Ecology

Field experiments: collective structuresWe found that self-assembled Eciton hamatum bridges adaptively adjust in response to shifts in the terrain on which they are built. Detailed methods are included in Methods: Field experiments. Briefly, we moved foraging trails onto an apparatus where we could introduce a terrain gap. We repeatedly changed the size of this gap by first incrementally increasing it to 30 mm, by 1 mm every 30 s, and then incrementally contracting it at the same rate (See Fig. 1, Methods: Field experiments, and Supplementary Movie 1). As the size of the gap was expanded (the period before the dotted line in Fig. 2a, b) both the volume and number of ants increased to mean maximum values of 1080 mm3 (standard error, s.e. 84) and 18.9 ants (s.e. 1.6), respectively. Ants typically began forming a bridge when the gap was ~5 mm. As the gap size was decreased (period after dotted line), volume and the number of ants decreased back to zero as ants left the bridge. These broad dynamics across the ten complete trials were similar (Fig. 2a, b, inset panels and Supplementary Figs. 2, 3). Additionally, bridge volume (Fig. 2a) strongly correlated with the number of ants in the bridge (Fig. 2b), indicating that the density of ants per unit volume in these structures is relatively consistent (Pearson correlation coefficients range from 0.88 to 0.98 across the ten trials, see also Supplementary Fig. 4). Bridges broke and quickly reformed in eight of the ten trials; breaks occurred in both experimental phases, and these broken periods were excluded from analyses. Overall, these results show that bridges adjust dynamically to changing terrain geometry, as stretching the bridges caused them to become larger, with more ants, and contracting bridges caused them to become smaller, with fewer ants.Fig. 1: Experimental procedure and data extraction summary.Experiments were conducted on robust E. hamatum foraging trails, which were moved onto the experimental apparatus while it was closed. a Experimental procedure: The size of the gap was increased by 1 mm every 30 s until the gap reached 30 mm (expansion phase), then decreased at the same rate till no gap remained (the contraction phase). b Field setup: Experiments were recorded from both the side and the top, examples of bridges during each phase of the same trial are shown. c Data extraction: Example images and silhouettes from the maximum size bridge (30 mm) of the same trial as the images of 20 mm bridges shown in panel a. The envelopes of the bridges were extracted at a temporal resolution of 1 s; for each focal second, image frames were averaged over 10 s to remove ants walking on the bridge from the extracted envelopes. Envelopes were automatically extracted using hue-saturation-value (HSV) thresholding, with thresholds checked independently for each trial due to lighting differences. Locations of fixed points on the platform were used to re-scale and combine data from the side and top views into a single coordinate system in which 100 pixels = 1 cm. Estimates of bridge volume, mean cross-sectional area, and relative height of the center of mass were recorded from the extracted envelopes as shown. See Methods: Data extraction and Supplementary Note 1 for additional details of the data extraction process, including additional bridge metrics.Full size imageFig. 2: Changes in collective structures in experiments.a, b Volume and group size of self-assembled bridges: a Estimated volume of collective bridge structures over time for one focal trial (main figure) and three other examples (inset). The dotted vertical line indicates the time when the experiment shifted from the expansion phase (increasing gap size) to the contraction phase (decreasing gap size). Gray shading indicates that the bridge was broken or recovering from a break; result metrics may be inaccurate during these periods and they were, therefore, excluded from analyses. b The number of ants in the bridge structure over time for the same focal trial (main figure) and three other examples (inset). c–f Hysteresis: Trials consistently show hysteresis, with bridge status at a particular gap size differing during the expansion and contraction phases, for volume (c), number of ants (d), mean cross-sectional area (e), and tautness, or the height of the center of mass of the bridge from the side view (f; lower values indicate bridge is hanging lower). c–f Panels show result metrics over gap size for the same focal trial as in panels a and b, as well as for three other examples (inset). Points show individual measurements, taken every second, lines are smoothed LOESS (local regression) for the expansion (orange points, dashed orange line) and contraction (green points, solid green line) phases. The area between the smoothed lines (shaded gray) shows the extent of hysteresis. c, e, f) Points are jittered to improve clarity. a–f See Supplementary Figs. 2, 3, 5–8 for all complete trials.Full size imageHowever, these changes were not symmetric—adjustments in the contraction phase were not the inverse of adjustments in the expansion phase. We found consistent hysteresis in several metrics; for a given gap size, bridges were larger and made up of more individuals during the contraction of the gap than the expansion (Fig. 2c, d; t-test for volume: mean extent of hysteresis = 0.43, 95% CI = 0.29 to 0.58, t = 6.7, df = 9, p  More

Model descriptionThe model involves a discrete time, discrete valued stochastic Markov process. Model variables and parameters are given in Tables 1 and 2 respectively. Both time and the number of individuals of each type are constrained to be integer valued. Death and reproductive mutation are stochastic processes derived from sampling from binomial distributions given by the relevant probabilities. All ensemble results give the 100-replicate average for the parameter choices in question.Growth of individuals from species ({S}_{1}) and ({S}_{2}) is proportional to the bio-available level of environmental substances ({R}_{1}) and ({R}_{2}) respectively. At time (t) (where time is in units of biological generations) the change in the number ({N}_{q,j}) of individuals of genotype (j) (non-producer, producer, plastic) within species (q) (({S}_{1}) or ({S}_{2})) can be written as a function of the state of the variables at the previous time-step:$${N}_{q,j}left(t+1right)=left(left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)cdot {G}_{q,j}left(tright)-{{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)-{delta }_{q,j}left(tright)+{{{{{{rm{{Upsilon }}}}}}}}_{q,xne j}left(tright)right)cdot left(1-frac{{S}_{q}left(tright)}{K}right)$$
(1)
The leftmost bracket on the right-hand side represents the number of individuals escaping starvation (death due to insufficient environmental substance) at the previous time-step and ({G}_{q,j}left(tright)) is the per capita reproductive growth rate. ({{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)) gives the number of mutant offspring individuals produced during reproduction from parent individuals of genotype (j). ({{{{{{rm{{Upsilon }}}}}}}}_{q,xne j}left(tright)) represents the number of (j) genotype individuals derived from mutation in parent individuals of other genotypes. ({delta }_{q,j}left(tright)) is the number of individuals of genotype (j) lost to random death, and the rightmost bracket relates the total number ({S}_{q}left(tright)) of individuals of species (q) to carrying capacity (K), which represents limitation of growth by any factor other than the relevant environmental substance, e.g. space. (The steady state population size in all simulations shown is below (K) and limited by the environmental substance influx. The carrying capacity is included in the model for computational reasons and as a “crash preventer” but has no qualitative effect on the results).The total number of individuals ({S}_{q}left(tright)) in species (q) is the sum of the number of individuals of each genotype (producer, non-producer and plastic, as discussed in the main text):$${S}_{q}left(tright)=mathop{sum }limits_{j=1}^{{j}_{{total}}}{N}_{q,j}left(tright)={N}_{q,{prod}}left(tright)+{N}_{q,{non}-{prod}}left(tright)+{N}_{q,{plast}}left(tright)$$
(2)
The genotype-specific reproductive growth rate ({G}_{q,j}left(tright)) (again for genotype (j) within species (q), time (t)), gives the number of offspring individuals produced per parent individual, per time-step. Growth rate is an increasing function of the bio-available level of environmental substance ({R}_{q,{BIOAVAIL}{ABLE}}) (the subscript (q) being identical because species ({S}_{1}) and ({S}_{2}) assimilate substances ({R}_{1}) and ({R}_{2}) respectively). Growth rate also includes a substance-to-biomass conversion efficiency parameter ({f}_{{conv}}) and a genotype-specific per capita term ({G}_{q,j,{PR}}) (number of offspring per parent, per unit environmental substance assimilated, per unit time). In the absence of growth-limitation by environmental substance levels, growth rate is capped at a genotype-specific maximum ({G}_{q,{jMAX}}):$${G}_{q,j}left(tright)={MIN}[{G}_{q,j,{PR}}cdot {R}_{q,{BIOAVAILABLE}}(t)cdot {f}_{{conv}},{G}_{q,{jMAX}}]$$
