Interactive influences of resource pulse quality and abruptness on aquatic ecosystem

Interactive influences of resource pulse quality and abruptness on aquatic ecosystem | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 17 July 2025 V1 Latest version Share on Interactive influences of resource pulse quality and abruptness on aquatic ecosystem Authors : Charlie Sarran 0009-0000-6573-7108 [email protected] , Beatrix Beisner , and Eric Harvey 0000-0002-8601-7326 Authors Info & Affiliations https://doi.org/10.22541/au.175273206.63627240/v1 323 views 149 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf Many ecosystems worldwide are interconnected by spatial pulses of energy and material capable of structuring and defining their main processes. With global anthropogenic change, the dynamics of such fluxes (i.e., resource type, magnitude and duration) will be modified. Yet, it is still unclear how simultaneous alterations in flux dynamics might interact to influence ecosystem functioning. Here, we provide experimental evidence for interactive effects of resource pulse quality and abruptness level (i.e., magnitude/duration) on aquatic ecosystem functions using experimental mesocosms. Mesocosms were exposed to three resource types (insects, leaves and a mix of both) based on three abruptness levels (high, medium, low) in a 3 x 3 factorial design. With an equal overall amount of resource added, nutrient inputs from insect decomposition spread quickly through ecosystem compartments, supporting the development of diverse autotrophic primary producers and increasing ecosystem stocks. In contrast, leaf decomposition confined nutrient flows within the benthic food web compartment, probably favoring more heterotrophic organisms linked to the decomposition of the vegetal matrix. Finally, our study also revealed that ecological responses arising from insect pulse are more sensitive to changes in abruptness variation – pulsed event of great magnitude over a short period of time leading to higher nutrient flow between ecosystem compartments, increasing the ecosystem impact of nutrient enrichment. Thus, considering both quality and abruptness level of allochtonous resource pulse is important in the understanding of ecological processes changes under such biomass-altering disturbance events. TITLE: Interactive influences of resource pulse quality and abruptness on aquatic ecosystem ABSTRACT Many ecosystems worldwide are interconnected by spatial pulses of energy and material capable of structuring and defining their main processes. With global anthropogenic change, the dynamics of such fluxes (i.e., resource type, magnitude and duration) will be modified. Yet, it is still unclear how simultaneous alterations in flux dynamics might interact to influence ecosystem functioning. Here, we provide experimental evidence for interactive effects of resource pulse quality and abruptness level (i.e., magnitude/duration) on aquatic ecosystem functions using experimental mesocosms. Mesocosms were exposed to three resource types (insects, leaves and a mix of both) based on three abruptness levels (high, medium, low) in a 3 x 3 factorial design. With an equal overall amount of resource added, nutrient inputs from insect decomposition spread quickly through ecosystem compartments, supporting the development of diverse autotrophic primary producers and increasing ecosystem stocks. In contrast, leaf decomposition confined nutrient flows within the benthic food web compartment, probably favoring more heterotrophic organisms linked to the decomposition of the vegetal matrix. Finally, our study also revealed that ecological responses arising from insect pulse are more sensitive to changes in abruptness variation – pulsed event of great magnitude over a short period of time leading to higher nutrient flow between ecosystem compartments, increasing the ecosystem impact of nutrient enrichment. Thus, considering both quality and abruptness level of allochtonous resource pulse is important in the understanding of ecological processes changes under such biomass-altering disturbance events. KEYWORDS: biomass-altering disturbances, pulse dynamic, nutrient flow, allochthonous perturbation, ecosystem processes, cascading impacts jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf 1 | INTRODUCTION Ecosystems are connected by spatial flows of energy and materials (i.e., through active organismal movement and/or passive flows of inorganic materials), playing significant roles in maintaining their structures and processes 1–3 . These exchanges are often discrete in time and can be considered as pulsed disturbance events when leading to biotic or abiotic changes in resource supply 4 . Such resource pulses can vary in quality (i.e., nutrient content and matrix degradability), frequency, magnitude and duration – attributes particularly relevant in the context of global anthropogenic changes characterized by shifts in resource distribution (land-use), phenology (seasonality), and an increased variance of climactic events 5,6 . Despite evidence regarding the importance of allochthonous resource diversity (i.e., quality) and pulse dynamics (i.e., quantity and magnitude of the disturbance) for connected ecosystems worldwide, it is still largely unclear how simultaneous alterations to their qualitative and quantitative dynamics might interact to influence ecosystem functions 3,7,8 . Terrestrial and freshwater ecosystems are highly connected across biological levels of organization, integrating both direct and indirect biotic and abiotic interactions. They are thus considered as useful meta-ecosystem models 9,10 . Previous studies have experimentally tested some facets of the dynamical aspect of cross-ecosystem fluxes of resources. Experimentally examining the effect of the resource quality (“poor” terrestrial leaves versus “high” terrestrial insects), Klemmer et al. (2020) found that, compared to leaf incorporation alone, insect addition increased the total biomass of emerging aquatic insects, an effect apparently dampened by the concomitant addition of leaves 11 . As such, they demonstrated that resource quality can modulate the reciprocal and cascading flow of energy and matter within and between adjacent ecosystems – and that joint additions of two types of resources did not necessarily result in an additive or