Enrichment of big-leaf mahogany (Swietenia macrophylla King) in logging gaps in Bolivia: The effects of planting method and silvicultural treatments on long-term seedling survival and growth

Forest Ecology and Management 262 (2011) 2271–2280

Contents lists available at SciVerse ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Enrichment of big-leaf mahogany (Swietenia macrophylla King) in logging gaps in Bolivia: The effects of planting method and silvicultural treatments on long-term seedling survival and growth Rafael M. Navarro-Cerrillo a,⇑, Daniel M. Griffith a,e, María José Ramírez-Soria a, William Pariona b, Duncan Golicher c,d, Guillermo Palacios e a

Department of Forestry, University of Córdoba, Ed. Leonardo da Vinci, Campus Universitario Rabanales, Carretera Madrid-Cádiz Km 396, Córdoba 14071, Spain Proyecto ICAA/Rainforest Alliance, C/Luis Crespo 20, AP. 1A. Sopocachi, La Paz, Bolivia c Área de Conservación de la Biodiversidad, El Colegio de la Frontera Sur, Carretera Panamericana y Periférico Sur s/n, San Cristóbal de las Casas, Chiapas 29290, Mexico d Centre for Conservation Ecology & Environmental Change, School of Applied Sciences, Dorset House DG 38B, Bournemouth University, Fern Barrow, Poole (Dorset) BH12 5BB, UK e Center for Applied Research in Agroforestry Development (IDAF), University of Córdoba, Ed. Leonardo da Vinci, Campus Universitario Rabanales, Carretera Madrid-Cádiz Km 396, Córdoba 14071, Spain b

a r t i c l e

i n f o

Article history: Received 13 April 2011 Received in revised form 11 August 2011 Accepted 13 August 2011 Available online 10 September 2011 Keywords: Tropical timber Seed sowing Seedling transplant Vegetation control Canopy cover Forest restoration

a b s t r a c t To insure adequate regeneration and future timber yields of mahogany (Swietenia macrophylla King), many logged forests will have to be restocked through enrichment planting and managed using silvicultural techniques that maintain this species’ long-term survival and growth. This study compared the effects of planting method and two silvicultural treatments on the survival and growth of mahogany seedlings in logging gaps in Bolivia. We tested the hypotheses that survival and growth will be higher among transplanted seedlings than seedlings established from sown seeds and higher in silvicultural treatments that reduce competing vegetation and increase light. The first silvicultural treatment consisted of gaps logged 6 months prior to planting, gaps logged just prior to planting, and gaps treated with herbicide prior to planting. The second treatment, applied 12 months after planting, consisted of manual vegetation cleaning around mahogany seedlings in half of the gaps. The first hypothesis was supported in terms of initial seedling growth but not survival, which was similar between planting methods during the 12–92 months after planting. Transplanted seedlings grew significantly faster than those established from sown seeds during the first year, but this growth advantage disappeared by the second year. Although transplants were 84 cm taller than seed-sown seedlings by the end of the study, this height gain was probably not worth the cost of growing and transplanting seedlings. The second hypothesis was supported in terms of both survival and growth. A significantly greater proportion of seedlings survived in herbicide (62%) compared to 6-month-old (46%) and recent gaps (18%) and in cleaned (51%) versus control gaps (39%). Seedlings initially grew faster in herbicide and recent gaps than in 6-month-old gaps. These differences among silvicultural treatments were largely explained by canopy cover, which, throughout the study, was at least 14% lower in herbicide gaps and 9% lower in cleaned gaps relative to their respective alternatives. By 64 months growth diminished to near zero and no longer differed among gap treatments, despite lower canopy cover in herbicide gaps. By 92 months, saplings in herbicide gaps were only 145 and 77 cm taller than those in recent and 6-month-old gaps, respectively. To maximize survival and growth of mahogany seedlings in logging gaps while minimizing costs, silvicultural strategies should focus on direct seed sowing and appropriately timed interventions (i.e. manual cleaning) to control competing vegetation. Ó 2011 Elsevier B.V. All rights reserved.

1. Introduction As one of the most commercially valuable timber species in Latin America, big-leaf mahogany (Swietenia macrophylla King) ⇑ Corresponding author. Tel.: +34 957 218657; fax: +34 957 218563. E-mail address: [email protected] (R.M. Navarro-Cerrillo). 0378-1127/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2011.08.020

has been severely locally depleted throughout much of its range. By the mid 1990s in Mesoamerica, deforestation and over harvesting had reduced mahogany’s original geographic range to roughly one third (Calvo and Rivera, 2000). In South America, the commercial range of mahogany has been reduced to an estimated 94 million hectares, or 34% of the historic range (Grogan et al., 2010). In both regions many natural populations of mahogany have been so

