr indicated larger effects (See Table 1; Supp. Fig. 2). The CT model obtained accurately classified 85.8% of the observations.
203
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
3.3. Stand diversity and composition In total, 53 species were identified. One species could not be identified beyond family level and five could not be identified beyond genus and were thus assigned to distinct morphotypes. Species richness varied widely between forest types: while broadleaf stands contained up to 65 species in a 1-ha plot, only 11 species occurred in the most species-rich 1-ha plot in pine stands (Table 1). Broadleaf stands were dominated by Dialium guianense, Spondias mombin, Terminalia amazonia and Ficus aff. yoponensis, (ranked by BA) and harbored numerous commercially valuable species including Hyeronima alchorneoides, Calophyllum brasiliense, Carapa guianensis, Swietenia macrophylla and Dipteryx oleifera (Supp. Table 1). Uprooting was frequent for Dialium guianense (40%, 72 individuals) as for Calophyllum brasiliense (33%, 4 individuals). Spondias mombin sustained the highest rate of snapping (22%, 37 individuals), and Ficus aff. yoponensis the highest rate of branch loss (33%, 16 individuals and also high mortality 45.8%). Overall, tree mortality varied across species, ranging from 4.3% in P. caribaea, to 39.1% in Terminalia amazonia (Supp. Table 1). 4. Discussion 4.1. Wind effects
Fig. 3. Percentage of wind affected trees sorted by height class (labels correspond to the upper limit of 2 m classes). Branch loss and uprooting was more frequent among trees >14 m (p < 0.01; unpaired t-test with unequal variances). The number of unaffected trees decreased after 14 m (p < 0.001; unpaired t-test with unequal variances). Median height differed significantly among all classes except uprooted vs. branch loss (P < 0.001, Kruskal-Wallis one way ANOVA).
Findings indicated that wind effects were influenced by tree size: larger DBH and height increased susceptibility to uprooting and branch loss. Whereas subcanopy light-wooded pioneers such as Cecropia insignis, Ochroma aff. lagopus and Hibiscus tiliaceous exhibited low uprooting and mortality (<5% trees), large hardwood species including Dialium guianense, Spondias mombin and Terminalia amazonia experienced higher levels of uprooting, snapping and branch loss. These results suggest that wind effects depend more on tree size and canopy position than species attributes (>12%; Supp. Table 1). The relationship between size and susceptibility
Table 1 Characteristics of sampled stands in Northern Nicaragua.
a
Plot #
Forest type
Land use
Stems/(ha)
Mean tree height (30 trees)
HDR Over.
HDR Under.
BA/ha (m2)
Species richness/ha
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl Bl P P P P P P P P P P P P P P P
SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW SL, FW L, G L, G L, G L, G L, G L, G L, G L L L L L, F L, F L, F L, F
402 391 287 333 276 283 264 364 283 292 311 236 623 311 355 45 93 83 45 63 34 39 411 607 589 472 731 932 827 783
20 20 19 23 22 24 32 28 27 38 45 32 21 22 22 11.2a 18.5a 21.1a 17.1a 16.8a 14.6a 16.4a 17 16 15 14 15 15 16 14
55.1 56.2 63.7 55.6 51.2 54.0 59.9 65.4 66.2 66.3 71.7 67.8 40.1 41.0 46.0 47.9 49.1 52.3 41.5 47.7 36.3 39.5 17.5 17.7 17.3 17.5 17.0 17.0 17.0 16.9
85.5 94.4 86.6 101.1 87.6 93.6 98.1 88.4 92.3 94.8 109.8 95.3 93.4 96.0 101.4 60.0 62.8 61.3 63.5 71.6 59.7 58.4 17.0 26.2 25.5 25.6 24.1 75.7 21.0 19.7
13.3 13.7 8.1 16.5 15.3 17.2 15.6 17.3 13.0 31.4 28.9 15.7 32.2 30.1 24.7 0.7 4.3 4.2 2.5 2.3 1.7 2.2 159.1 228.1 171.6 130.3 217.2 257.0 246.0 175.6
54 57 51 42 42 30 42 40 34 44 38 37 65 47 53 1 11 1 1 1 5 6 2 3 2 2 4 5 4 3
The 10 tallest trees were sampled in open stands. Bl-broadleaf, P-pine. (L-logging; SL-selective logging; FW-fuelwood; G-grazing; F-fire. HDR height to diameter ratio).
