Combined effects of salinization and temperature on microbial-mediated oak decomposition and invertebrate consumption

Water salinization is a recognized growing threat to freshwaters, whose consequences on streams’ function, per se or concomitantly with other stressors, are still far from clear. This microcosm study evaluated the combined effect of salinization (0 and 4 g/l NaCl) and temperature (5, 15, and 20 °C) on microbial-mediated oak leaf litter decomposition, with fungal biomass, sporulation, and microbial respiration as microbial descriptors. Invertebrate consumption was also assessed using the common shredder Sericostoma vittatum (Trichoptera, Sericostomatidae). Mass loss was affected by temperature and interaction between salinity and temperature. Under salt conditions, mass loss was higher at 15 °C and reduced (~ 10%) at 20 °C. Microbial activity was lower at 5 °C and higher at 15 and 20 °C, irrespective of salinity. Fungal biomass was affected by both temperature (5 < 20 < 15 °C) and salinity (4 < 0 g/l NaCl), although the interaction between both was not significant. The interaction of both variables affected the production of spores: salt addition strongly reduced sporulation rates at all temperatures despite a significant increase in conidial production with temperature. Invertebrate leaf consumption was significantly reduced only by salinization. Overall, our results seem to indicate that temperature may modulate the effect of salinization (at least at ≥ 4 g/l NaCl) on stream leaf decomposition. While stronger salinization effects may be observed at higher temperatures, a consistent strong inhibition of shredders’ feeding behavior promoted by salt, regardless of temperature, may anticipate important repercussions on streams’ secondary production throughout the year.