(3)
$${G}_{q,{non}-{prod},{PR}}={G}_{0}$$
(4)
$${G}_{q,{non}-{prod},{MAX}}={G}_{0}cdot {R}_{{assimMAX}}$$
(5)
({G}_{0}) is the baseline number of offspring, per parent, per unit substance assimilated. ({R}_{{assimMAX}}) is a universal maximum potential number of units of environmental substance that can be assimilated by a single individual per time-step (i.e. representing basic physiological constraints on growth). The producer genotype incurs a per capita reproductive growth rate cost ({kappa }_{{prod}}) relative to the non-producer:$${G}_{q,{prod},{PR}}={G}_{0}cdot (1-{kappa }_{{prod}})$$
(6)
$${G}_{q,{prod},{MAX}}={G}_{0}cdot (1-{kappa }_{{prod}})cdot {R}_{{assimMAX}}$$
(7)
This growth rate formulation is therefore a highly simplified linearization of the Michaelis-Menten kinetics normally used in models of resource and nutrient assimilation.The plastic genotype switches phenotype depending upon the level of environmental substance relative to a fixed threshold ({{R}_{q,{BIOAVAILABLE}}}_{{crit}}), in effect becoming a second non-producer genotype below this threshold and a second producer genotype above it:$${IF}[{R}_{q,{BIOAVAILABLE}}(t)ge {{R}_{q,{BIOAVAILABLE}}}_{{crit}}],{G}_{q,{plast}}left(tright)={G}_{q,{prod}}left(tright)$$$${ELSEIF}[{R}_{q,{BIOAVAILA}{BLE}}left(tright) , < , {{R}_{q,{BIOAVAILABLE}}}_{{crit}}],{G}_{q,{plast}}left(tright)={G}_{q,{non}-{prod}}left(tright)$$ (8) There is no spatial structure whatsoever, thus access to environmental substance is uniform across individuals. The bioavailable quantity of each environmental substance is simply the total amount ({R}_{q,{NET}}(t)) divided by the total number of individuals assimilating it:$${R}_{q,{BIOAVAILABLE}}left(tright)=frac{{R}_{q,{NET}}(t)}{{S}_{q}left(tright)}$$ (9) We allow the per capita reproductive growth rate to fall below ({G}_{q,j}left(tright)=1), which, if interpreted deterministically at the individual level would correspond to an individual failing to sustain its biomass to the next time-step and thus dying. However, a population-level average ({G}_{q,j}left(tright) , < , 1) is interpretable in terms of a thinning factor that maps between discretized individuals and continuously distributed environmental substance. Thus, a thinning factor of (left(1-{G}_{q,j}left(tright)right)) is used to calculate the total number of individuals dying of starvation ({rho }_{q,j}) (again genotype (j), species (q)). This represents pre-reproduction deaths, corresponding to the difference between the actual population size and the population size that the environmental substance pool is capable of supporting. ({rho }_{q,j}left(tright)) is constrained to be an integer and is zero for ({G}_{q,j}left(tright) , > , 1):$${rho }_{q,j}left(tright)={N}_{q,j}left(tright)cdot {MAX}left[0,left(1-{G}_{q,j}left(tright)right)right]$$
(10)
A subset of offspring are a different genotype from their parent via mutation. For parent genotype (j), the number ({{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)) of mutant offspring with genotype (ne j) is calculated using baseline mutation probability per reproductive event ({mu }_{0}), with the total number of new individuals produced by the parent individuals surviving starvation ({G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)). The total number of mutant offspring ({{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)) is thus a binomially distributed random variable with success probability ({mu }_{0}) and number of trials ({G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)). The expected value (Eleft[{{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)right]) is the product of these two numbers:$$ {{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright) sim Bleft({G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right),{mu }_{0}right), \ Eleft[{{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)right]={G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)cdot {mu }_{0}$$