204
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
Table 2 Percent basal area and wind effects for the most common species in the forest cover types. Selected wind effects are: BL – branch loss; SP – Snapped; UP – Uprooted. Values indicate the percentage of the total basal area covered by each species in the stands where it occurs, 100% corresponds to the total basal area of the stand. Wind effects correspond to the percent of trees in those stands that exhibited BL, SP or UP. Stand basal area and number of trees can be found in Table 1. Cover type
(1) (2) (3) (4)
Mixed broadleafa Dominated by Dialium Open pine Dense pine
No. of stands
9 6 7 8
Range of basal area (%)
Wind effects (%)
Dialium
Pinus
Byrsonima
Terminalia
Ficus
Spondias
BL
SP
UR
0–8.2 11.3–22 0 0
0 0 90.5–100 95.1–97.3
0.2–12.3 0–0.1 0 1.9–4. 9
0.9–24.4 0.5–7.7 0 0
0–30.6 0–10.2 0 0
0–26.5 0–5.2 0 0
7.3% 9.2% 2.5% 1.1%
7.1% 5.5% 0.5% 0.2%
13.3% 9.5% 0.2% 0.9%
to wind has been attributed to a ‘‘sail effect”, as large canopies present a greater surface to the wind (Yih et al., 1991; Foster and Boose, 1992). Overall, the height to diameter ratio best captured this ‘‘sail effect” and was perhaps the single most important variable for predicting wind effects (Fig. 4). This ratio depends on light availability because growing trees allocate photosynthates to primary growth (height), before secondary growth (diameter) (Mitchell, 1998; Wilson and Oliver, 2000). Light availability, in turn, is heavily dependent on stand density. Because of this, broadleaf trees in dense stands exhibited higher hdr than pines in open stands. On the other hand, tree susceptibility to snapping versus uprooting, could not be explained by measured attributes, as it might have been determined by a combination of microsite conditions and stochastic factors (Veblen et al., 2001). Although the role of tree architecture on wind effects was beyond the scope of this study, our data suggests that tree architecture was less important for two reasons. First, observations were consistent at the stand level, i.e., canopy trees in the same stand experienced similar effects. Second, crown size was more dependent on growing conditions than on species architecture (Supp. Fig. 2). For example, emergent trees such as Ficus aff. yoponensis were probably uprooted because of their large canopy size rather than crown architecture. The large size (dbh) and dominant canopy position of these trees can be a consequence of selective logging in the area. In the past, loggers probably avoided them because the abundant latex made them difficult to cut. In contrast, very large individuals from other species were extremely rare, suggesting that they were harvested for various purposes including fuelwood. Lastly, the relationship between tree size and wind susceptibility is consistent with observations in other locations (Mitchell, 1995; Foster and Boose, 1992). It is likely that the time since the last disturbance also shaped observed wind effects. Since broadleaf stands had not experienced
Fig. 4. Height to diameter ratios (hdr) among trees affected by hurricane Felix (2007). (1) Broadleaf; (2) Pines. Affected trees were snapped, uprooted or exhibited branch loss.
major hurricanes for at least a hundred years (Kar, 2010; Brennan et al., 2009; NOAA, 2016) and access difficulties prevented logging, especially in the central and northwestern part of the region, it was expected that broadleaf stands would experience larger wind effects (See timber extraction areas in Fig. 1). Pine savannas, in contrast, experienced a different disturbance regime characterized by frequent fires and complemented by logging and grazing (Myers et al., 2006); as a consequence, wind effects were less severe.
4.2. Regeneration and shade tolerance Characterizing life history traits such as tree regeneration and shade tolerance is relevant to understand how hurricane Felix could shape stand composition in the future. As documented in Caribbean forests, we found that tree regeneration through resprouting was common among broadleaf species and absent among pines (Boucher et al., 1990; Vandermer et al., 2000; Van Bloem et al., 2005). Vigorous resprouting was observed in 38 broadleaf species out of 54 and in 264 individuals, but was absent in pine stands. Snapped broadleaf trees were more likely to resprout than uprooted trees (52% vs. 16%) (Fig. 5).
Fig. 5. Observed resprouting in snapped and uprooted trees. Top: number of resprouting trees (n) by structural effect class. Bottom: Percentage of trees exhibiting resprouting by structural effect class. Note that the percentage of trees snapped and uprooted was calculated from the total number of affected trees (n = 1768). In contrast, the % of uprooted resprouting and % snapped resprouting was calculated from the number of trees per class (i.e. snapped: 341, snapped resprouting: 178), (uprooted: 652, uprooted resprouting: 107). The bottom part of the figure illustrates that snapped trees are much more likely to resprout than uprooted trees.