Introduction
Forested headwater streams are intimately connected with their riparian areas.As a consequence of the shadow promoted by the trees' canopies, light access to the channel is limited, resulting in stream communities highly dependent on allochthonous sources of organic matter, mainly in the form of senescent leaves (Abelho 2001;Drake et al. 2017;Marks 2019).Leaf litter decomposition is a key ecosystemlevel process in these systems, promoting the incorporation of terrestrial detritus into the brown aquatic food chains (Evans-White and Halvorson 2017;Marks 2019).This process integrates physico-chemical leaf litter changes and is executed by microbial (fungi and bacteria) and invertebrate protagonists (Gessner et al. 1999;Marks 2019;Swan et al. 2021), being closely modulated by environmental factors.The interaction between these biotic and abiotic drivers ultimately determines nutrient recycling and the transfer of energy across trophic levels (Perkins et al. 2010;Tiegs et al. 2019).
Several anthropogenic-related stressors presently affect stream ecosystems at an unprecedented level (Reid et al. 2019;Tickner et al. 2020).Among these, temperature increase (Gonçalves et al. 2015;Tiegs et al. 2019;Pérez et al. 2021;Niedrist 2023) and salinization (Cañedo-Argüelles et al. 2016;Kaushal et al. 2021;Thorslund et al. 2021;Carrier-Belleau et al. 2022) are major and intensifying threats at a global scale (Berger et al. 2019;Reid et al. 2019), directly affecting biotic communities composition, their distribution, and freshwater ecosystems processes such as leaf litter decomposition (Follstad Shah 116 Page 2 of 10 et al. 2017; Verberk et al. 2020;Wilmot et al. 2021).Many studies show that elevated temperature accelerates decomposition leading to the stimulation of leaf leaching and metabolic rates of the decomposer communities (Canhoto et al. 2016;Amani et al. 2019;Liu et al. 2022;Monroy et al. 2022), provided that their optimum temperature is not exceeded (Brown et al. 2004;Bonacina et al. 2023).Enhanced fungal sporulation rates, enzymatic activity, and biomass accrual in response to temperature are known to direct and indirectly, i.e., through the stimulation of invertebrate feeding rates, increase leaf processing (e.g., Fernandes et al. 2012;Gonçalves et al. 2013Gonçalves et al. , 2015;;Canhoto et al. 2016;Fenoy et al. 2016).A thermotolerant shredder community shows a generally positive response to temperature increase, usually correlated with an increase in respiration, growth, and developmental rates (Bonacina et al. 2023); this response, however, seems to be resource dependent (Landeira-Dabarca et al. 2019).Nonetheless, while thermal stress has been dealt with as a main concern by the scientific community, particularly in the scope of global changes, salinization is still comparatively overlooked and research fragmentary (Cañedo-Argüelles 2020; Canhoto et al 2021).Works assessing the effects of stream secondary salinization, i.e., increase in the concentration of inorganic ions promoted by human activities (Cañedo-Argüelles et al. 2013), have tended to focus on the effects of salt on invertebrates.Results consistently point to a reduction in their abundance (Piscart et al. 2006), diversity (Kefford et al. 2011), functional richness (Zhao et al. 2021), and distribution (Verberk et al. 2020), mainly due to physiological stress promoted by osmotic imbalances.As the stress threshold of invertebrates (namely shredders; Cummins 1974) is usually lower than that of microbial communities (Canhoto et al. 2021), a microbial-dominated depressed organic matter degradation, largely promoted by a halotolerant and impoverished fungal community, seems to dominate in salinized streams [> 3 g/l; Almeida Júnior et al. 2020)].Fungal tolerance to salt (Canhoto et al. 2017) seems to be species-specific and result from energetic tradeoffs between osmotic adjustments, mycelial growth and, particularly, sporulation (Canhoto et al. 2017(Canhoto et al. , 2021;;Gonçalves et al. 2019b;Carrier-Belleau et al. 2022).
To date, only limited information exists on the impacts of salt contamination co-occurring with other abiotic factors on stream communities and ecosystem function (Schäfer et al. 2012;Gonçalves et al. 2019b;Orr and Buchwalter 2020;Carrier-Belleau et al. 2022).To our knowledge, no previous approach concomitantly assessed the effects of salinization and temperature increase on litter decomposition dynamics.Nonetheless, a generalized overriding impact of the former over other stressors seems to occur in several previous tested/observed combinations (Velasco et al. 2019;Cuthbert et al. 2019), with some exceptions being reported (e.g., Gonçalves et al. 2019b).
In this study, we aimed to evaluate the combined effect of salinization [control (C; 0 g/l NaCl) or enriched (E; 4 g/l NaCl) media] and temperature (5, 15, and 20 °C) in the decomposition of oak leaves and associated microbial parameters.The effects of both factors on consumption rates by the shredder Sericostoma vittatum (Trichoptera; Sericostomatidae) were also assessed.We expected (1) salinity to have deleterious effects on leaf litter mass loss and related descriptors, particularly at the tested extreme temperatures (i.e., 5 and 20 °C), as fungal species from central Portugal streams have an optimal activity at around 11 °C (Dang et al. 2009;Gonçalves et al. 2015); (2) higher microbialmediated leaf litter quality at 15 °C to be translated into higher consumption rates by the shredder in both media; (3) salt enrichment to decrease shredder consumption due to osmotic energetic requirements, despite marked increases on invertebrate's metabolism occurring at 20 °C.