linear response. Another study found that an increase in pulse magnitude of common carp ( Cyprinus carpio ) carrion (defined as an increase in the total biomass of the pulse) could drive shallow lakes across alternative equilibria 12 . In that study, ecosystem responses to the pulse gradient generally led to both an increase in nutrient availability and primary productivity (while decreasing water clarity) – making the magnitude criteria an important factor influencing shallow eutrophic lakes 12 . Moreover, in one of the first experimental tests on the effects of single resource pulsed events duration on stream ecosystem mesocosm (i.e., considering the same total amount of terrestrial insect biomass input), a prolonged terrestrial insect resource pulse increased overall predation pressure by salmons on benthic invertebrates compared to a shorter event - thus reducing, via a trophic cascade, leaf breakdown rate 13 . Considered broadly, this finding demonstrated that the duration of such resource-pulse disturbance constitute itself an important factor for consumer population structure, community dynamics and ecosystem parameters within the recipient ecosystems. Therefore, these past studies showed how resource pulse duration, magnitude and quality each have significant implications for the receiving ecosystems. What remains largely unexplored however is how these allochthonous pulse attributes might interact to influence recipient ecosystem structure and dynamics. This is especially important in a context where human-induced disturbances are expected to globally transform land-use (resource quality/quantity), as well as the timing and duration of seasonal pulse (climate change). Theoretical works also provide clear evidences on the significance of pulse attributes for ecosystem dynamics 3,4,14 . Indeed, interactions between pulse attributes (e.g., between resource pulse quality and magnitude) may lead to additive or multiplicative ecological responses, themselves likely dependant on the temporal scaling of the pulse (relative to seasonal or phenological timing) 4 . In order to integrate those different attributes of resource pulse events within a single experimental design, we propose here to investigate the interactive effects of pulse quality and abruptness (magnitude/duration). To explore such interactions, we conducted an outdoor mesocosm experiment over four months wherein we fully crossed two disturbance attributes – the resource type (vegetal, animal and a combination of both) as our qualitative measure and the level of abruptness (“High”, “Medium” and “Low”), keeping equal the total amount of dried biomass added in each mesocosm. We hypothesized that the greater the resource pulse quality, the greater the impact on the recipient ecosystem (pelagic and benthic compartments). We predicted those impacts to be characterised as an increase of ecosystem stocks and processes compared to controls (such as a global increase in nutrient concentrations, primary production and respiration levels). Consequently, such increase in resource quality would lead the receiving aquatic ecosystem to approach, or even enter, a more productive state. On the other hand, according to previous theoretical predictions, we hypothesize that an increase in pulse abruptness would also increase the magnitude and rate of change in the water column (i.e., water chemistry, e.g., nutrients) – thus leading to a greater departure from unmanipulated controls and a decrease in the ecosystem recovery rate. Indeed, a greater pulse magnitude over a shorter period would increase the overall rate of nutrient released from the resource within the benthic system (consequence of more active biotic and passive abiotic processes) – thus leading to a stronger and variable benthic to pelagic flow at a given time. 2 | METHODS 2.1 | Experimental design of resource pulse dynamic To test the joint effects of terrestrial resource type quality (i.e., nutrient content and matrix accessibility) and abruptness on freshwater lentic ecosystems, a fully crossed 3 x 3 factorial design experiment was conducted. The aquatic mesocosm experiment consisted of three levels of allochthonous resource type (insect, plant and both) and three levels inputs (abruptness levels “Low”, “Medium” and “High”, details in Fig. 1) – for a total of nine treatments and one control to which no additions were made. Each treatment was replicated four times and situated within four blocks of ten mesocosms following a randomized complete block design (RCBD). Based on earlier trial runs, we set the total amount of dried biomass of each type over time added to each mesocosm at 40g. Across the treatments, all mesocosms received the same total resource biomass over the entire experiment but of a differing duration, dependent on the abruptness treatment, which necessarily led to a concurrent variation in magnitude (magnitude/duration = abruptness, see Fig. 1). To ensure a similar seasonality to the resource inputs, the peak timing of the different abruptness treatments was the same occurring in week 5 (Fig. 1). We used dried, commercially available insect larvae ( Tenebrio molitor ) as a proxy of high quality and easy degradable matrix resources. For a poorer and lower quality resource we collected and dried leaves of Acer saccharum during leaf fall of 2021 directly on site. As an intermediate type of resource, the “Mix” treatment was composed of an equal dried biomass of both leaves and insects. Pulse events occurred weekly on Wednesdays around 18h00 according to their abruptness levels (Fig. 1). Resources were weighted (5g, 10g or 20g) using an A&D Newton EJ-610 scale and placed in distilled water (250mL/5g of matter) 24 hours before addition to the mesocosms. This way, each resource input was composed of hydrated biomass (leaves, insects or a mix of both) and their corresponding rehydration water containing some dissolved nutrients. We also measured total nitrogen content (TN) of each of the allochthonous resources (leaves and insects) in their dried, original solid form. TN was measured by gas chromatography and electro-thermal detection using a FISONS EA1108 Elemental Analyser (Thermo-Fisher Scientific, USA). 