2272

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

depleted by legal and illegal logging as to be commercially and, in some cases, ecologically unviable (Veríssimo et al., 1995; Kometter et al., 2004; Grogan and Schulze, 2008; Schulze et al., 2008; Grogan et al., 2010). Although mahogany’s adaptability to a wide range of ecological conditions suggests greater resilience to logging than is generally acknowledged (Mayhew and Newton, 1998; Brown et al., 2003; Grogan et al., 2003a), mahogany seedlings are usually scarce or absent from heavily logged forests (Lamb, 1966; Dickinson and Whigham, 1999; Grogan et al., 2003b; Grogan and Galvão, 2006). Instead, a dense cover of secondary species with low commercial value prevails (Pariona et al., 2003). To insure adequate regeneration and future timber yields of this highly valuable species, many logged, low-volume forests will have to be restocked through enrichment planting (Negreros-Castillo et al., 2003; Park et al., 2005; Grogan et al., 2005,2008). As one of the principal bottlenecks in mahogany production, regeneration has been the focus of most studies investigating strategies to improve mahogany establishment and growth (Mostacedo and Fredericksen, 1999; Grogan et al., 2008; Cámara-Cabrales and Kelty, 2009). Natural regeneration in gaps has been investigated as an alternative to plantations, where seedlings are highly susceptible to attack by the mahogany shoot borer (Hypsipyla grandella Zeller) throughout mahogany’s native range (Mayhew and Newton, 1998). However, lack of seed trees and low early survival under closed canopies are major limitations to natural regeneration in logged forests (Dickinson and Whigham, 1999; Grogan et al., 2003b, 2005; Pariona et al., 2003; Grogan and Galvão, 2006). To overcome limited seed production and dispersal, enrichment planting has been conducted either by sowing seeds or transplanting seedlings in gaps created by selective logging (Lopes et al., 2008) or railroad tie extraction (Negreros-Castillo and Mize, 2008). However, germination of mahogany seeds depends mainly on soil moisture (Morris et al., 2000), which tends to be higher under a closed canopy due to reduced evaporation rates than in gaps, although root water extraction can have a confounding effect (Marthews et al., 2008). Seedling survival and growth are limited by low light conditions created by the rapidly growing competing vegetation characteristic of gaps (Negreros-Castillo et al., 2003; Snook and Negreros-Castillo, 2004; Lopes et al., 2008). As these bottlenecks become better understood, silvicultural strategies are becoming increasingly refined to minimize the tradeoffs between mahogany germination, establishment, and growth. To overcome microsite limitations to early survival and growth, particularly low light, several studies have implemented techniques to control competing vegetation. For example, Grogan et al. (2005) demonstrated that experimentally opening the forest canopy increased the survival and growth of naturally established seedlings in 2–3-year old logging gaps. Other studies have combined direct seeding or seedling restocking with treatments that control competing vegetation (Ramos and del Amo, 1992). For example, Snook and Negreros-Castillo (2004) found that survival and growth of transplanted seedlings was significantly higher in treatments that removed large areas of the overstory and limited initial and subsequent competition (slash and burn and clearing by machine) compared to a treatment that removed the overstory but allowed vigorous resprouting (fell and leave) and the control. Few studies, however, have investigated the effect of different planting methods and silvicultural techniques simultaneously. Negreros-Castillo et al. (2003) found that germination of buried seeds was significantly higher than that of surface-sown seeds, and survival but not height differed significantly among clearing treatments. A prevailing lesson from these studies is that because light is the principal limiting factor for mahogany seedlings, competing vegetation should be controlled in space and time to ensure seedling establishment and maximize early growth. A greater understanding of the factors that limit survival and growth of

mahogany seedlings at different stages will help managers design practices that optimize regeneration while minimizing costs. In this study, we simultaneously compare the effects of enrichment planting method and silvicultural treatments that control competing vegetation on the survival and growth of mahogany seedlings in gaps created by selective logging. We test two main hypotheses: (1) Given the advanced establishment of roots, shoots and leaves, transplanted seedlings will exhibit higher survival and growth than seedlings established from directly sown seeds. (2) As a shade-intolerant species, mahogany seedlings will experience higher survival and growth in silvicultural treatments that reduce competing vegetation and increase light availability relative to other treatments and to a control. We also examine how the planting method and silvicultural treatments used here affect height outcomes and exposure of seedlings to attack by the mahogany shoot borer. 2. Methods 2.1. Study area This study was conducted in a 100,000-ha forestry concession managed by Agroindustria Forestal La Chonta Ltda., located 30 km northeast of Ascensión de Guarayos in the Department of Santa Cruz de la Sierra, Bolivia (15° 470 S, 62° 550 W, 250 m elevation, Fig. 1). Mean annual precipitation is 1562 mm, with <10% falling between May and September, and mean annual temperature is 25.3 °C (Gil-López, 1998). Dominant soils are oxisols, ultisols, and inceptisols and the topography is rolling to flat. Forests of the study area are seasonally dry humid tropical forests, containing a mix of species characteristic of the Amazon Basin and the drier Chiquitano forests to the south, which are predominately evergreen with a deciduous component (Park et al., 2005). The entire concession has been subjected to selective logging, which began in 1974, and over one third of the area burned in both 1995 and 2004 and has been undergoing succession ever since (Gil-López, 1998; Verwer et al., 2008). More than 150 tree species have been identified at the site. Mahogany and Spanish cedar (Cedrela fissilis Vell.) were the principal species extracted until 1997, but their increasing scarcity has shifted focus to a group of 13 additional species (Park et al., 2005). The most commercially important of these are Ficus boliviana C.C. Berg, Cariniana ianeirensis R. Knuth., Hura crepitans L., Schizolobium parahyba (Vell.) S.F. Blake, and Terminalia oblonga (Ruiz & Pav.) Steud. Approximately 2370 ha are selectively logged in the concession each year according to a 30year harvest cycle (Gil-López, 1998). 2.2. Experimental design The experiment compared the effects of two planting methods (directly sown seeds and transplanted seedlings) and two rounds of silvicultural treatments applied at different times (Fig. 2) on the survival and growth of mahogany seedlings in logging gaps. Fifty-five gaps created by recent logging were randomly selected from a 1760-ha stand based on harvest maps. Gaps containing a relatively closed canopy and advanced regeneration of commercial species were avoided. Mahogany seeds and seedlings were planted in the area that had been scarified by a skidder during the extraction process, which was equivalent among all gaps and left the soil bare of most vegetation. At the time of planting in December 2001, the experimental design consisted of the following three silvicultural treatments, hereafter called the ‘‘gap treatments’’:

2273

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

Fig. 1. Map of Bolivia and Santa Cruz de la Sierra department with a black square indicating the location of La Chonta forestry concession.

Control

N=8

Cleaned

N=9

Control

N=10

Cleaned

N=9

Control

N=9

Cleaned

N=10

6-monthold gaps N=17

Recent gaps N=19

Silvicultural Treatments

Herbicide gaps N=19 Time (months)

Variables Measured

0

Planting

12

24

Germination (%)

42

64

92

Survival (%) Height Growth rate Canopy cover (%)

Fig. 2. Experimental design depicting the two rounds of silvicultural treatments (gap and cleaning treatments), timeline (axis not to scale), and measurements made throughout the study. Note: Canopy cover was not measured at 92 months. See text for further details.

