Mechanical site preparation severity mediates one-year-survival response to summer drought in planted tree seedlings

doi:10.21203/rs.3.rs-3796037/v1

In face of future climatic conditions, methods to ensure the success of forest plantation in warm and dry conditions are required. Mechanical site preparation (MSP) prior to planting is widely used around the world to enhance seedling establishment success. Our study aimed at identifying, among a set of MSP methods that are used in practical forestry, which methods ensure plantation success under dry weather conditions.

We evaluated the combined effects of summer drought (estimated using the Standard Precipitation Index) and MSP severity (estimated using vegetation cover in the immediate seedling vicinity one year after MSP) on 1-year seedling survival. We used a network of 20 experimental sites established in France over a 10-year-period, and where seedlings were planted after various MSP.

In all treatments (severe MSP, moderate MSP, no MSP), seedling survival was higher in years with rainy summers than in years with dry summers. In rainy years, both moderate and severe MSP methods slightly improved the seedling survival rate (95%) compared to the unprepared control (88%). In dry years, seedling survival was similar after moderate MSP or with no MSP (50 and 54%, respectively), whereas it was much higher after severe MSP (79%).

In practical forestry, severe MSP appears as an option to enhance early seedling survival, especially when summer precipitations are lower than the seasonal average, whereas moderate MSP does not significantly improve seedling survival compared to an unprepared control, in all summer weather conditions.

Standard precipitation index

climate change

reforestation

Global increases in temperature and drought, which are presently observed in western Europe and are expected to rise in the next decades (IPCC, 2023), have multifold impacts on forest ecosystems, threaten their resilience, and call for adaptation actions in forest management (Keenan, 2015; Lindner et al., 2014). Plantation of tree seedlings is a major tool to reinforce the resilience of forest ecosystems to current and future climatic conditions (Stanturf et al., 2014). It allows stand reforestation and restoration in all situations where natural regeneration from a seed bank or from sprouts does not provide the adequate material to meet the management objectives (Stanturf et al., 2001): lack of seed-producing trees, decision to change tree species, decision to increase tree species diversity, or use of genetically improved planting material.

Ensuring a high level of plantation success, even under hot and dry weather conditions, is a prerequisite to using plantation as a management measure to adapt forest ecosystems to future climatic conditions. Methods to produce and plant tree seedlings have evolved in the last decades and have significantly improved plantation success (Chirino et al., 2009; Grossnickle, 2018; South et al., 2023), but the risk of failure still remains high when conditions get harsh, and especially when rainfall is low and temperatures are high. There is ample evidence that hot and dry conditions during the growing season are deleterious to seedling establishment and planting success (Padilla and Pugnaire, 2007). In drylands, characterized by repeated seasonal drought, plantation establishment is severely limited by long, dry periods that occur during the spring and summer (del Campo et al., 2022; Pausas et al., 2004). In addition, the analysis of interannual variation in plantation success in most forestry regions reveals that the average seedling survival rate varies from year to year and primarily depends on the severity of spring and summer drought, with dry years resulting in significantly lower seedling survival than wet years (Zwolinski et al., 1994). Across Europe, in the Mediterranean (del Campo et al., 2022), temperate (Boutte et al., 2023) and boreal (Luoranen et al., 2023) zones, plantation success becomes critical in dry years, especially for drought-sensitive species (del Campo et al., 2020). These observations underline the urgent need to develop plantation methods that ensure plantation success during dry years for different forestry regions, including regions that were not previously affected by recurring seasonal drought.

Silvicultural methods to improve plantation success in dry conditions aim at improving the water status of the newly planted seedling by increasing the amount of water available to the seedling or by enhancing seedling water uptake capacity (Chirino et al., 2009). Mechanical site preparation (MSP) prior to planting is widely used around the world to enhance seedling establishment success. The aim of MSP is to reduce the various constraints that limit the survival, growth and development of young trees, i.e., problems with the physical characteristics of soils, competition from neighboring vegetation, or herbivory, and to create suitable microsites where the regenerating trees can thrive (Buitrago et al., 2014; Burton et al., 2000). More specifically, in plantation sites with limited water supply, MSP may improve seedling water status by reducing the abundance of vegetation in the immediate vicinity of the seedling, therefore improving soil water availability in the seedling rooting zone, as well as by reducing soil resistance to root growth, thus facilitating root system development (Löf et al., 2012). Numerous studies have documented the beneficial effects of MSP on seedling survival and early growth, especially in dry conditions (Querejeta et al., 2001). Although it has not been clearly shown in the previous literature, the positive impact of MSP on seedling establishment is expected to increase with the intensity of drought during the growing season in all sites with limited water supply.

A variety of MSP methods, which differ in the equipment used and in their implementation in the field, have been developed worldwide in response to specific biotic and abiotic constraints, in order to match site characteristics with silvicultural objectives (von der Gönna, 1990). MSP methods all create a disturbance around the seedlings but they markedly differ in the disturbances they induce, which may be described according to: (i) Compartments: MSP primarily affects the vegetation and the soil compartments; (ii) Impacts on the compartments: depending on the method, MSP may entirely remove the vegetation or only part of it (roots, shoots or leaves, living plant parts or organic material left by previous vegetation) (Chaves Cardoso et al., 2020), and may scalp, scarify, loosen, invert or mix the upper soil horizons (Sutton, 1993). In turn, these changes have profound impacts on the ecosystem characteristics and processes that occur in the seedling neighborhood (soil water availability, nutrient stocks and fluxes, microclimate, biotic interactions); (iii) Spatial scale: MSP may perform a continuous (over the entire surface area of the stand) or intermittent (on the planting lines or on spots in the immediate vicinity of the seedlings) preparation (Sikström et al., 2020); and (iv) Severity of impact: MSP methods may differ in their severity, which refers to the extent of the impacts that occur in the disturbed compartment (Frelich, 2002; Iwasaki and Noda, 2018). In most studies, the severity of MSP is expressed as a change in vegetation biomass or cover, or as the area, volume or depth of the disturbed soil.