Study site and experimental procedures
Oak leaves (Quercus robur L.) were collected after abscission and air-dried in the dark at room temperature.Leaves were placed in fine-mesh bags (10 × 15 cm, 0.5 mm mesh) and submerged in a reference oligotrophic stream (Ribeira do Candal, Lousã, central Portugal, 40° 4′ 44″ N, 8° 12′ 10″ W; 620 m asl) to allow microbial colonization.This secondorder stream runs through a mixed deciduous forest, mostly composed of Quercus robur L. and Castanea sativa Mill.trees.Mean stream flow was 0.50 ± 0.09 m/s, oxygen concentration 10.80 ± 0.40 mg/l (WTW Oxi 3310, Germany), pH 6.4 ± 0.1 (WTW pH 3110, Germany), and conductivity 198.00 ± 32.30 µS (Hanna HI 98192,Portugal).After 2 weeks, leaves were brought back to the laboratory.Using a cork borer, leaf discs (12 mm diameter) were punched out either randomly or as pairs cut symmetrically relative to the main leaf vein.In this case, discs from the same pair were assumed to have similar initial mass (Bärlocher et al. 2020) and were further used to assess consumption rates by the shredder S. vittatum (see below).
Twenty-four groups of 15 random leaf discs were distributed in Erlenmeyer flasks, containing either 40 ml nutrient medium [75.5 mg CaCl 2 , 10 mg MgSO 4 •7H 2 O, 0.5 g 3-morpholinopropanesulfonic acid, 5.5 mg K 2 HPO 4 , and 100 mg KNO 3 per liter of sterile distilled water (Dang et al. 2005)], corresponding to the control treatment (C medium, n = 12), or 40 ml nutrient medium enriched with 4 g/l NaCl (E medium, n = 12).This NaCl concentration, previously known to have an effect on fungal community structure and activity (Gonçalves et al. 2019b), corresponds to realistic conditions induced on streams by agriculture and road deicing practices (Gonçalves et al. 2019a;Oliveira et al. 2021).
Microcosms containing C and E media were randomly and equitably distributed to chambers at three different temperatures: 5, 15, and 20 °C.These temperature levels are common in low-order temperate streams across seasons and may provide insights into the effects of temperature increase in streams experiencing the effects of global warming.All microcosms, regardless of temperature and salinization treatment, were aerated on an orbital shaker (100 rpm) under a 12:12 h light/dark photoperiod; nutrient medium was renewed every 2 days under sterile conditions (Gonçalves et al. 2014), and Erlenmeyer flasks were always closed with sterile cotton plugs (Gonçalves et al. 2014).After 10 days of incubation in the laboratory, random discs were used to assess mass loss, microbial activity, fungal biomass, and sporulation rates.Six additional Erlenmeyer flasks, each with 10 pairs of leaf discs, were prepared, in the same conditions, for consumption tests.

Leaf mass loss
After the initial 14-day colonization period in the stream, 20 sets of 15 random discs obtained from stream-conditioned oak leaves were oven-dried (60 °C, 48 h) and weighed (± 0.1 mg); the average mass was considered as the initial dry mass (DM) of 15 discs after stream colonization.At the end of the experimental period, dry mass loss was estimated as the difference between the initial and final dry mass of the 15 discs from each microcosm, obtained after all microbial analyses (see below).

Microbial activity
Microbial activity was assessed by fluorescein acetate (FDA) hydrolysis (Datry et al. 2011;Simões et al. 2022).From each microcosm, a random subset of five leaf discs was placed in a sterile glass vial with 3 ml phosphate buffer (pH = 7.6) and 100 µl FDA stock solution (0.02 g FDA in 10 ml acetone).After incubation in the dark for 60-100 min, the reaction was stopped with 3 ml acetone.Absorbance (490 nm) was measured using a spectrophotometer (6400 Jenway, Dunmow, Essex, UK).Leaf discs were oven-dried (60 °C, 48 h) and weighed, and microbial activity was expressed as μmol FDA/h/g DM.

Aquatic hyphomycete sporulation
A final subset of five discs from each microcosm was incubated in 25 ml nutrient medium and maintained under the original salinization/temperature treatment.After 48 h, conidial suspensions were collected into 50-ml Falcon tubes and preserved with 2 ml 37% formalin.Subsamples were filtered (Millipore SMWP, 5 μm pore size, Billerica, MA, USA) and the retained spores stained (0.05% cotton blue in lactic acid 60%) and counted under a compound microscope at 250× magnification (Bärlocher et al. 2020).Discs were oven-dried (60 °C, 48 h) and weighed; aquatic hyphomycete sporulation rates were expressed as number of conidia/mg DM/day.