(1) Six-month-old gaps (N = 17) that had been opened by selective logging 6 months prior to the experiment with no

additional modification at this time. Regenerating vegetation in these gaps was dominated by non-woody plants

2274

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

(pteridophytes, Costus scaber Ruiz & Pav., Heliconia spp.), lianas (Pfaffia brachiata Chodat), and pioneer trees (Urera baccifera L.) Gaudich., Trema micrantha (L.) Blume, Cecropia spp.). (2) Recent gaps (N = 19) that had been opened by selective logging immediately prior to the experiment with no additional modification at this time. (3) Herbicide (N = 19), consisting of 5% glyphosate solution (RondopazÒ Pazchem Ltd., Askelon, Israel), was applied to all non-commercial vegetation within the scarified area of a random selection of 6-month-old gaps using a 20-L manual pump backpack sprayer. This treatment was conducted in mid-November 2001 three weeks prior to the enrichment planting of mahogany. In December 2002, one year after planting, a random selection of approximately half of the gaps within each gap treatment were subjected to a second round of silvicultural treatments, hereafter called the ‘‘cleaning treatments’’ (Fig. 2). Nine 6-month-old gaps, 9 recent gaps and 10 gaps sprayed with herbicide were cleaned with a machete, while the remaining gaps were designated as a control. Under this treatment all herbaceous and woody vegetation within approximately two meters around each mahogany seedling was cut near ground level. Stumps, roots and the soil were minimally affected. 2.3. Enrichment planting Mahogany seeds were collected from the forest floor at the study site towards the end of the dry season in August 2001. Seeds that appeared unviable were discarded. A portion of the apparently viable seeds were stored in burlap bags at ambient temperature and directly sown at the time of the experiment. The rest were planted in 20-cm plastic bags containing untreated soils from the forest concession and grown in an improvised nursery located 45 km from the study site at the La Chonta sawmill. At the time of planting, the seedlings were three months old and had a mean height and basal diameter of 17 and 0.25 cm, respectively. In early December 2001, the mahogany seeds and seedlings were planted in the scarified area of each gap according to a grid consisting of 2  2 m squares formed by evenly-spaced parallel and perpendicular lines. One mahogany seed or seedling was planted alternately at each point on the grid. The area where mahogany was planted was separated 4–5 m away from gap edges to avoid direct shading from the surrounding forest canopy. Gap areas ranged from 188 to 1074 m2. Mean area did not differ significantly among gap treatments (F2,52 = 0.08, P = 0.92) or cleaning treatments (F1,53 = 0.24, P = 0.63), and therefore is not considered to have influenced the effect of silvicultural treatments on seedling survival or growth. The number of seeds and seedlings planted per gap varied according to gap size, but the mean number per gap and the total number among all gaps were nearly equal (total transplanted seedlings = 669, mean per gap = 12.16 ± 0.52; total sown seeds = 672, mean per gap = 12.22 ± 0.57). Seeds were planted directly into the soil at a depth of 2–3 cm and seedlings along with the soil from the planting bag were inserted into 20-cm holes.

2002), 24 (December 2003), 42 (June 2005), 64 (April 2007), and 92 (August 2009) months after planting (Fig. 2). Seedlings were considered dead if their shoot was brown and shriveled. Immediately prior to the cleaning treatments in December 2002 and during each subsequent sampling period (except at 92 months), canopy cover in each gap was measured directly above each individual seedling with a spherical densiometer (Spherical Densiometer Concave Model C, Terra Tech). Attack by the shoot borer on individual seedlings was recorded 12 and 92 months after planting. Relative growth rates (RGR) in height were calculated according to Hoffmann and Poorter (2002). 2.5. Statistical analysis Analyses were carried out separately on canopy cover, seedling survival, seedling height, RGR in height, and shoot borer incidence at each sampling period to account for the different sequence of silvicultural treatments. The influence of planting method, gap treatment, and cleaning treatment on survival from 12 to 92 months after planting was tested using a Cox proportional hazard model, which takes into account both seedling longevity and status (dead or alive) at the final assessment of survival (Cox and Oakes, 1984). This analysis has been employed in previous studies of tree seedling survival (e.g. Clark, 2002), and is a suitable approach to evaluate survival patterns between treatments where the cumulative hazards over time (hazard functions) are generally proportional. Canopy cover, RGR and height were separately analyzed with linear mixed models to examine the effects of planting method, gap treatment, and cleaning treatment at each sampling period from 12 to 92 months after planting. The errors were modelled as random effects with individual mahogany seedlings nested within gaps, while planting method, gap treatment, and cleaning treatment were designated as fixed factors. Models containing the three-way interaction and all two-way interactions of the fixed factors were assessed at each time period using Akaike’s Information Criterion (AIC). Models were selected based on the lowest value of AIC, which indicates the optimal fit (Grogan and Landis, 2009). Although the manual cleaning was done at 12 months, the effect of this treatment was included in the models at this time to determine whether the dependent variables were different between cleaned and control gaps prior to the treatment. Basal diameter was highly correlated with height (r = 0.822, P 6 0.001) and is not discussed further. Canopy cover, survival, RGR and height were examined graphically to ascertain that the residuals were normally distributed and the variances were homogeneous. Height was subjected to log transformation to improve normality. Percent shoot borer incidence was analyzed separately at 12 and 92 months after planting with a Mann-Whitney Test using SPSSÒ (SPSS V.15.02, www.spss.com)to test for differences among planting methods and gap treatments, the latter with Bonferroni correction (alpha = 0.05/3 = 0.017). The Cox PH regression was run with the survival package (Therneau and Lumley, 2009) and linear mixed models were run using the nlme package (Pinheiro et al., 2009) in R version 2.13.0 (R Development Core Team, 2011).

2.4. Seedling measurements

3. Results

At the time of planting in December 2001, all transplanted mahogany seedlings were tagged and their height and basal diameter just above the root collar were measured. In March 2002, heights of germinants whose cotyledons had risen above the soil surface were measured. Thereafter, height, basal diameter, and survival of both transplanted seedlings and seedlings established from directly sown seeds (hereafter referred to as ‘‘seed-sown seedlings’’) were measured simultaneously at 12 (December

3.1. Canopy cover Twelve months after planting, mean canopy cover was significantly highest in 6-month-old gaps (73%), second highest in recent gaps (57%), and lowest in herbicide-treated gaps (33%; F2,51 = 34.2, P < 0.0001; Fig. 3A), and higher above seed-sown seedlings (57%) than transplanted seedlings (50%, F1,844 = 16.2, P = 0.0001; Fig. 3B). Differences in cover among gap treatments were presumably more