In the field of disturbance ecology, various descriptors have been developed to understand the effects of natural or anthropogenic disturbances such as fire, windstorm, drought and forest harvesting on ecosystems (Peters et al., 2011). Among these descriptors, the severity of the disturbance, often measured as the mortality rate of a population or the proportion of biomass removed, is an integrative indicator that has proven to be very useful to compare disturbances of different kinds (Kurth et al., 2020; Roberts, 2007). Following the same conceptual approach, Chaves Cardoso et al. (2020) suggested examining site preparation methods through the lens of disturbance ecology and, more specifically, using the severity of impact as an indicator to understand the effects of different site preparation methods, as well as evaluating their suitability for various forest management goals. In previous works, Mattson & Bergsten (2003) and Hjelm et al. (2019) used a severity index to compare MSP methods that differed in the equipment used and in their impact on the soil and the vegetation, and showed that the severity of the MSP method was positively related to early seedling performance and to long-term stand productivity.

The general objective of our study was to identify, among a set of MSP methods that are used in practical forestry, which methods ensure plantation success under dry weather conditions. These methods could then be recommended to forest managers to enhance plantation success. We focused on seedling survival during the first growing season. In many European countries, 1-year-survival is used as a criterion to evaluate planting success, with a threshold value set between 80 and 90% (Mataruga et al., 2023). Below this value, the plantation is considered to have failed, and refilling operations, which are very costly, are often implemented by the manager. Post-planting stress very often causes high seedling mortality rates during the first months, which may increase with adverse growing conditions (Grossnickle and Ivetić, 2022). Environmental factors that determine seedling survival after one year differ from those that determine survival or growth over longer periods (2 to 5 years) (Cole et al., 2018; Dumas et al., 2021), and 1-year-survival must be specifically analyzed in studies aimed at improving plantation methods. Our study was performed in France, between 2011 and 2019, a period that was characterized by intense summer droughts in some years that caused extensive seedling mortality during their first growing season (Tallieu et al., 2022).

We used the severity of the disturbance to compare the effects of various MSP methods and to identify the methods that ensure a high rate of seedling survival in the first year. We classified the MSP methods according the severity of the disturbance they induce in the immediate vicinity of the seedlings into two classes: severe and moderate. We expected that MSP methods that induce severe disturbances because they provide better access to soil water to the newly planted seedlings would ensure greater plantation success than MSP methods that produce moderate disturbance. We hypothesized that (1) MSP improves 1-year seedling survival, compared to an unprepared control; (2) the improvement in seedling survival is higher under dry weather and soil conditions; (3) severe MSP methods are more beneficial to seedling survival than MSP with moderate severity; and (4) the benefit increase from moderate to severe MSP methods is higher under dry conditions. To test these hypotheses, we used a network of field experiments that evaluate the performance of various MSP methods used in plantation forestry in France, and we compared first-year seedling survival in plots with severe, moderate or no MSP.

Study sites

The field experiments were established at 20 sites in France, with latitudes ranging from 44°N to 49°N, and longitudes from 1°W to 8°E (Table 1). Soil texture ranged from sandy to clayey and soil moisture regime ranged from water-logged to dry soils. Field vegetation was dominated by various grass, fern, bramble or woody species, depending on the site. Annual minimum and maximum temperatures ranged between 5.7 and 8.0° and between 14.6 and 19.3°, respectively, and annual precipitation between 633 and 1183 mm.

The experiments were established between 2011 and 2019 to evaluate the performance of various MSP methods. All MSP methods were applied between one and nine years after clear felling, depending on the site. In each site, MSP methods were applied in autumn and the seedlings were planted within a few months, in the following winter or early spring. Between four and six MSP methods and an unprepared control were tested in each site. Treatments were not replicated within each site. Each experimental site compared a specific set of MSP methods, selected among the methods used locally in practical forestry. Methods differed in the type of tool (disc, ripper, tooth, plough) used, the machine (tractor, excavator) used, the type of preparation performed (scarification, trenching, mounding, plowing), the spatial extent of the preparation (along the planting lines or on spots around the seedlings), and in the severity of the disturbance they induced.

We evaluated the severity of the disturbance of each MSP method on the basis of its impact on neighboring vegetation in the immediate vicinity of the seedlings during the first growing season after MSP. In each MSP method plot of each experimental site, we measured vegetation cover on ten 1-m²-subplots surrounding the seedlings one year after site preparation, and we computed the ratio between the average vegetation cover measured in the treatment and in the control. These ratios ranged from 0.02 to 1. Each MSP method was tested in several experimental sites, and we estimated the severity of the disturbance of MSP methods as the mean value of the ratios computed in all experiments where the method was tested. MSP methods with a mean ratio below 0.3, i.e., methods that removed more than 70% of the vegetation cover compared to the unprepared control, were classified as very severe and referred to as “S” for “Strong”. MSP methods with a mean ratio above 0.3, i.e., methods that removed less than 70% of the vegetation cover compared to the unprepared control, were classified as moderately severe, and referred to as “M” for “Moderate”. The threshold value of 0.3 was chosen in order to obtain a balanced number of M and S plots in our dataset. Selecting a lower or higher value would have forced us to either eliminate some experimental sites and lower the total number of observations, or obtain a strongly unbalanced design. The unprepared Control was referred to as “C”.

In each experimental site, we selected one S method, one M method, and the unprepared Control. In each site, we selected the S and M methods that were most commonly tested across the 20 experiments so as to avoid, whenever possible, methods that were tested in too few experiments. In three experiments, no M method was retained and in one experiment, no Control was available.