(11)
Mutation to genotype (j) from the other genotypes is calculated in exactly the same way using the number and reproductive growth rates of the relevant (other) genotypes. Any particular mutant offspring is randomly allocated to one of the other genotypes with equal probability ({p}_{kto j}=frac{1}{{j}_{{total}}-1}=0.5) (where ({j}_{{total}}=3) is the total number of genotypes per species). The expected number of offspring with genotype (j) produced by mutation within parent offspring of other genotypes (kne j) is therefore:$$E[{{{{{{rm{{Upsilon }}}}}}}}_{q,x , ne , j}left(tright)]={left(mathop{sum }limits_{k=1}^{{k}_{{to}{tal}}}{{{{{{rm{{Upsilon }}}}}}}}_{q,k}left(tright)right)}_{kne j}cdot frac{1}{{j}_{{total}}-1}$$
(12)
Independently of reproduction and assimilation of environmental substance, any given individual has a probability ({delta }_{0}) at each time point of death due to stochastic factors. The genotype/species specific number of such deaths is again a random sample from a binomial distribution, with success probability ({delta }_{0}):$${delta }_{q,j}left(tright) sim Bleft({N}_{q,j}left(tright),{delta }_{0}right),Eleft[{delta }_{q,j}left(tright)right]={N}_{q,j}left(tright)cdot {delta }_{0}$$
(13)
The net quantity of growth-limiting environmental substance at each time-step is given by the difference between total biotic assimilation ({A}_{{R}_{q}}) and the production ({P}_{{R}_{q}}) and abiotic input ({varphi }_{{R}_{q}}) fluxes:$${R}_{q,{NET}}(t+1)={varphi }_{{R}_{q}}(t)+{P}_{{R}_{q}}(t)-{A}_{{R}_{q}}(t)$$
(14)
The abiotic net influx is the sum of two fluxes. First, an input term that is the product of a baseline scaling factor ({{varphi }_{0}}_{{R}_{q}}) and a model forcing (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}) representing the mapping between abiotic-geological and biotic-evolutionary timescales. In practice (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}(t)) was set to either (1) or (0) or (in fluctuation runs) a time-dependent switching between the two. (More sophisticated implementations of (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}(t)), e.g. sinusoidal oscillations and stochastic time dependence, were attempted but made little qualitative difference to the results). Second, an abiotic removal term that scales linearly with the quantity of environmental substance:$${varphi }_{{R}_{q}}(t)={{varphi }_{0}}_{{R}_{q}}cdot frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}left(tright)-frac{{R}_{q,{NET}(t)}}{{R}_{q,{NET}0}}$$
(15)
where ({R}_{q,{NET}0}) is a normalization factor representing the sensitivity of the abiotic efflux to the influx. In the absence of any biota and for (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}=1) the steady state environmental substance level is immediately given by (15) as ({R}_{q,{NET}(t)}={R}_{q,{NET}0}cdot {{varphi }_{0}}_{{R}_{q}}), thus the numerical value of ({R}_{q,{NET}0}) corresponds to the abiotic steady state residence time.Total biotic assimilation ({A}_{R}) of each environmental substance is given by:$${A}_{{R}_{q}}left(tright)=mathop{sum }limits_{k=1}^{{j}_{{total}}}frac{{G}_{q,k}left(tright)cdot left({N}_{q,k}left(tright)-{rho }_{q,k}left(tright)right)}{{G}_{q,{kPR}}}$$
(16)
The numerator gives the total number of individuals produced as a result of biological assimilation of environmental substance ({R}_{q}) and the denominator is the genotype specific number of individuals produced per unit substance assimilated, dividing through by which therefore converts to total units of substance assimilated by the population as a whole.Net biotic ({P}_{{R}_{q}{NET}}) production of substance ({R}_{q}) by the producer genotype in the other species (p) is calculated equivalently, via the product of the per capita production rate ({P}_{q,{prod}}) and the total number of reproducing individuals:$${P}_{q,{prod}}left(tright)=frac{{MIN}[{G}_{q,{prod}}left(tright)cdot {f}_{{convprod}},{G}_{q,{prodMAX}}]}{{G}_{p,{PRODUCER;PR}}}$$
(17)
$${P}_{{R}_{q}{NET}}left(tright)={P}_{q,{prod}}left(tright)cdot left({N}_{p,{PRODUCER}}left(tright)-{rho }_{p,P{RODUCER}}left(tright)right)$$
(18)