2275

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

Percent canopy cover

A

C

B

Fig. 3. Percent canopy cover measured above individual mahogany seedlings according to: (A) gap treatments (6-month-old gaps, recently opened gaps, and herbicidetreated gaps at the time of planting), (B) planting method (seedlings established from sown seeds versus transplanted seedlings), and (C) cleaning treatments (control versus gaps cleaned with a machete 12 months after planting). Different letters represent significantly different post-hoc differences between the respective treatments at alpha = 0.05 based on a linear mixed model for a given sampling period.

pronounced at the time of planting. Cover was not significantly different between cleaned and control gaps immediately prior to the cleaning treatment at twelve months (F1,51 = 0.25, P = 0.62;

Fig. 3C). At 24 and 42 months, the difference between planting methods disappeared (Table 1) as well as the difference between 6-month-old gaps and recent gaps, but cover remained significantly

Table 1 Linear mixed models at each sampling period for canopy cover, RGR in height, and log height of mahogany seedlings with the following fixed effects: planting method (sown seeds versus transplanted seedlings), gap treatments (6-month-old gaps, recently opened gaps, and gaps sprayed with herbicide), and cleaning treatment (control versus gaps cleaned with machete). The structure of fixed main effects and interactions reflect model selection based on Akaine’s Information Criterion. Canopy cover was not measured at 92 months. Bold letters indicate significant post-hoc differences at alpha = 0.05 based on a linear mixed model for each sampling period (12-92 months). Months since planning:

12

24

42

64

92

Dependent variable

Fixed effect

F

P

F

P

F

P

F

P

F

P

Canopy cover

Planting method Gap treatment Cleaning treatment Planting  Gap Planting  Cleaning Gap  Cleaning Planting method Gap treatment Cleaning treatment Planting  Gap Planting  Cleaning Gap  Cleaning Planting method Gap treatment Cleaning treatment Planting  Gap Planting  Cleaning Gap  Cleaning

16.2 34.2 0.25 – – – 58.6 8.61 0.17 4.01 0.53 – 86.3 7.19 0.049 2.36 1.43 –

0.0001 <0.0001 0.62 – – – <0.0001 0.0006 0.68 0.019 0.47 – <0.0001 0.002 0.83 0.095 0.23 –

1.06 13.4 4.96 – – – 1.20 15.7 3.34 – 4.67 1.51 41.0 9.34 0.42 – – –

0.30 <0.0001 0.030 – – – 0.27 <0.0001 0.07 – 0.031 0.23 <0.0001 0.0003 0.52 – –

1.53 17.9 7.58 – – – 0.29 3.66 2.05 – – – 35.0 9.11 1.13 – – –

0.22 <0.0001 0.008 – – – 0.59 0.033 0.16 – – – <0.0001 0.0004 0.29 – – –

14.1 10.7 2.88 – – – 0.20 0.85 0.43 – – – 26.6 9.84 2.30 – – –

0.0002 0.0001 0.10 – – – 0.65 0.43 0.51 – – – <0.0001 0.0003 0.14 – – –

NA NA NA NA NA NA 0.14 2.73 1.59 – – – 27.2 9.19 2.99 – – –

NA NA NA NA NA NA 0.70 0.08 0.22 – – – <0.0001 0.0005 0.09 – – –

Relative growth rate in height

Log Height

2276

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

3.2. Survival

lower in herbicide-treated gaps (24 months: F2,51 = 13.4, P < 0.0001; 42 months: F2,51 = 17.9, P < 0.0001; Fig. 3A). During this same period, gaps that had been manually cleaned at 12 months had significantly lower cover (54% and 52% at 24 and 42 months, respectively) than the control (64% at both time periods; 24 months: F1,51 = 4.96, P = 0.030; 42 months: F1,48 = 7.58, P = 0.008; Fig. 3C). By 64 months, the last time canopy cover was measured, the difference between herbicide-treated gaps and the other two gap treatments persisted (F2,48 = 10.7, P = 0.0001), the effect of cleaning treatment disappeared, and cover was again higher over seedlings sown from seed than those that had been transplanted (F1,538 = 14.1, P = 0.0002).

The percentage of transplanted seedlings surviving between the time of planting and 12 months later was 85%, in contrast to 50% of sown seeds producing a live seedling at 12 months. Between 12 months and the end of the experiment 92 months later, seedlings suffered 54% mortality overall. According to the Cox proportional-hazards regression, percent survival was not significantly different between transplanted seedlings and seed-sown seedlings (v2 = 0.060, P = 0.807), which showed a notably similar pattern of decreasing survival at every sampling period (Fig. 4A). Seedling survival between 12 months after planting until the end of the

Transplants Seedlings Seeds Seed-sown

100

80

60

40

20 A A

0 100

12 12 12

24 24 24

42 42 42

64

Time (months)

92 Cleaned Control

Percent survival

80

60

40

20

B

B

0 100

12

24

42

64

Time (months)

Recent 92 6 months-old Herbicide

80

60

40

20

C

C

0 12

24

42

64

92

Time (months) Fig. 4. Cox proportional-hazards regression model comparing survival between 12 and 92 months after planting, with the following factors: (A) planting method (seed-sown versus transplanted seedlings), (B) cleaning treatments (control versus gaps cleaned with machete), and (C) gap treatments (6-month-old gaps, recently opened gaps, and gaps sprayed with herbicide at the time of planting). Post-hoc comparisons for gap treatment were made relative to recent gaps and those made for cleaning treatment were made relative to the cleaned treatment.