Finally, we selected nine MSP methods across the 20 experimental sites. The four M methods were: (1) disk pulled by a tractor (five experimental sites; mean ratio between vegetation cover in the treatment and in the control: 0.86); (2) scalper mounted on an excavator followed by a ripper-disk trencher pulled by a tractor (three sites; mean ratio: 0.38); (3) ripper pulled by a tractor (seven sites; mean ratio: 0.84); and (4) plow pulled by a tractor (two sites; mean ratio: 0.34). The five S methods were: (5) ripper-scarifier mounted on an excavator (seven sites; mean ratio: 0.23), (6) ripper-scarifier-disk trencher mounted on an excavator (two sites; mean ratio: 0.29); (7) scalper mounted on an excavator followed by a ripper-scarifier mounted on an excavator (one site; mean ratio: 0.13); (8) scarifier mounted on an excavator (two sites; mean ratio: 0.29); and (9) scarifier mounted on an excavator followed by a ripper-scarifier mounted on an excavator (eight sites; mean ratio: 0.14). Vegetation cover after one year averaged 84% in the Control, and 59% and 19% in all methods pooled from treatments Moderate and Strong, respectively.

The experimental sites were planted with different tree species: pines (Pinus sylvestris L., Pinus nigra var. corsicana (Loudon) Hyl., Pinus pinaster Aiton), oak (Quercus petraea (Matt.) Liebl., Quercus robur L.) and Douglas fir (Pseudotsuga menziesii (Mirb.) Franco). Seven sites were planted with two species, and 13 with one species. Six sites were refilled the following year because of high seedling mortality in some treatments.

We considered each site x MSP method x species x year combination as an observation. In total, 77 observations were available for our study.

Measurements and data

In each site and each treatment, between 95 and 388 seedlings of each planted species were selected and tagged immediately after planting. In the following winter, all seedlings were examined and their status (alive or dead) was recorded. In the six sites that were refilled, both the initial and the refilled seedlings were measured. Seedling survival rate after one growing season was computed for each observation.

We estimated meteorological summer drought in each site and each year using the Standardized Precipitation Index (SPI) computed over 3 months (June, July, August) (McKee et al., n.d.). SPI correlates well with plantation success (Del Campo et al., 2021). In a study that used data from 10,404 planting sites in conditions (geographic area, calendar years, planted tree species and MSP methods) that encompassed the conditions of our study, Tallieu et al. (2022) showed that SPI computed over June, July and August was a good estimator of seedling mortality in the first year after planting. SPI values for all sites and all years were extracted from the Météo-France SAFRAN database, where SPI values are computed over an 8 km x 8 km grid and use a 30-year period (1981-2010) to estimate the long-term rainfall distribution (Soubeyroux et al., 2012).

Statistical analyses

First, we analyzed the crossed effects of MSP and summer drought on seedling survival using a beta regression. We then ran a full model, where treatments (3 levels: S, M, C) and SPI estimated in each site and each year (continuous factor) were the predictors, and seedling survival estimated in each observation was the response variable. We performed a likelihood ratio test, which indicated that the interaction between the two predictors was statistically significant (p-value=0.024). As a consequence, three reduced models were separately fitted for treatments C, M and S, where SPI was the predictor and seedling survival the response variable. We used the logit function as a link function in all models.

Second, we used Wilcoxon signed rank tests to estimate whether treatments M and S induced higher seedling survival than treatment C, in the same site and same year. Unilateral paired tests were performed separately for treatments M and S, and separately for years with summer rain above (SPI>0) and below (SPI<0), the median long-term precipitation. We also tested whether treatment S induced higher seedling survival than treatment M in the same site and same year using the same tests.

All data treatments and statistical analyses were performed using the R environment (Cribari-Neto and Zeileis, 2010; R Core Team, 2023).

In all three treatments, average seedling survival in the first year progressively decreased as SPI decreased from 2 (very rainy year) to -1.8 (very dry year) (Fig. 1): it decreased from 89% to 34% in the unprepared control, from 97% to 32% in treatment M, and from 94% to 75% in treatment S. Parameters of the beta regression models expressed in the logit link are given for treatment C [intercept=0.59 (std-err: 0.21); slope=0.81 (std-err: 0.19)]; M [intercept=1.25 (std-err: 0.17); slope=1.25 (std-err: 0.16)]; and S [intercept=1.82 (std-err: 0.20); slope=0.46 (std-err: 0.15)].

After rainy summers (SPI>0), survival rates ranged between 64% and 100% in all treatments pooled. Seedling survival averaged 88, 95 and 95%, in treatments C, M and S, respectively. Unilateral paired Wilcoxon tests indicated that seedling survival was significantly higher in both treatments M and S, compared to treatment C, but that treatment S was not higher than treatment M (p-values of 0.0053, 0.0016 and 0.45, respectively).

After dry summers (SPI<0), seedling survival variability within each treatment was much higher than in rainy years, and survival rates ranged between 0% and 100%. Seedling survival averaged 50, 54 and 79%, in treatments C, M and S, respectively. Unilateral paired Wilcoxon tests indicated that treatment S was significantly higher than treatment C (p-value<0.0001) and treatment M (p-value=0.015), but that treatment M was not significantly higher than treatment C (p-value=0.15).

Impacts of MSP on seedling survival in interaction with summer drought

Our study showed that MSP enhances first-year seedling survival of species used to a large extent in European planted forests (sessile oak, Douglas fir, several pine species). It also showed that the increase in seedling survival following MSP is more pronounced when the summer is dry. These observations support our first and second hypotheses and agree with a large body of literature (Löf et al., 2012).

In years with rainy summers, early seedling survival in plots without MSP was high. MSP brought a slight but significant improvement and was similar for moderate and severe MSP. Seedling survival rates in plots without MSP decreased sharply as the amount of summer rain decreased, as previously shown by Zwolinski et al. (1994). In dry summers, seedlings planted after severe MSP had higher survival rates than those planted without MSP or with moderate MSP. All together, these observations partly invalidate our third hypothesis since severe MSP was more beneficial to seedling growth than moderate MSP only in dry summers, and they support our fourth hypothesis since the benefit from severe MSP, compared to moderate MSP was higher in years with summers drier than normal.