Where ({f}_{{conv},{PROD}}) is the per capita efficiency by which producers convert the environmental substance that they assimilate into the by-product they produce (note that the equivalent conversion efficiency for assimilation ({f}_{{conv}}) already appears in the growth functions of each genotype, therefore does not appear in Eq. (17)).The residence time ({T}_{{R}_{q}})of each environmental substance is given by the net quantity of this substance divided by the influx, to give units of the average number of biological generations a unit of environmental substance spends in the relevant pool before being removed. In those simulations in which the abiotic influx ({varphi }_{{R}_{q}}(t)) was set to zero (i.e. during the shut-off intervals) production ({P}_{{R}_{q}{NE}T}left(tright)) was used as an alternative denominator:$${IF}left[{varphi }_{{R}_{q}}left(tright) , > , 0,{T}_{{R}_{q}}left(tright)=frac{{R}_{q,{NET}}left(tright)}{{varphi }_{{R}_{q}}left(tright)}right]!,{IF}left[left(left({varphi }_{{R}_{q}}left(tright)=0right){& }left({P}_{{R}_{q}{NET}}left(tright) , > ,0right)right)!,{T}_{{R}_{q}}left(tright)=frac{{R}_{q,{NET}}left(tright)}{{P}_{{R}_{q}{NET}}left(tright)}right]{ELSE}[{T}_{{R}_{q}}left(tright)=0]$$
(19)
The cycling ratio ({{CR}}_{{R}_{q}}) of each substance is given by the ratio between net biotic assimilation of that substance ({A}_{{R}_{q}}left(tright)) and the abiotic influx of that substance ({varphi }_{{R}_{q}}left(tright)). As with the residence time, when the abiotic influx was zero, the input from biological production was used as an alternative denominator:$${IF}left[{varphi }_{{R}_{q}}left(tright) , > , 0,{{CR}}_{{R}_{q}}left(tright)=frac{{A}_{{R}_{q}}left(tright)}{{varphi }_{{R}_{q}}left(tright)}right]{IF}left[left(left({varphi }_{{R}_{q}}left(tright)=0right){{& }}left({P}_{{R}_{q}{NET}}left(tright) , > ,0right)right),{{CR}}_{{R}_{q}}left(tright)=frac{{A}_{{R}_{q}}left(tright)}{{P}_{{R}_{q}{NET}}left(tright)}right]{ELSE}[{{CR}}_{{R}_{q}}left(tright)=0]$$
(20)
Deterministic approximation to steady stateAssume that at steady state substance assimilation will reach a maximal state such that the level of environmental substance is limiting to population size. Assume that such a state is below the level ({{R}_{q,{BIOAVAILABLE}}}_{{crit}}) at which the plastic genotype effectively becomes a second non-producer genotype and can thus be subsumed into non-producer frequency, such that (2) becomes ({S}_{q}left(tright)=mathop{sum }nolimits_{j=1}^{{j}_{{total}}}{N}_{q,j}left(tright)={N}_{q,{prod}}left(tright)+{N}_{q,{non}-{prod}}left(tright)). Assume that there are non-zero starvations at each time-step for all genotypes, which implies growth rate ({G}_{q,j}left(tright) , < , 1,ll {G}_{q,{jMAX}},forall j,q), which gives by (3)({G}_{q,j}left(tright)={G}_{q,j,{PR}}cdot {R}_{q,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}}). Substituting this into (10), then the first bracketed term in (1), then labeling the post-starvation number of individuals as ({({N}_{q,j}left(tright))}_{{NET}}):$${({N}_{q,j}left(tright))}_{{NET}}=left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)={N}_{q,j}left(tright)cdot left(1-left(1-{G}_{q,j}left(tright)right)right)={N}_{q,j}left(tright)cdot {G}_{q,j}left(tright)$$ (21) Approximate (14) deterministically by a fixed fractional parameter corresponding to the baseline random death rate:$${delta }_{q,j}left(tright)approx {N}_{q,j}left(tright)cdot {delta }_{0}$$ (22) Doing the same for mutation:$${{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)={G}_{q,j}left(tright)cdot {left({N}_{q,j}left(tright)right)}_{{NET}}cdot {mu }_{0}={N}_{q,j}left(tright)cdot {{G}_{q,j}left(tright)}^{2}cdot {mu }_{0}$$ (23) The term in (12) for mutation to (j) from other genotypes simplifies to$${{{{{{rm{{Upsilon }}}}}}}}_{q,x ,ne , j}left(tright)={sum }_{k,=,1}^{{j}_{{total}}}frac{{G}_{q,k}left(tright)cdot {left({N}_{q,k}left(tright)right)}_{{NET}}cdot {mu }_{0}}{{j}_{{total}}-1}={N}_{q,k}left(tright)cdot {{G}_{q,k}left(tright)}^{2}cdot {mu }_{0}$$Substituting Eqs. (21–23) into (1):$${N}_{q,j}left(tright)cdot {{G}_{q,j}left(tright)}^{2}cdot (1-{mu }_{0})-{N}_{q,j}left(tright)cdot {delta }_{0}+{N}_{q,k}left(tright)cdot {{G}_{q,k}left(tright)}^{2}cdot {mu }_{0}=0$$ (24) Noting that by Eqs. (3)–(5) combined with the above assumptions, the growth rate of the producer can be written as:$${G}_{q,{prod}}left(tright)={G}_{q,{cheat}}left(tright)cdot (1-{kappa }_{{prod},q})$$ (25) Writing (24) explicitly for each genotype:$${N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot (1-{mu }_{0})-{N}_{q,{non}-{prod}}left(tright)cdot {delta }_{0}+{N}_{q,{prod}}left(tright)cdot {{G}_{q,{prod}}left(tright)}^{2}cdot {mu }_{0}=0$$$${N}_{q,{prod}}left(tright)cdot {{G}_{q,{prod}}left(tright)}^{2}cdot (1-{mu }_{0})-{N}_{q,{prod}}left(tright)cdot {delta }_{0}+{N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot {mu }_{0}=0$$Adding:$${N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot left(1-{mu }_{0}right)-{N}_{q,{non}-{prod}}left(tright)cdot {delta }_{0}$$$$+{N}_{q,{prod}}left(tright)cdot {left({G}_{q,{non}-{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)}^{2}cdot {mu }_{0}+{N}_{q,{prod}}left(tright)cdot {left({G}_{q,{non}-{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)}^{2}cdot left(1-{mu }_{0}right)$$$$-{N}_{q,{prod}}left(tright)cdot {delta }_{0}+{N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot {mu }_{0}=0$$Because the mutation terms cancel:$$left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot left({{G}_{q,{non}-{prod}}left(tright)}^{2}-{delta }_{0}right)=0$$ (26) Substituting in for the growth rate terms (3–7) gives:$$left({N}_{q,{non}-{prod}}(t)+{N}_{q,{prod}}(t)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot left({left({G}_{0}cdot {R}_{q,{BIOAVAILABLE}}(t)cdot {f}_{{conv}}right)}^{2}-{delta }_{0}right)=0$$ (27) By (16), (4), (6) and the above, total steady state assimilation of the growth limiting environmental substance by species (q) is:$${A}_{{R}_{q}}left(tright)=mathop{sum }limits_{k=1}^{{j}_{{total}}}frac{{G}_{q,k}left(tright)cdot left({N}_{q,k}left(tright)-{rho }_{q,k}left(tright)right)}{{G}_{q,{kPR}}}$$$$kern2.4pc=frac{{N}_{q,{non}-{prod}}left(tright)cdot {left({G}_{q,{non}-{prod}}left(tright)right)}^{2}}{{G}_{0}}+frac{left({N}_{q,{prod}}left(tright)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot {left({G}_{q,{non}-{prod}}left(tright)right)}^{2}}{{G}_{0}cdot left(1-{kappa }_{{prod}}right)}$$$$=left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {{G}_{0}cdot ({R}_{q,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}$$ (28) By (16–18), production of this substance by the producer allele in the other species (p , ne , q), assuming the various arguments above simultaneously apply to this species, is:$${P}_{{R}_{q}}left(tright)= frac{{G}_{p,{prod}}left(tright)cdot left({N}_{p,{prod}}left(tright)-{rho }_{p,{PRODUCER}}left(tright)right)}{{G}_{p,{prodPR}}}cdot {f}_{{conv},{PROD}}\ = {N}_{p,{prod}}left(tright)cdot left(1-{kappa }_{{prod},p}right) cdot {G}_{0}cdot {({R}_{p,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}cdot {f}_{{conv},{PROD}}$$ (29) Balance between input and output fluxes of each environmental substance requires ({varphi }_{{R}_{q}}left(tright)+{P}_{{R}_{q}}left(tright)={A}_{{R}_{q}}left(tright)), meaning that by substituting in ({A}_{{R}_{q}}left(tright)) from (28) it is possible to solve for bioavailable substance level, then substitute in the production flux of this substance derived from the producer allele in the other species (p,ne, q):$${R}_{q,{BIO}{AVAILABLE}}left(tright) =frac{1}{{f}!_{{conv}}}sqrt{frac{{varphi }_{{R}_{q}}left(tright)+{P}_{{R}_{q}}left(tright)}{left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {G}_{0}}}\ =frac{1}{{f}!_{{conv}}}sqrt{frac{{varphi }_{{R}_{q}}left(tright)+{N}_{p,{prod}}left(tright)cdot left(1-{kappa }_{{prod},p}right)cdot {G}_{0}cdot {({R}_{p,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}cdot {f}_{{conv},{PROD}}}{left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {G}_{0}}}$$ (30) Substituting this into (27) gives a symmetrical condition for steady state genotype frequencies and substance levels across the system:$$ left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot\ left(frac{{varphi }_{{R}_{q}}left(tright)+{N}_{p,{prod}}left(tright)cdot left(1-{kappa }_{{prod},p}right)cdot {G}_{0}cdot {({R}_{p,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}cdot {f}_{{conv},{PROD}}}{left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {G}_{0}}-{delta }_{0}right)=0$$ (31) This solution illustrates the intuitive ideas that growth and reproduction balance random death at steady state and that the associated producer frequency is lower than that of the non-producer by a factor of the cost. (This factor is of second order because the growth rate is used both directly and (by (10)) in the calculation of starvations). Because our model is a discrete stochastic process, (31) can be viewed as an approximation to a steady state condition, subject to the above assumptions combined with the continuous generation of producers by mutation at a sufficient rate to preclude their extinction. The key point is that over long timescales in the finite populations with which we deal, organism-level selection unavoidably favors the non-producer, with no possibility for multi-level fecundity selection. The producer’s stable presence is thus attributable to the combination of mutation and cycle-level selection. More