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

study was higher in cleaned gaps (51%) compared to control gaps (39%, v2 = 17.2, P < 0.001, Fig. 4B), and highest in herbicide-treated gaps (62%), intermediate in 6-month-old gaps (46%), and lowest in recent gaps (18%, v2 = 104, P < 0.001, Fig. 4C). 3.3. Growth rates and height Mean RGR in height peaked at 100 cm cm1 year1 during the first 12 months but declined to 12 cm cm1 year1 during the final sampling period from 64 to 92 months (Fig. 5). Transplanted seedlings initially had significantly higher RGR than seed-sown seedlings (12 months: F1,849 = 58.6, P < 0.0001), but the effect of planting method on height growth disappeared by 24 months (Table 1). Among gap treatments, seedlings at 12 months grew significantly faster in herbicide-treated and recent gaps compared to 6-month-old gaps (F2,51 = 8.61, P = 0.0006). The interaction between planting method and gap treatment was also significant (F2,849 = 4.01, P = 0.019), as transplanted seedlings grew faster in recent and 6-month-old gaps relative to herbicide gaps than seed-sown seedlings (Fig. 5). By 24 months, however, seedlings in herbicide-treated gaps were growing faster than those in the other two gap treatments (F2,51 = 15.7, P < 0.0001). Also at this time, the effect of manual cleaning had a slightly stronger effect on the RGR of seed-sown seedlings than transplanted seedlings (planting method  cleaning treatment: F1,757 = 4.67, P = 0.031), but the effect of this interaction disappeared by the next sampling period (Table 1). The effect of the gap treatment weakened by 42 months (F2,51 = 3.66, P = 0.033). In post-hoc comparisons none of the gap treatments were significantly different from one another at this time. By 64 and 92 months relative growth rates were no longer significantly different among any of the treatments. Overall mean height of seedlings was 90 cm after 24 months (N = 830) and 310 cm after 92 months (N = 410). Not surprisingly, transplanted seedlings, which were up to three months old at the time of planting, were taller on average than seed-sown seedlings at 12 months (F1,857 = 86.3, P < 0.0001, Table 1). This difference remained significant until the end of the study. Seedlings of both

2277

planting methods were significantly taller in the herbicide-treated and recent gaps than in 6-month-old gaps at 12 and 24 months, but by 42 months and thereafter, both types of seedlings were significantly taller in herbicide-treated gaps than in recent and 6month-old gaps (Table 2). By the end of the study, transplanted seedlings were 84 cm taller on average than seed-sown seedlings, and seedlings in herbicidetreated gaps were 145 and 77 cm taller than recent and 6-monthold gaps, respectively (Table 2). The cleaning treatment never had a significant effect on height. 3.4. Mahogany shoot borer incidence The overall incidence of shoot borer attack dropped from 36.6 ± 4.8% to 1.3 ± 0.5% of seedlings per gap between 12 and 92 months after planting, respectively. The proportion of 12-month-old seedlings infested differed by gap treatment (F2,103 = 6.275, P = 0.003) but not planting method (F1,103 = 0.051, P = 0.822). Attack by the shoot borer at 12 months was significantly higher in 6-month-old gaps and gaps sprayed with herbicide than recent gaps (Fig. 6). By the end of the study, when competing vegetation was denser and overtopped many mahogany seedling crowns, only 7 of 43 gaps (12 gaps were not measured for shoot borer incidence in the final sampling period) contained seedlings that had been attacked and the proportion of infested seedlings did not differ by gap treatment or planting method. 4. Discussion 4.1. Survival and growth of transplanted versus seed-sown seedlings The first hypothesis, that transplanted mahogany seedlings will experience higher survival and growth than seedlings established from directly sown seeds due to the advanced establishment of roots, shoots and leaves, was partially supported. The percentage of seedlings established 12 months after planting was higher for transplants (85%) than sown seeds (50%), the latter a result of

RGR in height per year

200

100

0

12

24

42

64

92

Months Fig. 5. Relative growth rate in height (cm cm1 year1) of mahogany seedlings between 12 and 92 months after planting. The six points within each time period represent different combinations of planting method and gap treatments: S6 = seed-sown seedlings in 6-month-old gaps, SH = seed-sown seedlings in herbicide gaps, SR = seed-sown seedlings in recent gaps, T6 = transplants in 6-month-old gaps, TH = transplants in herbicide gaps, TR = transplants in recent gaps.

2278

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

Table 2 Average height ± S.E. (cm) of mahogany seedlings by planting method, gap treatment, and cleaning treatment. Different letters indicate significant post-hoc differences between the gap treatments only (i.e. not between planting method or cleaning treatments) at alpha = 0.05 based on a linear mixed model for each sampling period (12-92 months). Months since planting Transplant 6-month-old gaps

Herbicide gaps

Recent gaps

6-month-old gaps

Cleaned

Control

Cleaned

Control

Cleaned

Control

Cleaned

Control

Cleaned

Control

Herbicide gaps Cleaned

Control

17.2 ± 0.3 NA 58.1 ± 2.2a 92.0 ± 5.3a 159 ± 12b 176 ± 15b 251 ± 25b

18.1 ± 0.3 NA 63.9 ± 3.6a 100 ± 7.5a 150 ± 13b 179 ± 12b 168 ± 27b

16.8 ± 0.3 NA 44.8 ± 2.0b 81.1 ± 5.7b 158 ± 15b 234 ± 23b 360 ± 35b

17.3 ± 0.4 NA 40.5 ± 1.8b 57.1 ± 3.2b 87.6 ± 6.7b 142 ± 13b 204 ± 20b

16.9 ± 0.2 NA 59.6 ± 2.5a 119 ± 6.6a 208 ± 13a 294 ± 19a 398 ± 24a

16.8 ± 0.3 NA 56.8 ± 2.5a 116 ± 6.1a 205 ± 12a 280 ± 18a 378 ± 24a

NA 19.0 ± 0.8 43.1 ± 2.9a 73.5 ± 6.1a 115 ± 12b 154 ± 19b 152 ± 26b

NA 21.0 ± 1.0 47.2 ± 3.4a 71.4 ± 7.0a 104 ± 11b 150 ± 18b 130 ± 47b

NA 19.8 ± 0.7 36.3 ± 1.8b 67.6 ± 5.8b 132 ± 15b 188 ± 22b 265 ± 37b

NA 20.6 ± 0.8 33.8 ± 1.8b 48.5 ± 3.7b 69.8 ± 6.5b 104 ± 10b 153 ± 19b

NA 20.3 ± 0.7 47.4 ± 2.0a 99.3 ± 5.7a 170 ± 11a 223 ± 18a 306 ± 24a

NA 19.3 ± 0.7 45.3 ± 1.8a 86.7 ± 4.4a 146 ± 10a 204 ± 16a 284 ± 24a

Shoot borer incidence (%)

0 3 12 24 42 64 92

Seed-sown

Recent gaps

Fig. 6. Percent shoot borer incidence among gap treatments 12 and 92 months after planting. Different letters indicate significant post-hoc differences between the respective treatments based on multiple comparisons using a Mann–Whitney Test with Bonferroni correction (alpha = 0.05/3 = 0.017).