The novelty of our study was to analyze the interaction between MSP severity and drought intensity. In rainy years, moderate MSP resulted in the same seedling survival rates as severe MSP, whereas in dry years, moderate MSP led to the same rates as the control without MSP. The interaction we observed between MSP severity and drought intensity suggests that moderate and severe MSP not only differ in the severity of their impact on the seedling environment but also in the type of impact they induce. MSP may improve seedling survival through the reduction of competition with neighboring vegetation, insect or rodent damage, soil compaction or soil water logging, and the various MSP methods differ in their impacts on these constraints (Löf et al., 2012). The relative importance of the constraints for seedling survival depend on the climatic conditions, soil characteristics and biotic interactions prevailing in the plantation site, and the impact of a MSP method on seedling survival in a specific set of environmental conditions may be indicative of which constraints are primarily affected by the MSP method.

In our study, moderate MSP was effective at reducing constraints that limited seedling survival in years with rainy summers (competition for nutrients and light from neighboring plants, biotic damage mediated by neighboring plants, allelopathy caused by neighboring plants, soil constraints), but could not significantly reduce constraints that limited seedling survival in years with dry summers (the same constraints as above, plus competition for water from neighboring plants). Constraints not related to drought were probably limited, as suggested by the magnitude of the gain in survival brought by MSP in rainy years (+ 7%). These constraints probably also occurred in dry years as well but could not be detected because they were small compared to the high variability in survival observed in these years. This observation may explain the absence of statistically significant effects of moderate MSP in dry years, compared to the unprepared control. A larger number of observations would be required to test their effect.

Severe MSP was as effective as moderate MSP to reduce constraints that limited seedling survival in rainy years. In addition, severe MSP significantly increased survival rates in dry years, showing that severe MSP efficiently alleviated drought constraints (i.e., mainly competition for water from neighboring plants).

Methodological approach

Our study followed an experimental approach to analyze the combined effects of MSP and drought intensity on seedling survival. In the place of an experimental network, previous studies (Boutte et al., 2023; Tallieu et al., 2022) used large-scale inventories of forest plantations managed by forest operators to examine the same effects as our study on a similar geographical area (entire France), a similar time period (2017–2021) and a list of tree species and MSP methods that included the species and MSP methods tested in our study. However, although they used a large dataset (more than 10,000 plantation sites), they did not observe any significant effects of MSP on seedling survival, once potential confounding factors (species, other silvicultural operations, geographical area) were taken into account in the models. In this work, the lack of response was ascribed to (1) correlations between the predictive factors throughout the dataset since, in practical forestry, MSP methods are selected according to the silvicultural context; and (2) strong variability in seedling survival observed within each level of the predictive factors. The experimental approach used in our study made it possible to overcome these limitations: (1) we balanced the tree species and the MSP methods along the geographical gradient explored by the experimental network, which successfully led to weak correlations among our predictive factors; and (2) the establishment of an unprepared control in each experimental site made it possible to pair MSP and control treatments, which absorbed a strong proportion of the observed intra-treatment variability in seedling survival. The approach brought a high statistic power to the analysis and allowed us to observed significant effects of MSP in interaction with drought intensity, using a relatively small number of experimental sites.

As expected from previous work in the field of disturbance ecology (Peterson and Anderson, 2009; Royo et al., 2016), the severity of impact was successfully used to compare disturbances (in our study: MSP methods) with distinctive features. We analyzed nine MSP methods that differed in the type of tool and prime mover used, and in the type and the spatial extent of the preparation performed. The severity index made it possible to synthesize their features into a single descriptor that could successfully predict the effects of MSP on planting success.

We selected the cut-off value to distinguish between moderate and severe MSP methods in order to obtain a sufficient number of observations in each severity class, for the sake of the statistical analysis, without any ecological or technical consideration. Although our study led to significant results, further work could be performed to test different cut-off values and to examine their ecological or technical significance in order to improve our global understanding of the processes that underlie seedling response to MSP.

Although MSP is known to affect several environmental factors in the immediate seedling vicinity (e.g., water, light and nutrient availability, soil features, susceptibility to biotic damage), we used a single indicator to estimate MSP severity, which was the vegetation cover one year after MSP. In the field of disturbance ecology, combining several indicators that reflect the different ecosystem processes affected by disturbances provides a consistent framework for analyzing, comparing and modeling the effects of disturbances on ecosystems (Roberts, 2007). Although different indicators are usually positively correlated along gradients of increasing severity, substantial variations may occur among indicators, and combining several indicators may help elucidate the factors and mechanisms that underlie ecosystem responses to disturbance (Roberts, 2004).

In our study, we used a single indicator of MSP severity: vegetation cover in the immediate seedling vicinity one year after MSP. It performed well because (1) following MSP, vegetation cover positively correlates with other indicators of MSP intensity such as the extent of soil disturbance (Chaves Cardoso et al., 2020); (2) vegetation control is one of the primary effects of MSP (Löf et al., 2012); (3) in our study, vegetation cover could be estimated with good accuracy; and (4) in our study, we explored a large gradient of vegetation cover values (ranging from 3 to 100%). In a preliminary analysis, we also used the amount of soil disturbed by the MSP method (estimated through the maximum soil depth reached by the MSP tool and through the surface area of soil disturbed), as suggested by the previous study of Sikstrom et al. (2020). However, the soil indicators ranged along a small gradient of soil disturbance values (due to the selection of MSP methods in our study) and were difficult to estimate accurately (due to the measurement methodology in our study). Eventually, the soil disturbance indicators did not further discriminate between MSP methods, and we eliminated them from further analysis and instead used a single indicator based on vegetation cover. Selecting a different set of MSP methods that showed more pronounced differences in soil disturbance could possibly have led to different conclusions on the efficiency of the soil indicators used to estimate MSP severity.