2.2 | Experimental setup The experiment was conducted at the Laurentian Biological Station (SBL), Quebec, Canada from June 29 to October 26, 2022 (17 weeks). On May 31st, 40 tanks of approximately 450 L within a floating platform on Triton Lake (45°59’17.1”N 74°00’23.7”W; see Fig. S1 for pictures) were filled with unfiltered littoral lake water (Fig. 2). This first step ensured an initial inoculation of pelagic organisms and suspended matter. To decrease variability between tanks, on June 10 and 14 we mixed 1L of water from each mesocosm and then reinoculated 1L of the mixture to each mesocosm. To enable connectivity with the local environment (e.g. rainfall, aquatic insect colonisation.), mesocosms were not covered during the experiment. To create a benthic community in each mesocosm, we added homogenized benthic material directly collected from the littoral zone of the Triton Lake to all tanks on June 9 & 13 (substrate sizes between 500um and 2mm). Even if this protocol allowed small odonates larvae to be inoculed in each mesocosm, we also added one dragonfly larvae of 20mm or larger, collected on site, to ensure at least one large top predator before the beginning of the experiment. We used custom-made artificial benthic habitats (Fig. S2) that were kept in the littoral zone of the lake to allow colonization over a 3-week period. These were then added at the bottom of each tank on June 7 th . To permit the assessment of benthic periphyton growth, a porcelain tile (10cm*30cm) was placed at the bottom of each tank on June 29th. Finally, for monitoring of several key ecosystem parameters, two dataloggers (Onset HOBO U26-001 Dissolved Oxygen/Temperature and MX2501 pH/Temperature) were placed at the center of each mesocosm. jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf 2.3 | Response variable measurements To assess the impacts of resource pulse dynamics on ecosystem processes, we sampled all mesocosms five times during the experimental period, in weeks 1, 5, 9, 12 and 17 (Fig. 1), over a week-long period. Each time, four types of data were collected: (1) nutrients and carbon, (2) in vivo chlorophyll a (Chl. a .) and turbidity, (3) periphyton biomass and (4) temperature, pH and dissolved oxygen. If occurring during a sampling week, to avoid the effects of precipitations on mesocosm water quality (e.g., impacting nutrients, chlorophyll and turbidity), we sampled data (1) and (2) at least 24h following a rainfall event. Unfiltered water was collected at the center of each mesocosm, approximately 10-15cm below the surface in acid-washed 125mL Nalgene HDPE bottles (for Total Nitrogen and Total Phosphorus – TN and TP, respectively) and acid washed 40mL amber glass vials (for Total Organic Carbon – TOC). Samples were stored in the dark at − 20 °C for TN and TP analyses and at 4 °C for TOC. Subsequently, TN concentrations were measured using a flow injection analyzer (Lachat QuikChem 8000), TP concentrations using a segmented flow analyzer (Astoria 2) and TOC concentrations through membrane conductometric detection using a Sievers M5310 C analyser (Veolia). Chl. a . and turbidity measures were both done at the same sampling depths and on the same day using a handheld fluorometer/turbidimeter (Turner Design AquaFluor®). To assess periphyton growth as an indicator of a benthic primary production, we scraped the top surface of each tile with a rubber spatula. Cleaned tiles were then returned to their respective mesocosms. Collected biomass was placed onto a pre-weighed coffee filter and then oven-dried for 24h at 30°C. Dried filters were weighed using a Sartorius Entris® II (BCE224I-1S) balance to estimate periphyton biomass. Water temperature, pH and dissolved oxygen (DO) were collected every 6-hours with the submersed sondes, starting at midnight, for the 17 weeks of experiment. To compare with other variables collected during the same week, the 6-hour frequency sonde data was time-averaged. Leaf decomposition rates was assessed in autumn, following the allochthonous input manipulations. For this, we placed one plastic leaf litter bag (≈ 500μm mesh size; 20g ± 0.02 of dried leaves of Acer saccharum ) in each mesocosm at the end of week 12. All litter bags were removed 33 days later and rinsed three times in clean water to remove particles and organisms smaller than the mesh size. Litter bags were then air (24h) and oven (24h at 30°C) dried to a constant mass (A&D Newton EJ-610). 2.4 | Data analysis All data manipulation and statistical analyses were performed using R version 4.2.2 (2022-10-31, R Core Team). 2.4.1 | Global dispersion of ecosystem processes Principal Components Analysis (PCA) on centered and reduced (i.e., correlation matrix) data from week 05 (middle disturbance) to week 17 (end of experiment) was performed to distinguish the most significant sets of measured variables responsible for the treatment dynamics and visualize these differences in a reduced space (through appreciation of main variables contribution to each axis; ‘prcomp’ function in the R package ‘stats’). We then used permutational multivariate analysis of variance (PERMANOVA; ‘adonis’ function from ‘Vegan’ R package) to test for significant differences between treatments (i.e., Resource type, Abruptness level, Time and crossed effects). Finally, to approximate and compare relative ecosystem variation between treatments, we assessed for differences in dispersion (distance to centroids) and estimated the spread of datapoints among group centroids through a test of homogeneity of multivariate dispersion (‘betadisper’ function from the ‘Vegan’ R package). 2.4.2 | Testing for treatment effects Linear Mixed-Effects Models (LMMs; ‘lmer’ function from the ‘lme4’ R package) with experimental blocks (9 treatments + 1 control, n=4 replicates) as random effect were used to test whether the temporal dynamics of primary production ( Chl. a . and periphyton biomass), weekly DO descriptors (using daily mean values and their associated standard errors (SE)), nutrient concentrations (TP, TN, TOC) and associated stoichiometric molar ratios (C:N, C:P and N:P) were affected by the treatments. For each model, asymptotic normality of errors was confirmed by a Q-Q plot and associated tests: Kolmogorov-Smirnov (KS) for uniformity, Dispersion for over/under-dispersion and Outlier for significant points deviations (‘simulateResiduals’ function from ‘DHARMa’ R package). Subsequently, though generally robust to violation of the normality assumption, some response variables in the models were log-transformed 15 . To quantify the proportion of variance explained by our statistical models, conditional (including both fixed and random effects) and marginal (considering only the fixed effects) R 2 values were calculated according to Nakagawa et al.