Invertebrate consumption test
Sericostoma vittatum larvae were collected from Ribeira do Candal (Lousã, central Portugal, 40° 4′ 44″ N, 8° 12′ 10″ W; 620 m asl) and brought to the laboratory.Individuals were acclimatized in containers with aerated filtered stream water at 5, 15, or 20 °C for 1 week.Larvae were fed ad libitum with colonized leaves collected from the same stream and exposed to a 12:12 h light/dark photoperiod.
A total of 60 invertebrates were individually allocated into plastic containers (i.e., feeding chambers; 70 mm diameter × 85 mm high), each containing 200 ml filtered (Whatman GF/F 50 mm) and aerated stream water.The bottom of each feeding chamber was covered by a fine layer of sieved (< 2 mm diameter) and ashed stream sediment.In half the containers for each temperature treatment, the stream water was enriched with 4 g/l NaCl.Just before the beginning of the tests, the diameter of the larval case opening (CO; mm) was measured, and individual DM (0.0115 ± 0.0004 g) was estimated using the regression model DM = CO × 0.0136 -0.0162 (Gonçalves et al. 2011).Individuals were starved for 14 h before the beginning of the tests.For each temperature and salt treatment, 10 replicates were assessed.Each microcosm was supplied with a pair of leaf discs (previously conditioned in C or E medium at the three temperature levels): one oak disc from each pair was offered to the invertebrate by directly placing it in the sediment, while the other symmetrical disc was enclosed in a small finemesh bag (0.5 mm mesh), allowing continued microbial 116 Page 4 of 10 decomposition while restricting consumption by the invertebrate (Bärlocher et al. 2020).
Leaf discs were removed from the containers when half of the exposed discs were 50% consumed (visual assessment); the discs from each replicate and treatment were recovered, oven-dried (60 °C, 48 h) and weighed.Invertebrate consumption was estimated as the difference between the final DM of the control and corresponding disc offered to the invertebrate.Individual relative consumption rates (RCR) for each treatment were expressed as g leaf DM consumed/g invertebrate DM/day (Bärlocher et al. 2020).

Statistical analyses
Differences in mass loss, microbial activity, fungal biomass, sporulation, and consumption rates were each analyzed using two-way analysis of variance (ANOVA), with temperature (5, 15, and 20 °C) and salinization (0 and 4 g/l) as categorical variables.Tukey's HSD tests were applied whenever significant (p < 0.05) statistical differences were found to determine differences between treatments.Normality and homoscedasticity assumptions were met in all cases (assessed through histograms and Bartlett's test).Statistical analyses were performed with Statistica 7 software (StatSoft, Tulsa, OK, USA).

Microbial activity
Microbial activity was not significantly affected by salinity or the interaction between both factors (2-way ANOVA, p > 0.12).Temperature affected microbial activity (2-way ANOVA, p < 0.001); it was significantly lower at 5 °C (Tukey's test, p < 0.001), with no differences between the values observed at 15 and 20 °C (Tukey's test, p = 0.06).
Minimum microbial activity values (1.65 ± 0.30 µmol FDA/h/g DM) were found associated with oak leaf discs exposed to E medium at 5 °C; maximum values (4.42 ± 0.16 µmol FDA/h/g DM) were found in the same salinization treatment at 15 °C (Fig. 2a).
Fungal biomass was highest (103.69 ± 6.01 mg/g DM) in oak leaves exposed to the C medium at 15 °C and lowest (31.62 ± 0.41 mg/g DM; Fig. 2b) in leaves exposed to salt (E medium) at 5 °C (Fig. 2b).

Consumption rates
Consumption rates were significantly affected by salinity (2-way ANOVA, p < 0.001) but not by temperature or the interaction between both factors (p = 0.176).Regardless of the temperature, consumption rates were significantly reduced (> 35%) in the E medium (Tukey's test, p < 0.001; Fig. 3).