Recommendations for management

We observed a strong decrease in seedling survival when summer drought increased, which is in full agreement with the long-term monitoring of plantation success in France (Boutte et al., 2023; Tallieu et al., 2022). In our study, we were not able to observe species-specific responses due to the small number of experimental sites available but, in their long-term monitoring, Tallieu et al. (2022) showed that some species (including Quercus petraea and Pseudotsuga menziesii.) are very responsive to summer drought, while others (including Pinus sylvestris, Pinus nigra and Pinus pinaster) show constant survival rates irrespective of summer drought. These observations provide robustness to our results since our results apply to tree species that show a wide range of responses to summer drought. Similarly, due to the limited number of experimental sites, we were not able to observe any impacts of vegetation or soil type, although these factors are known to have primary effects on seedling response to MSP (South et al., 2023), and forest managers should use their own expert knowledge to translate the results of our study to specific site conditions.

In many European countries, 1-year-survival is used as a criterion to evaluate planting success, with a threshold value set between 80 and 90% (Mataruga et al., 2023). In France, in most reforestation contracts, a threshold value of 80% seedling survival is used, below which the company that was responsible for the plantation is committed to paying for all refilling operations. We will use this threshold value as a basis to discuss practical recommendations based on our results.

In our study, the average survival rate in plots without MSP was 50% in dry years, which is well below the threshold value of 80%. In rainy years, seedling survival was above the threshold of 80% in most plots without MSP (13 out of 17 plots). All together, these observations reaffirm that the main issue for forest managers in France is to ensure plantation success during dry years for species that are most susceptible to summer drought, in order to avoid the extra costs due to plantation refill.

In rainy years, a positive and statistically significant effect of MSP on seedling survival was observed, irrespective of MSP severity. However the magnitude of the effect was small (+ 7%) and its ecological or practical significance (sensus Dixon and Pechmann (2005) and Wei et al. (2015)) is questionable. Defining a threshold value above which the effect of MSP may be considered as practically significant requires expert knowledge and data on long-term plantation dynamics and economic value that were not available for our study. Previous studies performed on Pinus taeda L. (South et al., 2023) suggest that the benefit of MSP during the rainy years that we observed in our study was too small to substantially increase stand value at harvest and, therefore, could be considered as negligible.

In dry years, severe MSP made it possible to significantly increase seedling survival compared to plots without MSP, and the magnitude of the effect (+ 30%) is large and, most probably, significant for the forest practice. Only three plots out of 11 that received severe MSP showed a survival rate below the threshold value of 80%. In practical forestry, these three plots would be refilled, and the eight remaining plots would avoid the expense of refilling operations. In comparison, eight out of nine plots in the control and seven out of eight plots in the moderate MSP treatment were below the threshold value and would need to be refilled. Applying a severe MSP method reduces the number of plantations to be refilled after a dry summer by 60%, compared to applying a moderate or no MSP. However, we can observe that, even if severe MSP significantly improves seedling survival, it does not guarantee obtaining a survival rate above the 80% threshold in dry years.

MSP has a financial cost that must be balanced against its beneficial effect on seedling survival. In our study, severe MSP methods used tools mounted on medium sized (7 to 8 tons) or large (20 to 22 tons) excavators, whereas moderate MSP methods (except for treatment #2) used tools pulled by tractors. Using an excavator instead of a tractor approximately doubles the working time (Puyal et al., 2022) and the financial cost (Suadicani, 2003) of the preparation. In rainy years, there is no benefit to using excavators that perform severe MSP instead of tractors that perform moderate MSP because they are more costly and do not improve early seedling survival. On the contrary, in dry years the benefit of using excavators may be high because they increase seedling survival, although they are more costly. In each reforestation project, the MSP method is chosen in the autumn or the winter prior to planting, whereas the potential benefit on seedling survival can only be seen several months later during the summer that follows planting. Determining, prior to planting, whether a MSP is required and choosing the most appropriate MSP method for each plantation project would require a risk analysis that takes into account (1) the probability of summer drought occurrence; (2) the probability of seedling survival according to drought intensity and the MSP method; (3) the cost of the MSP methods; and (4) their long-term economic return as a function of initial seedling survival. A research effort to gather such data is presently required for most plantation systems in order to develop decision support tools that make it possible to choose the best MSP method in each plantation site. Meanwhile, before such practical tools are available, forest managers may secure their plantation projects in dry and in rainy years by applying a severe MSP or by opting for an insurance contract against plantation failure (Bréteau-Amores et al., 2023).

Finally, there is presently a growing emphasis on the maintenance of soil fertility and biodiversity, prompting forest managers to select silvicultural methods that preserve both soil and biodiversity (Duncker et al., 2012), and we also need to consider environmental impacts of MSP, when recommending appropriate MSP methods. In our study, MSP severity was assessed in the immediate seedling vicinity (1-m2-subplots surrounding the seedlings), which corresponds to the rooting zone of the newly planted seedlings and the potentially competing neighboring plants. We found that MSP severity was related to seedling survival. We must stress that our results do not suggest that implementing a MSP method beyond the area immediately around the seedlings would further improve seedling survival. Therefore, implementing MSP on a larger surface area (along the planting line or throughout the entire stand area) should be decided cautiously, given its demonstrated strong and long-lasting environmental impact, particularly on soil structure and fertility (Collet et al., 2020; Sutinen et al., 2019). In this respect, intermittent or spotted MSP methods (Sikström et al., 2020) that implement a severe preparation around the planted seedlings while leaving the rest of the surface area in the stand undisturbed, appear to be a good option for forester managers looking to ensure high seedling survival while simultaneously preserving the environment.

Competing Interest information

The authors have no competing interests to declare that are relevant to the content of this article.