(‘r2_nakagawa’ function from ‘performance’ R package) 16 . However, in our models, for all response variables, except for the TOC and Chl. a ., the random effect was undetectable (variance and Std. deviation = 0). These singularities – probably due to the negligible contribution of the random effect to the explained variance – meant that we could not obtain a conditional R 2 for these variables and therefore only the marginal R 2 is presented. Lastly, to obtain a global range of possible values revolving around the fixed effect estimates (slopes) for each model, we used the associated fixed effect standard errors (SE) to calculate 95% confidence intervals (CI) following this equation: CI=Estimate ±1.96 ×Std.Error. To assess divergence by treatment of the selected response variables at the end of the experiment (week 17), we performed an Analysis of Variance (ANOVA, ‘aov’ function from the ‘stats’ R package) to determine differences between group means (experimental blocks were included in the model to account for random effects). Following the same approach as for the LMMs, we systematically checked for normality and calculated R 2 (conditional and adjusted). However, because our goal was to focus solely on comparing treatment effects, we excluded the controls from the models, keeping them only as a visual reference in the figures. Finally, to assess the consequences of the treatments on leaf litter decomposition during autumn, we again used ANOVA to test the main and interaction treatment effects after 33 days of immersion. 3 | RESULTS 3.1 | Effect of allochthonous matter treatments Assessment of the TN content of the dried allochthonous resources demonstrated that lowest concentrations were found in the leaf material (0.006 g. g-1(dw)). Insect biomass had 13 times more TN content (0.087 g. g-1(dw)). The realized pulse magnitude (TN mass added each week) was included on a biplot along with the associated abruptness treatment to visualize the treatment ranges (‘pulsedness’ sensus Jentsch and White 2019) (Fig. 3A). This confirmed that the Insect treatment had a higher relative pulsedness range compared to the Leaf treatment, and with the mix treatment being intermediate. The first PCA axis (41.97% of explained variation), captured variables distribution after disturbance, from week 5 (mid-time point, by which each tank had received the same amount of material but with varying abruptness) to week 17 (end of experiment, 9 weeks after the last resource pulse). On the first PCA axis, nutrients and associated stoichiometric molar ratios best discriminated the treatments (Fig.3B; loadings for TN = 0.393 and TP = 0.375, C:P = -0.359 and C:N = -0.346; see Fig. S3 for details). Weekly DO descriptors were best correlated with the second PCA axis (loadings for DO mean=-0.604 and variance= 0.419; PC2 = 17.61% of explained variations) and primary production estimates were best correlated with the third axis (loadings for Chl. a .= 0.545 and Periphyton = -0.451; PC3 = 14.25% of explained variations). The PERMANOVA showed significant differences between group centroids for all treatments and their paired interactions, but the resource type displayed the greatest coefficients of determination (R 2 = 0.35, see Fig. S4 for details). As also seen in the PCA, following the disturbance event, the Insect treatment had globally greater effects (i.e., greater divergence from the controls) compared to Leaf and Mix treatments (Fig 3B polygons). While the Insect and Mix treatments partially overlapped, the Leaf treatment partially overlapped the controls and was clearly different from the two latter owing to higher stoichiometric molar ratios. The analysis of multivariate homogeneity of group dispersions also showed significant differences for most of the considered factors, with exception of “Time” and the “Resource type * Abruptness level” terms (Fig. S5). More precisely, the distance to centroid increased from control (1.70) and Leaf (1.33) to Mix (2.17) and finally to Insect (3.55) treatment. 3.2 |Resource pulse effects on water nutrients Throughout the experiment, Control mesocosms displayed an overall stable concentration of TOC (LMM slope: 0.006 ± 0.011) but showed significant declines in TN and TP (LMMs slopes: -0.020 ± 0.015 and -0.080 ± 0.032 respectively; Fig.4 and Fig. S6). Thus, C:N and C:P slopes both exhibited significant increases (Fig. 5, LMMs slopes: 0.498 ± 0.274 and 153.041 ± 32.770, respectively). The observed N:P slope for the Control also showed a significant increase (LMM slope: 4.823 ± 1.241), suggesting that phosphorus declined at a faster rate than nitrogen throughout the duration of the experiment. Considering Insect and Mix treatments, similar patterns of nutrient dynamics were observed. Over the entire experimental period, the two treatments including insect addition (“insect” and “mix”) increased all nutrients and carbon in the mesocosms (Fig.4 – positive LMM slopes for TN, TP and TOC) as well as a global decrease of the stoichiometric molar ratios (Fig.5 – negative LMM slopes for C:N, C:P and N:P). Insect additions increased phosphorus at faster rate than nitrogen, which, in turn, increased faster than carbon (LMM slopes = 0.123 ± 0.011(TP), 0.1 ± 0.005 (TN) and 0.06 ± 0.004 (TOC)). The Leaf treatments led to a different nutrient dynamic (Fig.4, Fig.5). A general decrease of both TN and TP concentrations was observed following leaf addition, as in the Control (Fig. 4B1 and B2). However, leaf additions appeared to dampen the reduction of TP over time (Leaf TP LMM slope = -0.035 ± 0.022, Fig. 4B2) relative to Control. On the opposite, leaf additions led to TOC increases over time with a slope (0.029 ± 0.008) that was significantly higher than the Control and similar to the Mix treatments (Slope = 0.041 ± 0.008, see Fig. 4B3). As a result, the C:N molar ratio slope in the Leaf treatment increased at a faster rate than the Control (1.321 ± 0.187, see Fig. 5B1), while the C:P trend in the Leaf treatment had a significantly lower slope than the Control (90.362 ± 22.421) (Fig. 5B2, see Table