69% germination minus 19% early mortality of germinants (Negreros-Castillo et al., 2003). However, the proportion of those seedlings that survived until the end of the study was very similar between planting methods (Fig. 4A). Transplanted seedlings had higher growth rates than seed-sown seedlings, at least initially, and were quickly able to take advantage of the high light conditions present during the first year. They invested in rapid vertical growth, which enabled them to obtain a height advantage over seed-sown seedlings that persisted until the end of the study (Table 2). This greater size allowed them to overtop some understory vegetation at 12 months, as reflected by significantly lower canopy cover above transplants than seed-sown seedlings (Fig. 3B). Transplants also grew faster than seed-sown seedlings at this time in the shadier 6-month-old and recent gaps, relative to herbicide gaps, probably because tranplants’ advanced development allowed them to better tolerate shade (Fig. 5). However, this initial advantage in growth did not persist and did not translate into greater survivorship. 4.2. Seedling survival and growth among silvicultural treatments The second hypothesis, that mahogany seedlings will experience higher survival and growth in silvicultural treatments that reduce competing vegetation and increase light availability relative to other treatments and to a control, was supported but with caveats. Among gap treatments, differences in survival probability partially reflected differences in canopy cover. The highest proportion of surviving seedlings was found in herbicide-treated gaps (Fig. 4C), which had significantly lower canopy cover than 6month-old and recent gaps throughout the study (Fig. 3A). Survival was also higher in 6-month-old gaps than in recent gaps, possibly

because cover increased to 82% in the latter while remaining constant at 67% in the former (P > 0.05) at 64 months – the same time the proportion of surviving seedlings began to diverge substantially between these treatments(Fig. 4C). This level of canopy cover in recent gaps may represent a threshold of shade tolerance for mahogany. Further support of the second hypothesis was provided by the greater survival in manually cleaned (51%) versus control gaps (39%, Fig. 4B), a likely result of generally lower cover in the former (Fig. 3C). By comparison, Snook and Negreros-Castillo (2004) recorded 50% survival over 5 years for mahogany seedlings planted in 5000-m2 clearings produced using bulldozers or local slash and burn techniques. Well-timed clearing with machete may be a cost-effective strategy to control competing vegetation in gaps (see below). As with survival, differences in growth and height outcomes reflected canopy cover among gap treatments. The cleaning treatment had a relatively minor effect. The experiment was begun under favorable conditions for mahogany establishment and growth, such as reduced vegetation and the timing of planting to coincide with the rainy season (Fredericksen and Pariona, 2002).Precipitation was 70% higher than average during the year of planting (Verwer et al., 2008), which contributed to high establishment rates of mahogany but also rapid regeneration of ‘‘exploitative gap invaders’’ that dominated gaps following timber extraction (Fredericksen and Mostacedo, 2000; Grogan et al., 2003b; Pariona et al., 2003). While gaps were still relatively open during the first year, mean RGR of mahogany reached a maximum of100 cm cm1 year1but declined quickly thereafter (Fig. 5). RGR and heights were higher in recent and herbicide gaps than in 6month-old gaps where understory vegetation was denser and the

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

canopy had already begun to close (Table 2). Shoot borer incidence, which was highest in herbicide gaps and intermediate in 6-monthold gaps at this time, did not explain these differences (Fig. 6). By 24 months growth in recent gaps lagged behind that of herbicide gaps, suggesting that the glyphosate herbicide was effective in curtailing competing vegetation. By 64–92 months growth diminished to the point where all differences between gap treatments disappeared, although the earlier gains in height remained in the herbicide gaps. Although manual cleaning did not have a significant effect on mahogany growth or height, seedlings in 6-month-old gaps tended to grow faster in cleaned gaps compared to the controls. By 92 months, seed-sown seedlings and transplants in cleaned gaps were over 110 and 150 cm taller than controls, respectively (Table 2). No such height advantage existed in recent or herbicide gaps. This result suggests that where canopy cover is initially high, as in older gaps, manual cleaning provides a degree of competitive release for mahogany that persists long after the intervention. While seedling survival and growth among silvicultural treatments appeared to be governed mainly by differences in light, the effect of root competition for soil moisture and nutrients cannot be ruled out. In a deciduous, secondary forest similar to the one studied here, Gerhardt (1996) attributed the positive effect of root trenching on mahogany seedling survival and growth to reduced root competition for water. In this study, surrounding trees and especially the regenerating understory within gaps – both included within the measurement of canopy cover – probably competed with mahogany for soil moisture and nutrients (Barberis and Tanner, 2005). The degree to which light or root competition influenced mahogany survival and growth remains a question for further study. 4.3. Torward a silviculture of enrichment of S. macrophylla Given the low to non-existent natural regeneration of many commercial species in Neotropical forests, enrichment planting is increasingly needed at large scales to insure sustained timber yields (Mostacedo and Fredericksen, 1999; Fredericksen and Mostacedo, 2000; Pariona et al., 2003; Fredericksen and Putz, 2003; Park et al., 2005; Dauber et al., 2005; Verwer et al., 2008). This is particularly the case for mahogany, many populations of which have been heavily exploited (Dickinson and Whigham, 1999; Grogan and Galvão, 2006). Scaling up enrichment planting efforts requires a greater understanding of the tradeoffs of different planting methods. This study explores the question of whether the establishment rates and heights achieved for transplanted seedlings justify the expense for the material, transport, and labor required to raise mahogany seedlings in a nursery and transplant them to gaps. Planting method had no effect on survival of established seedlings between 12 and 92 months and the only growth advantage of transplants was an 84-cm increment in height, which is probably not important in the lifespan of a mahogany tree. However, the 35% lower proportion of one-year-old seedlings yielded by sown seeds compared to transplants, despite a high germination rate due to favorable rainfall (Verwer et al., 2008), highlights the risk of relying solely on sown seeds. Yet if seed sowing is timed with the onset of the rainy season, seeds are buried just below the soil surface (Negreros-Castillo et al., 2003), and competing vegetation is kept under control (Snook and Negreros-Castillo, 2004), then the likelihood of germination failure and early seedling mortality can be minimized. Despite carrying some risk, direct seed sowing represents the most cost-effective method for enrichment planting of mahogany at large scales. Enrichment planting by itself, however, is insufficient to sustain mahogany regeneration without some form of vegetation control to overcome microsite limitations to survival and growth. The effi-