Acknowledgements

The authors heartily thank Léon Wehrlen, Sébastien Harnist, Lindsay Godard and Erwin Thirion (INRAE), Didier François, Gwénaëlle Gibaud, Jérôme Piat, Quentin Girard, Lucie Arnaudet, Erwin Ulrich and Claudine Richter (ONF), Jean-Yves Fraysse, Mathilde Duverger and Xavier Montagny (FCBA), and Christophe Vidal (IDF-CNPF) for their invaluable help when designing, establishing and running the 20 experimental sites.

Author Contribution information

All authors contributed to the study conception and design, to material preparation and data collection. Analyses were performed by CC. The first draft of the manuscript was written by CC, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the French Ministry in charge of forests (“Soutien au Pôle Renfor” annual grant), France Bois Forêt (grants 16RD496 and 17RD696), the Grand-Est region (“IPLor” grant), and the Fonds National pour l’Aménagement du Territoire (“IPLor” grant).

Data availability

The datasets generated during the current study are available from the corresponding author upon reasonable request.

Boutte, B., Husson, C., Saintonge, F.X., 2023. Quelle a été l’évolution des taux de succès des plantations de l’année au cours des quinze dernières années ? in: Expertise collective CRREF «Coupes Rases et Renouvellement des peuplements Forestiers en contexte de changement climatique », Rapport scientifique de l’expertise, editors: Landmann G et al. Paris : GIP ECOFOR, RMT AFORCE (mai 2023) 624–635.
Bréteau-Amores, S., Brunette, M., Andrés-Domenech, P., 2023. A Cost Assessment of Tree Plantation Failure under Extreme Drought Events in France: What Role for Insurance? Forests 14, 308. https://doi.org/10.3390/f14020308
Buitrago, M., Paquette, A., Thiffault, N., Bélanger, N., Messier, C., 2014. Early performance of planted hybrid larch: effects of mechanical site preparation and planting depth. New Forests 46, 319–337. https://doi.org/10.1007/s11056-014-9463-3
Burton, P., Bedford, L., Goldstein, M., Osberg, M., 2000. Effects of disk trench orientation and planting spot position on the ten-year performance of lodgepole pine. New Forests 20, 23–44.
Chaves Cardoso, J., Burton, P.J., Elkin, C.M., 2020. A Disturbance Ecology Perspective on Silvicultural Site Preparation. Forests 11, 1278. https://doi.org/10.3390/f11121278
Chirino, E., Vilagrosa, A., Cortina, J., Valdecantos, A., Fuentes, D., Trubat, R., Luis, V.C., Puértolas, J., Bautista, S., Baeza, J.L., Peñuelas, J.L., Vallejo, V.R., 2009. Ecological restoration in degraded drylands: the need to improve the seedling quality and site conditions in the field. in: Forest management. Editor: Grosberg, S.P., Nova Science Publishers, Inc. 85–158.
Cole, E., Lindsay, A., Newton, M., Bailey, J.D., 2018. Eight-year performance of bareroot Douglas-fir and bareroot and plug western larch seedlings following herbicide applications, northeast Oregon, USA. New Forests 49, 791–814. https://doi.org/10.1007/s11056-018-9653-5
Collet, C., Vast, F., Richter, C., Koller, R., 2020. Cultivation profile: a visual evaluation method of soil structure adapted to the analysis of the impacts of mechanical site preparation in forest plantations. Eur J Forest Res. https://doi.org/10.1007/s10342-020-01315-2
Cribari-Neto, F., Zeileis, A., 2010. Beta regression in R. Journal of Statistical Software 1–24.
Del Campo, A.D., Segura-Orenga, G., Bautista, I., Ceacero, C.J., González-Sanchis, M., Molina, A.J., Hermoso, J., 2021. Assessing reforestation failure at the project scale: The margin for technical improvement under harsh conditions. A case study in a Mediterranean Dryland. Science of The Total Environment 796, 148952. https://doi.org/10.1016/j.scitotenv.2021.148952
del Campo, A.D., Segura-Orenga, G., Ceacero, C.J., González-Sanchis, M., Molina, A.J., Reyna, S., Hermoso, J., 2020. Reforesting drylands under novel climates with extreme drought filters: The importance of trait-based species selection. Forest Ecology and Management 467, 118156. https://doi.org/10.1016/j.foreco.2020.118156
del Campo, A.D., Segura-Orenga, G., Molina, A.J., González-Sanchis, M., Reyna, S., Hermoso, J., Ceacero, C.J., 2022. On the Need to Further Refine Stock Quality Specifications to Improve Reforestation under Climatic Extremes. Forests 13, 168. https://doi.org/10.3390/f13020168
Dixon, P.M., Pechmann, J.H.K., 2005. A statitical test to show negligible trend. Statistical Reports 86, 1751–1756.
Dumas, N., Dassot, M., Pitaud, J., Piat, J., Arnaudet, L., Richter, C., Collet, C., 2021. Four-year-performance of oak and pine seedlings following mechanical site preparation with lightweight excavators. Silva Fenn. 55. https://doi.org/10.14214/sf.10409
Duncker, P.S., Raulund-Rasmussen, K., Gundersen, P., Katzensteiner, K., De Jong, J., Ravn, H.P., Smith, M., Eckmüllner, O., Spiecker, H., 2012. How Forest Management affects Ecosystem Services, including Timber Production and Economic Return: Synergies and Trade-Offs. Ecology and Society 17. https://doi.org/10.5751/ES-05066-170450
Frelich, L.E., 2002. Forest dynamics and disturbance regimes. Cambridge University Press, 266 p.
Grossnickle, S., 2018. Seedling establishment on a forest restoration site – An ecophysiological perspective. REFOR 110–139. https://doi.org/10.21750/REFOR.6.09.62
Grossnickle, S.C., Ivetić, V., 2022. Root system development and field establishment: effect of seedling quality. New Forests 53, 1021–1067. https://doi.org/10.1007/s11056-022-09916-y
Hjelm, K., Nilsson, U., Johansson, U., Nordin, P., 2019. Effects of mechanical site preparation and slash removal on long-term productivity of conifer plantations in Sweden. Can. J. For. Res. 49, 1311–1319. https://doi.org/10.1139/cjfr-2019-0081
IPCC, 2023. Summary for Policymakers. In: Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, H. Lee and J. Romero (eds.)]. IPCC, Geneva, Switzerland, pp. 1-34, doi: 10.59327/IPCC/AR6-9789291691647.001.
Iwasaki, A., Noda, T., 2018. A framework for quantifying the relationship between intensity and severity of impact of disturbance across types of events and species. Sci Rep 8, 795. https://doi.org/10.1038/s41598-017-19048-5
Keenan, R.J., 2015. Climate change impacts and adaptation in forest management: a review. Annals of Forest Science 72, 145–167. https://doi.org/10.1007/s13595-014-0446-5
Kurth, V.J., D’Amato, A.W., Bradford, J.B., Palik, B.J., Looney, C.E., 2020. Assessing the ecological impacts of biomass harvesting along a disturbance severity gradient. Ecological Applications 30, e02042. https://doi.org/10.1002/eap.2042