S4 for LMMs results). Finally, Leaf treatment N:P was stable over the experimental period – with a slope significantly lower than the Control (0.101 ± 0.849), slightly higher than the Insect and most similar to the Mix treatments (Mix N:P LMM slope = -1.135 ± 0.849) (Fig. 5B3). At the end of the experiment (i.e., week 17) both Resource type and Abruptness level factors resulted in divergences in nutrient concentrations and their ratios (except for C:N where only the resource type factor was significant) without any significant interaction (Fig. 4C, Fig. 5C) – therefore exhibiting independent and predictable outcomes from each experimental factor (Fig. S7 for details). Within each ANOVA model where both factors were significant, F-values indicated a stronger effect of the Resource type factor (see associated Wilcoxon tests for paired Type comparisons) in contrast to the Abruptness levels with the exception of the N:P ratio (F Resource Type /F Abruptness level ≈ 19.34 (TN), 4.90 (TP), 4.06 (TOC), 21.76 (C:P) and 0.78 (N:P)). Finally for these five models, while usually of a lesser statistical importance (based on F values) compared to the Resource type, trends suggested that an increase in pulse abruptness level favored greater divergence from the Control (Fig. 4C, Fig. 5C). 3.3 | Resource pulse effects on DO and primary production Over the 17 weeks of the experiment, temporal dynamics and LMMs slopes displayed similar treatment (and Control) trends in weekly DO means, with only the Insect treatments shifting significantly apart from the Leaf treatments, increasing mean concentrations (see Fig. S8 for details). On the other hand, DO weekly variance (based on standard deviation) clearly showed distinct temporal dynamics for all three resource types, increasing progressively from Leaf, to Mix and Insect treatments. In all cases, DO variance peaks occurred at mid-point of the resource pulse event (week 5) followed by a progressive decrease to the end of the experiment (Fig. S8). At the end of the experiment, a significant effect of Resource type on DO (weekly mean and variance) was observed (ANOVA) but no effect of Abruptness level. Mix and Insect treatments had significantly higher weekly mean and variance in DO over the Control and the Leaf treatment, while the latter systematically overlapped with the Control. Chl. a . (in vivo chlorophyll a concentration) increased over time in the Insect treatments (LMM slope = 0.018 ± 0.011), with concentrations peaking at week 9 before decreasing until the final date (Fig. S9). Chl. a . LMM slopes declined in the Leaf treatment (slope = -0.017 ± 0.011) and the Control (slope = -0.035 ± 0.016) but remained constant in the Mix treatment (slope = -0.011 ± 0.011). At the end of the experiment, the only significant difference was higher Chl. a. values for the Insect relative to the Leaf treatment, with the Mix treatment being intermediate. The Abruptness factor only displayed a significant Chl. a. effect (p = 0.009) with a higher pulsed abruptness leading to a greater departure (i.e., increased concentration) from the undisturbed Control. Periphyton biomass increased in both Insect (slope = 0.231 ± 0.053) and Mix (slope = 0.124 ± 0.053) treatments while the Leaf (slope = 0.003 ± 0.053) treatment and the Control (slope = 0.002 ± 0.077) were stable (Fig. S9). At the end of the experiment, Mix and Insect treatments had higher periphyton biomass (2.100g ± 0.645 vs 4.683g ± 1.260, respectively, p<0.05) than the Leaf treatment and the Control (0.382g ± 0.046 vs 0.388 ± 0.158, respectively). The Abruptness factor did not lead to any significant differences in periphyton biomass. 3.4 | Resource pulse effects on leaf litter decomposition Resource type affected leaf decomposition (p = 0.0167; Fig. S8. for details) with the Leaf treatment having significantly greater decomposition percentage relative to the Insect treatments (Wilcoxon paired tests; mean and SE: 36.88 ± 0.22 % and 34.55 ± 0.56 %, respectively, Fig. 6). The Control (36.04 ± 0.34 %) and Mix (35.61 ± 0.65 %) treatment were intermediate and not significantly different from each other. jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf 4 | DISCUSSION Because of their ubiquity, diversity and potential effect on every scale and level of biological organization, understanding biomass-altering disturbances dynamics such as resource pulses is essential in an era of rapid global changes 4,17,18 . In freshwater ecosystems, resource pulse disturbances may occur as episodic allochthonous inputs of autotrophically and heterotrophically derived biomass (e.g., leaf litter or insect remains). This litter may have contrasting quality (i.e., nutrient content and decomposition rate) 19 for the recipient aquatic ecosystem. While it is well-known that terrestrial leaf litter inputs may broadly alter the structure and functioning of diverse freshwater ecosystem compartments (ex., subsidizing macro-invertebrates and microbial activity, for the benthic part), knowledge about the direct consequences of animal-type matrices decomposition on ecosystem processes remains scarce 19–21 . Constrained by our experimental design, we observed that, for an equivalent dried matter biomass, a higher quality resource type (the result of the interaction between the tissue matrix and the nutrient content) led to a higher range of expected disturbance pulsedness range. In turn, our experiment showed that resource pulses of higher-quality also resulted in a greater departure of the impacted ecosystems from the unmanipulated state (i.e., Control) together with an increase of the system stochasticity (i.e., as measured by within treatment dispersion, sensus Chase, 2010, and defined by Dietze, 2017 and Munch et al., 2022) 2223,24 . This larger response to high quality resources was amplified by greater resource pulse abruptness (shorter duration of an equivalent pulse amplitude). Thus, our results suggest that pulsed resources of high quality will tend to push the receiving ecosystem further away from its background (Control) state than either would lower quality resources, or the same resource pulsed over a longer duration (lower abruptness). Indeed, independently of the resource dynamic per se, the high nutrient levels arising from the decomposition of the high-quality resource (e.g., Insect treatment) was likely mainly