Non‐Abrupt Vegetation Changes due to Altered Nutrient Balance Make Complex Scale‐Dependent Warming and Cooling Effects

Land-atmosphere exchanges are mediated by biophysical properties (e.g., albedo change, evaporative cooling) and biogeochemical cycle (e.g., CO₂ fluxes), with both processes exerting global feedback as radiative forcing ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>RF</mml:mi></mml:mrow> <mml:annotation>$$ RF $$</mml:annotation></mml:semantics> </mml:math> ). While most research on <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>RF</mml:mi></mml:mrow> <mml:annotation>$$ RF $$</mml:annotation></mml:semantics> </mml:math> concentrated on the impact of abrupt vegetation changes, this study investigates the effects on non-abrupt changes due to altered nutrient levels (i.e., nitrogen [ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>N</mml:mi></mml:mrow> <mml:annotation>$$ N $$</mml:annotation></mml:semantics> </mml:math> ] and phosphorus [ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>P</mml:mi></mml:mrow> <mml:annotation>$$ P $$</mml:annotation></mml:semantics> </mml:math> ] deposition). We examined impacts of these changes by assessing <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>RF</mml:mi></mml:mrow> <mml:annotation>$$ RF $$</mml:annotation></mml:semantics> </mml:math> , representing global effects, and linked it with surface temperature ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>Ts</mml:mi></mml:mrow> <mml:annotation>$$ Ts $$</mml:annotation></mml:semantics> </mml:math> ), reflecting local influence. We hypothesized there are scale-dependent warming and cooling effects due to surface-atmosphere interactions. We explored this question using a 9-year dataset (2014-2023) from a large-scale nutrient manipulation experiment in a semi-arid savanna, Spain. Three co-located eddy-covariance sites are established: control, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>N</mml:mi></mml:mrow> <mml:annotation>$$ N $$</mml:annotation></mml:semantics> </mml:math> -added ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>NT</mml:mi></mml:mrow> <mml:annotation>$$ NT $$</mml:annotation></mml:semantics> </mml:math> ), and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>N</mml:mi> <mml:mo>+</mml:mo> <mml:mi>P</mml:mi></mml:mrow> <mml:annotation>$$ N+P $$</mml:annotation></mml:semantics> </mml:math> -added ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>NPT</mml:mi></mml:mrow> <mml:annotation>$$ NPT $$</mml:annotation></mml:semantics> </mml:math> ). The results indicate domination of changes in surface albedo over CO₂ fluxes, producing paradoxical effects: a net cooling at global scale ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>RF</mml:mi></mml:mrow> <mml:annotation>$$ RF $$</mml:annotation></mml:semantics> </mml:math> differences are [mean ± SD]-0.46 ± 0.08 W m^-2 [global] m^-2 [surface] at <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>NT</mml:mi></mml:mrow> <mml:annotation>$$ NT $$</mml:annotation></mml:semantics> </mml:math> and -0.39 ± 0.09 W m^-2 m^-2 at <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>NPT</mml:mi></mml:mrow> <mml:annotation>$$ NPT $$</mml:annotation></mml:semantics> </mml:math> ) due to higher surface reflectivity, but localized warming at understory ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>Ts</mml:mi></mml:mrow> <mml:annotation>$$ Ts $$</mml:annotation></mml:semantics> </mml:math> differences are 0.63°C ± 0.46°C at <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>NT</mml:mi></mml:mrow> <mml:annotation>$$ NT $$</mml:annotation></mml:semantics> </mml:math> and 0.80°C ± 0.77°C at <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>NPT</mml:mi></mml:mrow> <mml:annotation>$$ NPT $$</mml:annotation></mml:semantics> </mml:math> ) driven by shifts in energy partitioning. Furthermore, our findings indicate that <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>N</mml:mi></mml:mrow> <mml:annotation>$$ N $$</mml:annotation></mml:semantics> </mml:math> -only addition has more canopy-level <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>Ts</mml:mi></mml:mrow> <mml:annotation>$$ Ts $$</mml:annotation></mml:semantics> </mml:math> cooling than <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>N</mml:mi> <mml:mo>+</mml:mo> <mml:mi>P</mml:mi></mml:mrow> <mml:annotation>$$ N+P $$</mml:annotation></mml:semantics> </mml:math> treatment, although <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>Ts</mml:mi></mml:mrow> <mml:annotation>$$ Ts $$</mml:annotation></mml:semantics> </mml:math> increases at the understory. These contrasting responses reveal a layered and scale-dependent interplay of surface-atmosphere interactions. They highlight the critical role of nutrient stoichiometry in shaping climate feedbacks despite the vegetation changes are not abrupt, and emphasize that what cools the globe may still warm the land beneath our feet.