Lindner, M., Fitzgerald, J.B., Zimmermann, N.E., Reyer, C., Delzon, S., van der Maaten, E., Schelhaas, M.-J., Lasch, P., Eggers, J., van der Maaten-Theunissen, M., Suckow, F., Psomas, A., Poulter, B., Hanewinkel, M., 2014. Climate change and European forests: What do we know, what are the uncertainties, and what are the implications for forest management? Journal of Environmental Management 146, 69–83. https://doi.org/10.1016/j.jenvman.2014.07.030
Löf, M., Dey, D., Navarro, R.M., Jacobs, D.F., 2012. Mechanical site preparation for forest restoration. New Forests 43, 825–848.
Luoranen, J., Riikonen, J., Saksa, T., 2023. Damage caused by an exceptionally warm and dry early summer on newly planted Norway spruce container seedlings in Nordic boreal forests. Forest Ecology and Management 528, 120649. https://doi.org/10.1016/j.foreco.2022.120649
Mataruga, M., Cvjetković, B., De Cuyper, B., Aneva, I., Zhelev, P., Cudlín, P., Metslaid, M., Kankaanhuhta, V., Collet, C., Annighöfer, P., Mathes, T., Marianthi, T., Despoina, P., Jónsdóttir, R.J., Cristina Monteverdi, M., de Dato, G., Mariotti, B., Dina Kolevska, D., Lazarević, J., Sundheim Fløistad, I., Klisz, M., Gil, W., Paiva, V., Fonseca, T., Nicolescu, V.-N., Popović, V., Devetaković, J., Repáč, I., Božič, G., Kraigher, H., Andivia, E., Diez, J.J., Böhlenius, H., Löf, M., Bilir, N., Villar-Salvador, P., 2023. Monitoring and control of forest seedling quality in Europe. Forest Ecology and Management 546, 121308. https://doi.org/10.1016/j.foreco.2023.121308
Mattsson, S., Bergsten, U., 2003. Pinus contorta growth in northern Sweden as affected by soil scarification. New Forests 26, 217–231. https://doi.org/10.1023/A:1024425205712
McKee, T.B., Doesken, N.J., Kleist, J., n.d. The relationship of drought frequency and duration to time scales. Eighth Conference on Applied Climatology, 17-22 january 1993, Anaheim, California, 6p.
Padilla, F.M., Pugnaire, F.I., 2007. Rooting depth and soil moisture control Mediterranean woody seedling survival during drought. Functional Ecology 21, 489–495. https://doi.org/10.1111/j.1365-2435.2007.01267.x
Pausas, J.G., Bladé, C., Valdecantos, A., Seva, J.P., Fuentes, D., Alloza, J.A., Vilagrosa, A., Bautista, S., Cortina, J., Vallejo, R., 2004. Pines and oaks in the restoration of Mediterranean landscapes of Spain: New perspectives for an old practice – a review. Plant Ecology 171, 209–220.
Peters, D.P.C., Lugo, A.E., Chapin, F.S., Pickett, S.T.A., Duniway, M., Rocha, A.V., Swanson, F.J., Laney, C., Jones, J., 2011. Cross-system comparisons elucidate disturbance complexities and generalities. Ecosphere 2, art81. https://doi.org/10.1890/ES11-00115.1
Peterson, C.E., Anderson, P.D., 2009. Large-scale interdisciplinary experiments inform current and future forestry management options in the U.S. Pacific Northwest. Forest Ecology and Management 258, 409–414. https://doi.org/10.1016/j.foreco.2008.12.029
Puyal, M., Agro, C., Collet, C., Dumas, N., Pitaud, J., Vidal, C., 2022. Productivité horaire de la préparation mécanisée du site avant plantation : premiers résultats du réseau Pilote. Internal Report INRAE, 66p.
Querejeta, J.I., Roldan, A., Albaladejo, J., Castillo, V., 2001. Soil water availability improved by site preparation in a Pinus halepensis afforestation under semiarid climate. Forest Ecology and Management 149, 115–128.
R Core Team, 2023. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. <https://www.R-project.org/>.
Roberts, M.R., 2007. A conceptual model to characterize disturbance severity in forest harvests. Forest Ecology and Management 242, 58–64. https://doi.org/10.1016/j.foreco.2007.01.043
Roberts, M.R., 2004. Response of the herbaceous layer to natural disturbance in North American forests. Can. J. Bot. 82, 1273–1283. https://doi.org/10.1139/b04-091
Royo, A.A., Peterson, C.J., Stanovick, J.S., Carson, W.P., 2016. Evaluating the ecological impacts of salvage logging: can natural and anthropogenic disturbances promote coexistence? Ecology 97, 1566–1582. https://doi.org/10.1890/15-1093.1
Sikström, U., Hjelm, K., Holt Hanssen, K., Saksa, T., Wallertz, K., 2020. Influence of mechanical site preparation on regeneration success of planted conifers in clearcuts in Fennoscandia – a review. Silva Fenn. 54. https://doi.org/10.14214/sf.10172
Soubeyroux, J.M., Kitova, N., Blanchard, M., Vidal, J.P., Martin, E., Dandin, P., 2012. Sécheresses des sols en France et changement climatiqu. Résultats et applications du projet ClimSec. La météorologie 78, 21–30.
South, D.B., Starkey, T.E., Lyons, A., 2023. Why Healthy Pine Seedlings Die after They Leave the Nursery. Forests 14, 645. https://doi.org/10.3390/f14030645
Stanturf, J.A., Palik, B.J., Dumroese, R.K., 2014. Contemporary forest restoration: A review emphasizing function. Forest Ecology and Management 331, 292–323. https://doi.org/10.1016/j.foreco.2014.07.029
Stanturf, J.A., Schoenholtz, S.H., Schweitzer, C.J., Shepard, J.P., 2001. Achieving restoration success: myths in bottomland hardwood forests. Restoration Ecology 9, 189–200.
Suadicani, K., 2003. Site preparation and planting in a Picea abies shelterwood stand. Scandinavian Journal of Forest Research 18, 247–259.
Sutinen, R., Gustavsson, N., Hänninen, P., Middleton, M., Räisänen, M.L., 2019. Impact of mechanical site preparation on soil properties at clear-cut Norway spruce sites on mafic rocks of the Lapland Greenstone Belt. Soil and Tillage Research 186, 52–63. https://doi.org/10.1016/j.still.2018.10.013
Sutton, R.F., 1993. Mounding site preparation: a review of European and North American experience. New Forests 7, 151–192.
Tallieu, C., Collet, C., Renaud, J.P., Pitaud, J., 2022. Conception d’indices météorologiques pour prendre en compte le risque de sécheresse estivale dans la garantie de reprise des plantations. Internal report INRAE, 70p.
von der Gönna, M.A., 1990. Fundamentals of mechanical site preparation (No. FRDA report No. 178). British Columbia Ministry of Forests. 29 p.
Wei, L., Villemey, A., Hulin, F., Bilger, I., Yann, D., Chevalier, R., Archaux, F., Gosselin, F., 2015. Plant diversity on skid trails in oak high forests: A matter of disturbance, micro-environmental conditions or forest age? Forest Ecology and Management 338, 20–31. https://doi.org/10.1016/j.foreco.2014.11.018
Zwolinski, J., South, D.B., Barber, B.L., 1994. Drought and survival of loblolly pine seedlings after planting. Eighth Biennal Southern Silvicultural Resarch Conference, Auburn, AL, Nov. 1-3. 419–423.