bounded to the fluxes from the benthic to the pelagic compartment. During the time of the experiment, these nutrients had probably been integrated through distinct pathways by some nutrient-acquisitive taxa present in the mesocosms, forming green-based food-webs: like through the proliferation of filamentous green algae or cyanobacteria, as observed during the study and confirmed by periphyton and Chl. a . measures. As such, throughout the experiment, because of their improved rate of resource uptake at high resource availability compared to more conservative species, it is conceivable that fast-colonizing organisms led to the observed higher and more stochastic divergence state of the disturbed ecosystems processes compared to the Control. Those nutrient dynamic alterations sometimes induced changes in the trophic status of the mesocosms (from mesotrophic to mid-eutrophic systems, based on TLI parameter ranges of TN and TP 25 ) leaded by the increased benthic and/or pelagic primary production. Moreover, the Insect and Mix treatments led to a gradual decline of C:N and C:P compared to both the Control and the Leaf treatment as well as a progressive decrease in TP limitation as inferred by N:P ratios . Alltogether, these observations support previous results about the probability of N and P co-limitation occurring during the eutrophication process in lakes (i.e., through the lens of algal productivity) 26 and also support our hypothesis of green-based food webs developing in mesocosms impacted by resources pulses of high quality. Overall, the independent effect of abruptness was subtle, but nonetheless important. Our results suggest that low abruptness (longer duration of the resource pulse) could dampened the overall nutrient concentration released to the water column for any given resource pulse. Past theoretical work has shown that the greater the abruptness (high magnitude over a short period of time) of a high-quality resource influx, the greater the expected growth rates of generalist and opportunist populations through resource uptake at the benthic and pelagic levels, owing to elevated nutrient flux 4 . As such, in accordance with the work of Robert D. Holt (2008), a lowered and prolonged input of the insect resource probably reduced the excess of nutrients available at a given time through decomposition, limiting the amount of nutrients circulating from the benthic to pelagic compartments and allowing the primary consumers to better regulate the primary producers abundance 14 . In contrast, an already poor quality and recalcitrant resource such as leaves remains less sensitive to abruptness variation (at least during the time of our experiment), the nutrients (N and P) substantially remaining in the benthic compartment and supporting the development of a more heterotrophic food web structure initiated by benthic decomposers and detritivores. Leaf inputs, a more recalcitrant resource influx (higher C:N molar ratio compared to Insect inputs; likely owing to leaf lignin content) and nutrient-poor (Leaf inputs containing approximately 13.5 times less TN.g -1 dw than Insect inputs), did not prevent the global decrease of TN and TP in the water column also observed under unmanipulated, controlled conditions. As demonstrated by previous studies, these specific trends can be seen through the lens of the assimilation of labile solutes from the water column by microbes associated with detrital organic matter as well as more mechanically by the direct planktonic and other organic matter sedimentation 27–29 . As such, we hypothesize that leaf additions supported the development of the benthic pathway with nutrient fluxes confined to specific leaf litter decomposers. Supporting this hypothesis, a study assessing regional variation in plant detritus decomposition rates in lake littoral zones found that there was less water column nutrient release from leaf breakdown in lower nutrient lakes because the P and N was retained and sequestered by benthic microbial organisms 30 . Within the Leaf treatments, the macroinvertebrates decomposers inoculated before the experiment presumably contributed to leaf breakdown, along with fungi and bacteria also linked to detritus decomposition who probably led to the growth of a biofilm around the leaves. Assimilating available nutrients, these communities assemblages where possibly competing for resources bound within the leaves, subsequently restraining a direct resource flow to the water column during the time of the experiment 31–33 . As such, compared to the Insect treatments, Leaf treatments led to higher decomposition rates of the leaf litter bags, observation likely linked to the support of a more benthic and heterotrophic-based pathway. On the other hand, in mesocosms with resource influxes of insect biomass, decomposition rates of leaf litter (leaf bags) were reduced. We hypothesize that this is a result of the development of both benthic and pelagic autotrophic pathways, as shown by increases in periphyton and chl. a. , as well as greater weekly DO variance that point to high levels of respiration at night and oxygen release during the day. Although we monitored periphyton growth during the experiment, our sampling protocol did not allow for a direct verification of this aforementionned hypothesis. The tiles were positioned at the edges of the tanks and thus had no direct contact with the leaf resource added at the bottom of the mesocosms. They can only provide a proxy of the benthic biomass development capacity between each sampling period. Periphyton (including diverse attached diatoms, green algae and/or cyanobacteria 34,35 ) are capable of uptake of nutrients from the water column that would have arisen from decomposition of high quality influxing resources (i.e., Mix and Insect treatments). In contrast, Leaf treatments led to a global decrease of chl. a. , while DO followed the same trends as in the Control – indicating a reduced autotrophic oxygen production development (with reduced weekly variance). As such, even though we did not specifically scrutinize the periphyton communities’ composition nor measured it directly on the leaves surface, we can speculate that in this case the initial benthic compartment structure supported the subsequent development of a nutrient-demanding heterotrophic brown food web (e.g., fungi and