Discussion
The results from this study confirm that the process of leaf decomposition is affected by temperature and salinization; more importantly, they show that the combined effect of both stressors may differentially affect leaf mass loss, particularly at higher temperatures.In addition, salinization was deleterious to detritivore consumption, while the thermal conditions did not affect consumption in either treatment.
Microbial-mediated mass loss was globally not significantly impacted by the addition of salt by itself.This result is in line with previous laboratory (Gonçalves et al. 2019b;Martínez et al. 2020) and field (Berger et al. 2019;Vander Vorste et al. 2019) works, although contradictory results have also been found (Cañedo-Argüelles et al. 2014;Gómez et al. 2016;Sauer et al. 2016;Canhoto et al. 2017Canhoto et al. , 2022)).This disparity may be explained by the short exposure to salinization in this work (i.e., 10 days) but also by potential changes in fungal community composition and/or interactions among fungal species (Gonçalves et al. 2019a), frequently occurring at the salt contamination level (4-6 g/l) used in these tests (Gonçalves et al. 2019a;Canhoto et al. 2021).As fungal biomass decreased globally with salt addition, it seems likely that, under salt stress, fewer tolerant fungal species, less expressive in terms of biomass, may dominate and guarantee oak leaf degradation (Canhoto et al. 2021) eventually through functional redundancy (Dang et al. 2005;Allison and Martiny 2009;Gonçalves et al. 2013Gonçalves et al. , 2015)).Also, the negative effects of salinization seem to be less evident in low (i.e., oak; present study) than high (i.e., alder) quality leaf litter (Oliveira et al. 2021), strengthening the idea that intrinsic leaf physico-chemical characteristics are central determinants of decomposition in salt-contaminated freshwaters (Stoler et al. 2017) and streams (Almeida Júnior et al. 2020;Monroy et al. 2022).
Previous works suggest 11 °C (e.g., autumn/spring) to be the optimum temperature for leaf decomposition promoted by most aquatic hyphomycete species isolated from Central Portuguese streams (Gonçalves et al. 2015b;Simões et al. 2022).According to our hypothesis, it seems that the negative impact of salinization on microbial-mediated litter decomposition tends to be buffered when water temperatures are close to this optimum, probably as a result of thermal metabolic stimulation, suggested by the increased microbial activity and fungal biomass.In fact, decomposition efficiency was higher at 15 °C regardless of salinization treatment.On the other hand, at both temperature extremes tested, the effect of salinization on oak mass loss depended on temperature: compared to the control media, salt addition resulted in 15-times higher mass loss at 5 °C but 3-times lower at 20 °C.The first result, although not significant, cannot be ignored, as it mimics what commonly happens during winter when salt is used as deicer agent leading to the salinization of mountain streams.The potential high ability of aquatic hyphomycetes to sequester Na (Kaspari et al. 2009), associated with physiological mechanisms to enhance tolerance to cold temperatures [e.g., accumulation/increase of the disaccharide trehalose, polyols, antifreeze proteins, and/or unsaturated lipids (Robinson 2001;Senik et al. 2019)], may have facilitated the maintenance of hyphal integrity, increasing the degradative capacity in salt-rich medium compared to the control at 5 °C, although at a cost of lower biomass accrual and vestigial sporulation rates.A more deleterious effect of salinization was observed under warmer water temperatures (20 °C).Despite a similar starting fungal community due to the leaves' stream incubation, this could translate a fungal community with a higher proportion of dormant species (Fenoy et al. 2022).However, as microbial activity was only slightly (and nonsignificantly) reduced from 15 to 20 °C, we suggest that the fungal community was restricted to active thermotolerant species that presented lower degradative capacity because of physiological strategies to cope with thermal stress (del Campo et al. 2021).Also, as happens with other stressors, higher temperatures may determine increased salt toxicity (Orr and Buchwalter 2020).This is particularly worrisome as salinization is expected to increase in most persistent and intermittent streams during warmer seasons as a result of reduced dilution ability of aquatic systems due to lower precipitation and runoff, increased evaporation and water abstraction (Jeppesen et al. 2020).However, a subtle increment in conidial production paralleled the temperature increase in salt-rich media.This may suggest a thermal metabolic stimulation that leads to the production and release of conidia as a potential strategy to overcome unfavorable ionic conditions (Bärlocher et al. 2008;Canhoto et al. 2016).