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
Snapped trees were much more likely to resprout than uprooted trees, presumably because the former maintain an intact root system; yet resprouting also occurred among some fallen trees. Since species with resprouting capacity can rapidly occupy available growing space it is possible that these species will become dominant in the next few decades (See Oliver and Larson, 1996). In addition regeneration strategies, tree responses to wind also depend on shade tolerance and maximum potential height growth (Clark and Clark, 1992; Oliver et al., 2005; Gunter et al., 2011). While shade intolerant species grow rapidly under full sun, shade tolerant species can grow slowly in the understory (Pacala et al., 1996). Our classification of shade tolerance indicated that 17 species could be regarded as shade intolerant, 12 as shade tolerant and 14 as intermediate (Clark and Clark, 1992; Gunter et al., 2011). Among shade intolerant species, Spondias mombin, Terminalia ama-
205
zonia and Hyeronima alchorneoides, exhibited abundant resprouting (Table 3). Based on our observations of regeneration and shade tolerance among canopy trees we propose two hypotheses regarding the role of hurricanes in the dynamics of these stands. Although these hypotheses remain speculative they were included to inform future research. First, species with resprouting capacity and fast growth are likely to become dominant after large blowdowns and subsequently exclude other tree species. In this way, hurricanes could favor the establishment of shade intolerant species. This has been observed in Southern Mexico, where the recruitment of Swietenia macrophylla seems to be favored by large windstorms and subsequent fires (Negreros-Castillo et al., 2003; Snook, 2003). On the other hand, in our study area six stands out of fifteen were dominated by shade tolerant species, particularly Dialium
Table 3 Autoecology of dominant tree species. Canopy position: emergent (E), canopy (C), understory (U).
b
#
Species name
Resprouting (# trees)a
Canopy position
Autoecology (shade tolerance)b
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
Pinus caribaea Byrsonima crassifolia Cecropia insignis Dialium guianense Spondias mombin Inga densiflora Terminalia amazonia Homalium racemosum Muntingia calabura Castilla elastica Ochroma lagopus Bursera simaruba Hibiscus tiliaceus Miconia dodecandra Hirtella triandra Brosimum alicastrum Vochysia ferruginea Rheedia sp. Tabernaemontana sp. Tetragastris panamensis Luehea seemannii Virola koschnyi Zuelania guidonia Albizia adinocephala Hyeronima alchorneoides Ficus aff. yoponensis Apeiba tibourbou Lacmellea panamensis Psychotria chiapensis Schizolobium parahybum Morinda panamensis Lysilóma auritum Coccoloba belizensis Symphonia globulifera Calophyllum brasiliense Cupania cinerea Malvaceae sp1. Carapa guianensis Dendropanax arboreus Grias cauliflora Vochysia guatemalensis Oreopanax capitatus Pouteria sapota Lonchocarpus phaseolifolius Guazuma ulmifolia Pouteria campechiana Ehretia austin-smithii Ceiba pentandra Croton panamensis Ochroma sp. Swietenia macrophylla Didymopanax morototoni Dipteryx panamensis Cordia bicolor
0 3 2 32 32 11 27 9 1 16 0 13 0 1 18 4 0 1 0 16 15 2 9 6 5 0 1 1 1 2 1 1 5 10 2 3 0 2 2 0 0 0 0 1 2 0 0 0 0 1 1 0 3 2
C U U E, C C C C U C, U C, U U C, U C, U U C, U C, U C, U C, U C, U C, U C, U C, U C, U C, U E, C E, C C, U C, U U C C, U C, U C, U C, U E, C C, U C, U C C, U C, U C, U C, U C, U C, U C, U C, U U E, C C, U U C, U U E, C C, U
Intolerant Intolerant Intolerant (pioneer) Tolerant Intolerant Shade intermediate Intolerant Shade intermediate Unknown Unknown Intolerant Intolerant Shade intermediate Toleranta Toleranta Shade intermediate Shade intermediate Unknown Unknown Intolerant Unknown Shade intermediate Unknown Unknown Intolerant Tolerant Unknown Unknown Tolerant Intolerant Unknown Intolerant Tolerant Tolerant Tolerant Tolerant Unknown Shade intermediate Tolerant Tolerant Shade intermediate Tolerant Shade intermediate Intolerant Intolerant Shade intermediate Intolerant Intolerant (long lived pioneer) Shade intermediate Intolerant Shade intermediate Intolerant (pioneer) Intolerant Intolerant
Clark and Clark, 1992; Gunter et al., 2011.