bacteria), resulting in its localised growth, while the decomposition of insects in other mesocosms gave rise to a more autotrophic dominated periphyton. This specific microbial benthic food web that developed in the Leaf treatment was probably also adapted to the decomposition of the subsequent leaves addition in the water, hence explaining the higher break-down rate as opposed to the Insect treatment. In addition, the significant increase of TOC in the water, probably due to the release of C-recalcitrant compounds such as tannins contained within the leaves (also seen by the water brownification of the impacted mesocosms), did not change the overall trend of the Control treatment (i.e., a progressive increase of the limitation of TN and TP relative to TOC) indicating a minor impact of this type of disturbance on the water column temporal stoichiometric dynamic. To our knowledge, this study is the first to test and untangle the crossed effects of resource pulse disturbances quality and abruptness level within complex ecological systems such as aquatic freshwater mesocosms – suggesting that considering both resource quality and dynamic can be consequential for understanding ecosystem functions in a global change context. Undeniably, our results suggest that for a given amount of allochthonous resource, differences in resource pulse quality play an overriding role in determining the fate of the receiving ecosystem states and processes. As such, nutrient fluxes arising from high and labile resources would spread quickly across the ecosystem compartments and likely support the development of regionally more divergent and stochastic autotrophic-based food webs. An alteration in the phenology of a high quality resource pulse could potentially dampen those effects through a change in abruptness. On the contrary, more recalcitrant and nutrient-poor resource would confine the nutrient flow within the receiving (i.e., benthic) compartment, favoring heterotrophic-specialized organisms to develop, and making the system less responsive to a change in phenology (abruptness). REFERENCES 1. Loreau, M., Mouquet, N. & Holt, R. D. Meta-ecosystems: a theoretical framework for a spatial ecosystem ecology. Ecol. Lett. 6 , 673–679 (2003).2. Massol, F. et al. Linking community and ecosystem dynamics through spatial ecology. Ecol. Lett. 14 , 313–323 (2011).3. Richardson, J. S. & Sato, T. Resource subsidy flows across freshwater–terrestrial boundaries and influence on processes linking adjacent ecosystems. Ecohydrology 8 , 406–415 (2015).4. Jentsch, A. & White, P. A theory of pulse dynamics and disturbance in ecology. Ecology 100 , e02734 (2019).5. Behera, S. K. Understanding the impact of climate change on extreme events. Front. Sci. 2 , (2024).6. Zhou, S., Yu, B. & Zhang, Y. Global concurrent climate extremes exacerbated by anthropogenic climate change. Sci. Adv. (2023) doi:10.1126/sciadv.abo1638.7. Talluto, L. et al. Towards (better) fluvial meta-ecosystem ecology: a research perspective. Npj Biodivers. 3 , 1–10 (2024).8. Allen, D. C. et al. Global patterns of allochthony in stream-riparian meta-ecosystems. Ecol. Lett. 27 , e14401 (2024).9. Sullivan, S. M. P. & Manning, D. W. P. Aquatic–terrestrial linkages as complex systems: Insights and advances from network models. Freshw. Sci. 38 , 936–945 (2019).10. Adams, H. & Leroux, S. J. Integrating Field Data and a Meta-ecosystem Model to Study the Effects of Multiple Terrestrial Disturbances on Small Stream Ecosystem Function. Ecosystems 27 , 951–968 (2024).11. Klemmer, A. J., Galatowitsch, M. L. & McIntosh, A. R. Cross-ecosystem bottlenecks alter reciprocal subsidies within meta-ecosystems. Proc. R. Soc. B Biol. Sci. 287 , 20200550 (2020).12. Weber, M. J. & Brown, M. L. Effects of resource pulse magnitude on nutrient availability, productivity, stability, and food web dynamics in experimental aquatic ecosystems. Hydrobiologia 814 , 191–203 (2018).13. Sato, T., Ueda, R. & Takimoto, G. The effects of resource subsidy duration in a detritus-based stream ecosystem: A mesocosm experiment. J. Anim. Ecol. n/a , (2021).14. Holt, R. D. Theoretical Perspectives on Resource Pulses. Ecology 89 , 671–681 (2008).15. Knief, U. & Forstmeier, W. Violating the normality assumption may be the lesser of two evils. Behav. Res. Methods 53 , 2576–2590 (2021).16. Nakagawa, S., Johnson, P. C. D. & Schielzeth, H. The coefficient of determination R2 and intra-class correlation coefficient from generalized linear mixed-effects models revisited and expanded. J. R. Soc. Interface 14 , 20170213 (2017).17. Gough, C. M., Buma, B., Jentsch, A., Mathes, K. C. & Fahey, R. T. Disturbance theory for ecosystem ecologists: A primer. Ecol. Evol. 14 , e11403 (2024).18. Yang, L. H. & Naeem, S. The Ecology of Resource Pulses. Ecology 89 , 619–620 (2008).19. Benbow, M. E., Receveur, J. P. & Lamberti, G. A. Death and Decomposition in Aquatic Ecosystems. Front. Ecol. Evol. 8 , (2020).20. Becu, M. H. J. & Richardson, J. S. Leaf litter decomposition rates in freshwaters differ by ecosystem. Aquat. Sci. 85 , 101 (2023).21. Stoler, A. B. & Relyea, R. A. Reviewing the role of plant litter inputs to forested wetland ecosystems: leafing through the literature. Ecol. Monogr. 90 , e01400 (2020).22. Chase, J. M. Stochastic Community Assembly Causes Higher Biodiversity in More Productive Environments. Science 328 , 1388–1391 (2010).23. Dietze, M. C. Prediction in ecology: a first-principles framework. Ecol. Appl. 27 , 2048–2060 (2017).24. Munch, S. B., Rogers, T. L., Johnson, B. J., Bhat, U. & Tsai, C.-H. Rethinking the Prevalence and Relevance of Chaos in Ecology. Annu. Rev. Ecol. Evol. Syst. 53 , 227–249 (2022).25. Zhang, Y. et al. A Critical Review of Methods for Analyzing Freshwater Eutrophication. Water 13 , 225 (2021).26. Zhou, J., Han, X., Brookes, J. D. & Qin, B. High probability of nitrogen and phosphorus co-limitation occurring in eutrophic lakes. Environ. Pollut. 292 , 118276 (2022).27. Tank, J. L. et al. Partitioning assimilatory nitrogen uptake in streams: an analysis of stable isotope tracer additions across continents. Ecol. Monogr. 88 , 120–138 (2018).28. Pastor, A. et al. Stream carbon and nitrogen supplements during leaf litter decomposition: contrasting patterns for two foundation species. Oecologia 176 , 1111–1121 (2014).29. Bastias, E., Ribot, M., Bernal, S., Sabater, F. & Martí, E. Microbial uptake of nitrogen and carbon from the water column by litter-associated microbes differs among litter species. Limnol. Oceanogr. 