The overpowering effect of salinization in relation to temperature, particularly visible in shredder feeding behavior, confirmed previous studies suggesting Trichoptera to be more sensitive to salt than temperature (Fenoy et al. 2020;Verberk et al. 2020).Consumption rates were consistently reduced in salinized treatments regardless of temperature.This may have occurred directly through waterborne toxicity, indirectly due to the reduced quality of offered food (leaves conditioned by less palatable, salinity-tolerant fungal assemblages), or a combination of both.In fact, elevated ion concentrations may determine osmoregulatory stress on several insect groups, caddisflies included, even at levels below the ones used in our tests (Tyree et al. 2016;Dowse et al. 2017;Jackson and Funk 2019;Feld et al. 2023).Although our experimental design does not allow a causal discrimination (i.e., food vs. waterborne effect), we foresee that reduced food intake under salinized conditions will likely compromise larval performance and consequently litter decomposition, particularly if conditions persist.It was interesting to notice that, although non-significant, consumption rates in salt-enriched medium increased with temperature (Orr and Buchwalter 2020).This shredder species is known to tolerate our tested temperature levels (Dunlop et al. 2008); nonetheless, the increased metabolism at higher temperatures and the maintenance of an osmotic balance occur at an energetic cost that may require a compensatory food intake, particularly when facing leaves of poorer quality (Swan and Palmer 2006;Tyree et al. 2016;Landeira-Dabarca et al. 2019;Orr and Buchwalter 2020).
If confirmed, this supports previous findings that indicate lower effects of salinity at lower temperatures (Jackson and Funk 2019;Velasco et al. 2019) and litter quality as an important modulator of leaf processing in streams affected by salinity per se or co-occurring with other stressors (Landeira-Dabarca et al. 2019;Almeida Júnior et al. 2020;Oliveira et al. 2021).
Our results globally confirm a dominant negative effect of salinity over temperature on microbial and detritivore-mediated decomposition of oak leaves, while also strengthening the crucial role of fungi, over invertebrates, on leaf litter processing in salinized streams.Based on our results, we expect a higher resilience and activity of the microbial communities to salt stress at intermediate temperatures (15 °C), frequent in autumn when the organic matter is more abundant in temperate freshwaters.Increased salt concentration occurring at higher temperatures seems to be more deleterious to microbial leaf degradation than at lower ones, while potentially resulting in a frail stimulation of S. vittatum larvae (reduced) consumption rates, as a compensatory feeding behavior on lower-quality oak detritus.Salt effects on leaf consumers were obvious but need to be amplified and further scrutinized; invertebrates' response to salinization under different thermal scenarios is likely species specific and occurs via the still not disentangled direct waterborne toxicity and/or food contamination pathways, with consequences propagating through the aquatic food webs.
Salinization and increased water temperature are abiotic stressors closely linked with global change, able to alter, per se and in combination, key processes and services of stream ecosystems.These consequences are far from elucidated, despite the urgency to adopt legislative measures and specific environmental guidelines to protect and manage these vulnerable ecosystems.

Fig. 2
Fig.2Microbial activity (a), fungal biomass (b), and sporulation rate (c) associated with oak leaf discs exposed to two salinization (C, 0 g/l NaCl; E, 4 g/l NaCl) and three temperature (5, 15, and 20 °C) treatments.Values are means ± SE.Different letters indicate significant (p < 0.05) differences between temperature treatments regardless of salinization for microbial activity; between temperature (lower-case letters) and between salinization treatments (uppercase letters) for fungal biomass; and between all treatments, for sporulation rate

Fig. 3
Fig. 3 Relative consumption rate by Sericostoma vittatum of oak leaf discs exposed to two salinization (C, 0 g/l NaCl; E, 4 g/l NaCl) and three temperature (5, 15, and 20 °C) treatments.Values are mean ± SE.Different uppercase letters indicate significant (p < 0.05) differences between salinity treatments