206
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207
guianense (Table 2) as documented by previous studies in the area (Pérez-Flores, 2001). The dominance of shade tolerant canopy trees can be facilitated by hurricanes that cause partial removal of the overstory and release understory species. Since tree species living in the understory are mostly shade tolerant, hurricanes can allow these species to reach the canopy and occupy available space. Therefore, wind disturbances can shape stand composition in two contrasting ways: while large blowdowns favor the regeneration of shade intolerant species, partial canopy removal favors shade tolerant species present in the understory. It is likely that due to differences in stand exposure both of these processes take place at the landscape level (Guariguata and Ostertag, 2001; Baker et al., 2005; Lugo, 2008). These hypotheses could be tested experimentally (Huston and Smith, 1987; Kraft et al., 2008; Comita et al., 2010) and by reconstructing stand history (Baker et al., 2005).
4.3. Management implications Previous studies using aerial photographs and satellite imagery showed that hurricane effects were more severe in broadleaf stands than in pine stands (Inafor, 2007; Rossi et al., 2013). Our data confirmed these observations but indicated that these studies overestimated hurricane effects in managed stands (Table 2). Pine resistance to wind was caused by a long history of land use that included logging, frequent surface fires, and grazing. The combined effects of these factors lowered stand density and eliminated young seedlings. However, because of its fast-growth and wide distribution, pine stands still offer various opportunities for management. Since seedlings of Pinus caribaea need three or four years to become resistant to surface fires (O’Brien et al., 2008), regeneration depends on seedlings escaping fire mortality (Myers et al., 2006; Veldman et al., 2015). Consequently, in remote locations or in stands specifically destined for timber production, recruitment can be favored by reducing fire frequency and grazing intensity. Since the land belonging to these indigenous communities comprises hundreds to thousands of hectares, the conflict between alternative land uses should be minimal. Grazing, timber production and agriculture could coexist in this landscape. Taken together, observed wind effects can be explained by marked differences in stand structure. While the dense and tall structure of broadleaf stands increased susceptibility to wind, the open and short structure of pines minimized it. These structural features were best captured by the height to diameter ratio which is perhaps the best indicator of tree and stand susceptibility to wind (Fig. 4; Supp. Fig. 2). Therefore, managers can increase wind resistance through early thinning in young stands or by removing trees with high ratios where appropriate (Mitchell, 1998; Wilson and Oliver, 2000). In sum, our results describe how strong wind coupled with selective logging and fires have shaped stand structure and illustrate potential effects of hurricane Felix on stand composition. Lastly, the findings of this stand-scale study highlight the importance of examining large disturbances across a range of spatial scales.
Acknowledgements This research was supported by a Prusser Fellowship from Clark University and by the Finfor project at CATIE, Costa Rica. The following people provided valuable support: Y. Ordóñez, W. Lau, and Canales J. Special thanks to D. Salazar, E. Cordon, W. Watler, F. Bascope and the late Rodolfo Cruz ‘‘Chaparro”, from Moss. Two anonymous reviewers provided thoughtful suggestions.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.foreco.2017.05. 034. References Alvarado, C., 2010. Manejo forestal: los primeros pasos de una comunidad indígena enfrentada a grandes dificultades. En Sabogal, C y Casaza, J. Casos ejemplares de manejo forestal sostenible en America latina y el Caribe. Junta de Castilla y Leon. FAO. ISBN: 978-92-5-306651-3
E. Rossi et al. / Forest Ecology and Management 400 (2017) 199–207 distribution of tropical biomes. Ecol. Lett. 15, 759–768. http://dx.doi.org/ 10.1111/j.1461-0248.2012.01789.x. Hubbell, S.P., 1979. Tree dispersion, abundance and diversity in a tropical dry forest. Science 203, 1299–1309. http://dx.doi.org/10.1126/science.203.4387.1299. Huston, M., 1979. A general hypothesis of species diversity. Am. Nat. 113, 81–101
207
Pacala, S.W., Canham, C.D., Silander, J.A., 1993. Forest models defined by field measurements: I. The design of a northeastern forest simulator. Can. J. For. Res. 23, 1980–1988. http://dx.doi.org/10.1139/x93-249. Pacala, S.W., Canham, C.D., Saponara, J., Silander, J.A., Kobe, R.K., Ribbens, E., 1996. Forest models defined by field measurements: II. Estimation, error analysis and dynamics. Ecol. Monogr. 66 (1), 1–43. Pérez-Flores, M.A., 2000. Composición y diversidad de los bosques de la región autónoma del Atlántico Norte nicaragüense: una base para el manejo sostenible. ‘‘Composition and diversity of the forests of The Autonomous Region of the Nicaraguan North Atlantic: a base for the sustainable management”. Master’s thesis. CATIE, Turrialba, Costa Rica. 155 p. Pérez-Flores, M.A., Finegan, B., Delgado, D., Louman, B., 2001. Composición y diversidad de los bosques de la región autónoma del Atlántico norte de Nicaragua: una base para el manejo sostenible. Revista forestal centroamericana 34 (Abr–Jun), 66–72.