65 , 1891–1902 (2020).30. DeGasparro, S. L., Beresford, D. V., Prater, C. & Frost, P. C. Leaf litter decomposition in boreal lakes: variable mass loss and nutrient release ratios across a geographic gradient. Hydrobiologia 847 , 819–830 (2020).31. Wissinger, S. A., Klemmer, A. J., Braccia, A., Bush, B. M. & Batzer, D. P. Relationships between macroinvertebrates and detritus in freshwater wetlands. Freshw. Sci. 40 , 681–698 (2021).32. Hayer, M. et al. Microbes on decomposing litter in streams: entering on the leaf or colonizing in the water? ISME J. 16 , 717–725 (2022).33. Marks, J. C. Revisiting the Fates of Dead Leaves That Fall into Streams. Annu. Rev. Ecol. Evol. Syst. 50 , 547–568 (2019).34. Zanden, M. J. V. & Vadeboncoeur, Y. Putting the lake back together 20 years later: what in the benthos have we learned about habitat linkages in lakes? Inland Waters 10 , 305–321 (2020).35. Allan, J. D., Castillo, M. M. & Capps, K. A. Primary Producers. in Stream Ecology : Structure and Function of Running Waters (eds. Allan, J. D., Castillo, M. M. & Capps, K. A.) 141–176 (Springer International Publishing, Cham, 2021). doi:10.1007/978-3-030-61286-3_6. Fig.01. Details of the disturbance dynamics design. From (a) to (c), the different abruptness levels and their temporal scaling during the 17 weeks of experiment, respectively on y and x axis. The weeks indicated in dark-gray represent the 5 sampling periods. The dotted lines indicate every under-event of each disturbance (occurring on every Wednesday) and the gray area represent the total pulse occurrence. (d), relative positions of the different abruptness levels regarding the experimental design (figure inspired from the theoretical work of Jentsch and White 4 ). Abruptness levels are constructed as follow: Magnitude/duration of the disturbance – with the duration being equal to the pulse total weeks (being also equal to the number of under-events). Fig.02. Schematic perspectives of a mesocosm. (a) lateral view and details of the timeline relative to the benthic and pelagic compartments inoculation. (b) Upper view with details of the structures placed in each tank. The dotted line represents the rope holding the two immerged dataloggers. Fig.03. For both figures colors represents the resource types (red, purple, orange and black for Insect, Mix, Leaf and Control, respectively). A: Disturbance pulsedness range based on Total Nitrogen (TN dried weight) content of the resource types. Dotted lines represent a theoretical gradient of pulsedness constructed with minimal and maximal values obtained from the resource types content. Point shapes represent the abruptness levels based on total biomass (round, square and triangle for High, Medium and Low, respectively). B: principal component analyses (PCA) based on data collected from sampling 02 (i.e., week 5, middle pulse) to sampling 05 (i.e., week 17, end of the experiment) – each illustrated by the “+” symbol. Axis exhibits the two main components (PC1 and PC2) and polygons highlight the spaces occupied by the resource types, including a polygon for the Control treatment. jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf Fig.04. Decomposition of the three nutrients measured in this study: TN, TP and TOC (row 01, 02 and 03, respectively). Colors represents the resource types (red, purple, orange and black for Insect, Mix, Leaf and Control, respectively). Point shapes and line types represents abruptness levels: round-solid, square-dashed, triangle-dotted and diamond-longdashed for High, Medium, Low and Control, respectively. Radiant points on panel B stand for resource types without consideration of the abruptness levels. Panel A (left column), nutrient temporal dynamics for each treatment (mean and associated Std. Errors, SE). The grey frames give a visual on the duration for each abruptness levels. The solid line at the end of week 12 indicates the inoculation of the leaf-litter bags in the mesocosms. Panel B (center column), fixed effects estimates and associated confidence intervals (CI) of each resource types extracted from Linear Mixed Models (LMMs). Panel C, details of the concentrations measured at the end of the experiment (week 17). Boxplots are built regardless of the different abruptness levels (N = 12). For a given boxplot, points and associated error bars represent values for each abruptness levels. Grey dotted zone represents the range of the Control treatment (based on mean and SE). Fig.05. Decomposition of the stoichiometric molar ratios arising from the nutrients measured in this study – C:N, C:P and N:P (row 01, 02 and 03, respectively) – following the same procedure presented in figure Fig. n. For a better visual appreciation, values on the y axis of the panel C figures are log-transformed. Fig.06. Leaf break-down percentage in relation to resource types (with associated Wilcoxon pairwise comparisons tests). Point shapes and associated error bars represents the three abruptness levels: round, square and triangle for High, Medium and Low, respectively. Grey dotted zone represents the range of the Control treatment (based on mean and SE). Information & Authors Information Version history V1 Version 1 17 July 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords allochthonous perturbation biomass-altering disturbances cascading impacts ecosystem processes nutrient flow pulse dynamic Authors Affiliations Charlie Sarran 0009-0000-6573-7108 [email protected] Université de Montréal View all articles by this author Beatrix Beisner University of Quebec at Montreal View all articles by this author Eric Harvey 0000-0002-8601-7326 Université du Québec à Trois-Rivières View all articles by this author Metrics & Citations Metrics Article Usage 323 views 149 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Charlie Sarran, Beatrix Beisner, Eric Harvey. Interactive influences of resource pulse quality and abruptness on aquatic ecosystem. Authorea . 17 July 2025. DOI: https://doi.org/10.22541/au.175273206.63627240/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175273206.63627240/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a004bb34492758f4',t:'MTc3OTU0NjM0MA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();