2279

cacy of silvicultural practices must be weighed against relative economic and environmental costs. Glyphosate herbicide in this study effectively suppressed competing vegetation and increased light penetration to the understory. While this significantly benefited survival and early growth, the rapid decline in growth and the small 145 and 77-cm height advantages of seedlings in herbicide gaps over recent and 6-month-old gaps, respectively, call the necessity of herbicide use into question (Pariona et al., 2003). Manual cleaning with machete also significantly improved mahogany survival, although its effects on growth and height were more subtle. Economically, a three-person team in Bolivia can clean up to 15 gaps per day at a cost of US$1.23 per gap using manual cleaning and US$2.13 per gap using glyphosate (Pariona et al., 2003). Glyphosate also has serious external costs such as detrimental effects on the environment, including soil microorganisms that may be beneficial to plant growth (Altieri, 2009). In the final assessment, manual cleaning can be as effective as glyphosate herbicide if applied properly, and is preferable in terms of economic and environmental costs. The small area of logging gaps restricts the spatial impact that silvicultural practices can have on competing vegetation, which shifts the manager’s focus to the method and timing of interventions. Studies of mahogany regeneration concur that practices intended to reduce competing vegetation usually have a transient effect that lasts at most a few years (Negreros-Castillo et al., 2003;Pariona et al., 2003; Snook and Negreros-Castillo, 2004; Grogan and Galvão, 2006). In this study, a second intervention to clean competing vegetation is recommended around 42 months, or shortly before, to prevent stagnation of growth rates (Fig. 5). Intensive silvicultural intervention, such as cutting vines and girdling trunks of non-commercial species that shade the gaps (Peña-Claros et al., 2008), will further reduce competing vegetation and allow more light to reach seedlings (Fredericksen and Putz, 2003). Well-timed, repeated manual cleaning of competing vegetation in gaps (i.e. 3–4 years after initial planting and less frequent interventions thereafter) will provide microsites with high light that optimize mahogany seedling survival and growth (Grogan et al., 2005). These techniques should be further studied and refined to insure that saplings reach the higher strata of the canopy, where growth rates are maximized, in order to obtain future harvests in a timely manner. Acknowledgments This study was a collaboration between the Autonomous University of Gabriel Rene Moreno (Santa Cruz de la Sierra, Bolivia) and the University of Córdoba-Campus de Excelencia CeiA3 (Spain). Funding was provided by the Spanish Agency for International Development Cooperation (AECID) and logistical support was provided by La Chonta forest concession. We thank the Bolivian Institute of Forest Research (IBIF) for its key role in the experimental design, data collection, and use of facilities in the field. We are also grateful to two anonymous reviewers for constructive comments on earlier versions of the manuscript. This research was conducted as part of the undergraduate thesis of M. J. Ramírez-Soria at the University of Córdoba. References Altieri, M.A., 2009. The ecological impacts of large-scale agrofuel monoculture production systems in the Americas. Bulletin of Science, Technology & Society 29, 236–244. Barberis, I.M., Tanner, E.V.J., 2005. Gaps and root trenching increase tree seedling growth in Panamanian semi-evergreen forest. Ecology 86, 667–674. Brown, N., Jennings, S., Clements, T., 2003. The ecology, silviculture and biogeography of mahogany (Swietenia macrophylla): a critical review of the evidence. Perspect. Plant Ecol. Evol. Syst. 6, 37–49.

2280

R.M. Navarro-Cerrillo et al. / Forest Ecology and Management 262 (2011) 2271–2280

Calvo, J.C., Rivera, H., 2000. Evaluation of mahogany (Swietenia macrophylla King) in Mesoamerica: general overview. Tropical Science Center and Programa Ambiental Regional para Centroamerica/Central American Protected Areas System (PROARCA/CAPAS), San José, Costa Rica. Cámara-Cabrales, L., Kelty, M.J., 2009. Seed dispersal of big-leag mahogany (Swietenia macrophylla) and its role in natural forest management in the Yucatan Peninsula, México. Journal of Tropical Forest Science 21, 235–245. Clark, P.J., 2002. Experiments on tree and shrub establishment in temperate grassy woodlands: seedling survival. Aust. Ecol. 27, 606–615. Cox, D.R., Oakes, D., 1984. Analysis of Survival Data. Chapman and Hall, New York, NY. Dauber, E., Fredericksen, T.S., Peña, M., 2005. Sustainability of timber harvesting in Bolivian tropical forests. For. Ecol. Manage. 214, 294–304. Dickinson, M.B., Whigham, D.F., 1999. Regeneration of mahogany (Swietenia macrophylla) in the Yucatan. Int. For. Rev. 1, 35–39. Fredericksen, T.S., Mostacedo, B., 2000. Regeneration of timber species following selection logging in a Bolivian tropical dry forest. For. Ecol. Manage. 131, 47–55. Fredericksen, T.S., Pariona, W., 2002. Effect of skidder disturbance on commercial tree regeneration in logging gaps in a Bolivian tropical forest. For. Ecol. Manage. 171, 223–230. Fredericksen, T.S., Putz, F.E., 2003. Silvicultural intensification for tropical forest conservation. Biodivers. Conserv. 12, 1445–1453. Gerhardt, K., 1996. Effects of root competition and canopy openness on survival and growth of tree seedlings in a tropical seasonal dry forest. For. Ecol. Manage. 82, 33–48. Gil-López, P., 1998. Plan general de manejo forestal empresa Agroindustria Forestal La Chonta Ltda. Empresa Agroindustria Forestal La Chonta Ltda.: Ascención de Guarayos, Bolivia. 44 pp. Grogan, J., Ashton, M.S., Galvão, J., 2003a. Big-leaf mahogany (Swietenia macrophylla) seedling survival and growth across a topographic gradient in southeast Pará. Brazil. For. Ecol. Manage. 186, 311–326. Grogan, J., Galvão, J., Simões, L., Veríssimo, A., 2003b. Regeneration of Big-Leaf Mahogany in Closed and Logged Forests of Southeastern Pará, Brazil. In: Lugo, A., Figueroa Colón, J.C., Alayón, M. (Eds.), Big-Leaf Mahogany: Genetics, Ecology and Management. Springer-Verlag, New York, NY, USA, pp. 193–208. Grogan, J., Landis, R.M., Ashton, M.S., Galvão, J., 2005. Growth response by big-leaf mahogany (Swietenia macrophylla) advance seedling regeneration to overhead canopy release in southeast Pará. Brazil. For. Ecol. Manage. 204, 399–412. Grogan, J., Galvão, J., 2006. Factors limiting post-logging seedling regeneration by big-leaf mahogany (Swietenia macrophylla) in southeastern Amazonia, Brazil, and implications for sustainable management. Biotropica 38, 219–228. Grogan, J., Schulze, M., 2008. Estimating the number of trees and forest area necessary to supply internationally traded volumes of big-leaf mahogany (Swietenia macrophylla) in Amazonia. Environmental Conservation 35, 26–35. Grogan, J., Jennings, S., Landis, R.M., Schulze, M., Baima, A., do Carmo, J., Lopes, A., Norghauer, J.M., Oliveira, L.R., Pantoja, F., Pinto, D., Silva, J.N., Vidal, E., Zimmerman, B.L., 2008. What loggers leave behind: Impacts on big-leaf mahogany (Swietenia macrophylla) commercial populations and potential for post-logging recovery in the Brazilian Amazon. For. Ecol. Manage. 255, 269– 281. Grogan, J., Landis, R.M., 2009. Growth history and crown vine coverage are principal factors influencing growth and mortality rates of big-leaf mahogany Swietenia macrophylla in Brazil. J. Applied Ecol. 46, 1283–1291. Grogan, J., Blundell, A.G., Landis, R.M., Youatt, A., Gullison, R.E., Martinez, M., Kómetter, R., Lentini, M., Rice, R.E., 2010. Over-harvesting driven by consumer demand leads to population decline: big-leaf mahogany in South America. Conservation Letters 3, 12–20.