Table 1

Site characteristics: site number; geographical location (latitude and longitude were blurred and rounded to the nearest degree); MSP treatments applied; year indicates the first growing season after plantation; Standardized Precipitation Index (SPI) computed over June, July and August of the same year; planted species. MSP treatment codes are: 1: disk pulled by a tractor; 2: scalper mounted on an excavator followed by a ripper-disk trencher pulled by a tractor; 3: ripper pulled by a tractor; 4: plow pulled by a tractor; 5: ripper-scarifier mounted on an excavator; 6: ripper-scarifier-disk trencher mounted on an excavator; 7: scalper mounted on an excavator followed by a ripper-scarifier mounted on an excavator; 8: scarifier mounted on an excavator; 9: scarifier mounted on an excavator followed by a ripper-scarifier mounted on an excavator).
		MSP treatment
Site number	Location	Moderate severity	Strong severity	Year	SPI	Species
1	49°N 1°E	3	5	2014	2,01	Quercus petraea
2	49°N 1°E	/	9	2011	1,21	Pinus nigra Quercus petraea
3	47°N 6°E	2	8	2014	1,8	Quercus petraea
4	47°N 0°E	2	7	2015	0,47	Pinus pinaster
5	49°N 3°E	3	5	2014	1,9	Quercus petraea
6	49°N 3°E	3	5	2014	1,91	Quercus petraea
6	49°N 3°E	3	5	2016	0,1	Quercus petraea
7	44°N 1°W	4	9	2012	-0,8	Pinus pinaster Quercus robur
8	48°N 3°E	3	8	2014	1,82	Quercus petraea
9	48°N 6°E	3	5	2014	0,71	Pseudotsuga menziesii
10	49°N 8°E	1	9	2012	1,14	Pinus sylvestris
10	49°N 8°E	1	9	2013	-0,56	Quercus petraea
11	49°N 8°E	1	9	2012	1,14	Pinus sylvestris
11	49°N 8°E	1	9	2013	-0,56	Quercus petraea
12	49°N 8°E	1	9	2012	1,14	Pinus sylvestris
12	49°N 8°E	1	9	2013	-0,57	Quercus petraea
13	49°N 7°E	/	6	2019	-1,71	Quercus petraea
14	46°N 2°E	/	5	2017	-0,35	Pseudotsuga menziesii
15	48°N 2°W	1	9	2012	0,57	Pinus sylvestris Quercus petraea
16	48°N 2°W	2	5	2016	-0,26	Pinus pinaster
17	44°N 1°W	4	9	2012	-0,79	Quercus robur
18	49°N 6°E	3	6	2018	-1,17	Pinus sylvestris
18	49°N 6°E	3	6	2019	-1,23	Pinus sylvestris
19	48°N 2°W	1	9	2012	0,47	Pinus sylvestris Quercus petraea
20	47°N 4°E	3	5	2018	-0,45	Pseudotsuga menziesii
20	47°N 4°E	3	5	2019	-1,81	Pseudotsuga menziesii

No competing interests reported.

Mechanical site preparation severity mediates one-year-survival response to summer drought in planted tree seedlings

Status:

Version 1

Abstract

Figures

Introduction

Material And Methods

Results

Discussion

Declarations