Hoffmann, W.A., Poorter, H., 2002. Avoiding bias in calculations of relative growth rate. Ann. Bot. 80, 37–42. Kometter, R.F., Martinez, M., Blundell, A.G., Gullison, R.E., Steininger, M.K., Rice, R.E., 2004. Impacts of unsustainable mahogany logging in Bolivia and Peru. Ecol. Soc. 9. http://www.ecologyandsociety.org/vol9/iss1/art12. Lamb, F.B., 1966. Mahogany of Tropical America: Its Ecology and Management. University of Michigan Press, Ann Arbor, Michigan, USA. Lopes, J., Jennings, S.B., Matni, N.M., 2008. Planting mahogany in canopy gaps created by commercial harvesting. For. Ecol. Manag. 255, 300–307. Marthews, T.R., Burslem, D.F.R.P., Paton, S.R., Yanyuez, F., Mullins, C.E., 2008. Soil drying in a tropical forest: three distinct environments controlled by gap size. Ecological Modelling 216, 369–384. Mayhew, J.E., Newton, A.C., 1998. The Silviculture of Mahogany. CABI Publishing, Wallingford, UK. Morris, M.H., Negreros-Castillo, P., Mize, C., 2000. Sowing date, shade and irrigation affect big-leaf mahogany (Swietenia macrophylla King). For. Ecol. Manage. 132, 173–181. Mostacedo, B., Fredericksen, T.S., 1999. Regeneration status of important tropical forest tree species in Bolivia: assessment and recommendations. For. Ecol. Manage. 124, 263–273. Negreros-Castillo, P., Snook, L., Mize, C., 2003. Regenerating mahogany (Swietenia macrophylla) from seed in Quintana Roo, Mexico: the effects of sowing method and clearing treatment. For. Ecol. Manage. 183, 351–362. Negreros-Castillo, P., Mize, C., 2008. Regeneration of mahogany and Spanish cedar in gaps created by railroad tie extraction in Quintana Roo. Mexico. For. Ecol. Manage. 255, 308–312. Pariona, W., Fredericksen, T.S., Licona, J.C., 2003. Natural regeneration and liberation of timber species in logging gaps in two Bolivian tropical forests. For. Ecol. Manage. 181, 313–322. Park, A., Justiniano, M.J., Fredericksen, T.S., 2005. Natural regeneration and environmental relationships of tree species in logging gaps in a Bolivian tropical forest. For. Ecol. Manage. 217, 147–157. Peña-Claros, M., Peters, E.M., Justiniano, M., Bongers, F., Blate, G., Fredericksen, T., Putz, F., 2008. Regeneration of commercial tree species following silvicultural treatments in a moist tropical forest. For. Ecol. Manage. 255, 1283–1293. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D.the R Core team, 2009. Linear and Nonlinear Mixed Effects Models. R package version 3, 1–96. R Development Core Team. 2011. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3900051-07-0, URL http://www.R-project.org. Ramos, J.M., del Amo, S., 1992. Enrichment planting in a tropical secondary forest in Vera Cruz. Mexico. For. Ecol. Manage. 54, 289–304. Schulze, M., Grogan, J., Landis, R.M., Vidal, E., 2008. How rare is too rare to harvest? Management challenges posed by timber species occurring at low densities in the Brazilian Amazon. For. Ecol. Manage. 256, 1443–1457. Snook, L.K., Negreros-Castillo, P., 2004. Regenerating mahogany (Swietenia macrophylla King) on clearings in Mexico’s Maya forest: the effects of clearing method and cleaning on seedling survival and growth. For. Ecol. Manage. 189, 143–160. Therneau, T., original R port by Lumley, T. 2009. Survival: Survival analysis, including penalised likelihood. R package version 2.35-8. http://CRAN.Rproject.org/package=survival. Veríssimo, A., Barreto, P., Tarifa, R., Uhl, C., 1995. Extraction of a high-value natural resource in Amazonia: the case of mahogany. For. Ecol. Manage. 72, 39–60. Verwer, C., Peña-Claros, M., van der Staak, D., Ohlson-Kiehn, K., Sterck, F.J., 2008. Silviculture enhances the recovery of overexploited mahogany Swietenia macrophylla. J. Applied Ecol. 45, 1770–1779.