Continental, regional, and local spatiotemporal patterns of carbon lossFor Africa’s primary tropical humid forest, carbon losses due to forest disturbances reached 42.2 ± 5.1 MtC yr−1 (mean ± standard deviation, where MtC yr−1 is one million metric tons of carbon loss per year) in 2019 and 53.4 ± 6.5 MtC yr−1 in 2020. Just 9 countries out of the 23 analyzed accounted for 95.0% of total gross losses in 2019 and 94.3% in 2020. These countries contain about 95.7% of all primary tropical humid forests of Africa, with the DRC accounting for 52.8%, Gabon 11.8%, the Republic of the Congo 11.0%, and Cameroon 9.8%. Of these, DRC and Cameroon were responsible for 49.3% and 19.1% of losses in 2019 and 44.7% and 20.6% in 2020. DRC and Cameroon had an annual increase of 15.0% and 36.5% respectively, between 2019 and 2020. From countries with at least 1 MtC emitted in the two years analyzed, Madagascar had the highest annual increase in carbon loss (+153.9%), while Equatorial Guinea is the only country with a decrease in carbon loss (−20.1%). Extending the carbon loss analysis for both past and future will help to better understand these variations and whether the COVID-19 global pandemic had any influence on the general increase between 2019 and 202019. While the absolute numbers for carbon loss estimates should be treated carefully and a sample-based approach should be preferred for an unbiased estimate of absolute numbers20, we focused our analysis on the trends of carbon loss at the continental, country, and local scale (Fig. 1 and Supplementary Fig. 1).Fig. 1: Carbon loss across Africa’s rainforests.We analyzed 23 countries containing primary moist forest. The aboveground carbon stock (green palette) underlies the carbon loss estimations (red palette). Several hotspots can be seen across these regions. The uncertainties of the carbon loss estimations are expressed as standard deviations and shown in Supplementary Fig. 1.Full size imageThe high temporal detail of the analysis revealed various monthly patterns of carbon losses for countries, highly related to local rainfall patterns18 (Fig. 2). Countries like Cameroon, Liberia, Nigeria, Central African Republic (CAR), and Madagascar showed a clear dry-wet seasonal variation in carbon loss per year, while the Republic of the Congo and the DRC, due to their latitudinal extent, exhibited two dry-wet season variations per year with varying intensities (Fig. 2). The seasonal variation can be explained by higher accessibility to forests during the dry months when activities related to smallholder agriculture and logging are more feasible than in the wet season when many roads become inaccessible.Fig. 2: Temporal patterns of carbon loss for the top 10 countries.We show monthly statistics for 2019 and 2020 and the associated uncertainty (black lines). We separate between high (red bars) and low (yellow bars) confidence alerts, the latter showing up for the last 3 months of 2020.Full size imageOne of the highest differences between the months with the most and the least carbon losses was found for Madagascar (72 times more carbon loss in March compared to November 2019). In CAR, the three consecutive months with the highest cumulative carbon loss (January to March 2020) contributed to 75.7% of the total annual loss (between February and April 2020), in Nigeria 73.9% (January to March 2020), Liberia 73.1% (February to April 2020), Madagascar 70.7% (September to November 2020), and Cameroon 62.2% (January to March 2020). Lower percentages were found for countries with mixed seasonality and patterns, like DRC 36.7% (January to March 2020), and the Republic of the Congo 32.8% (January to March 2020) (Fig. 2). For the latter two countries, we expect better-defined peaks of carbon loss at local scales, where climatic conditions are not mixed. The annual cumulative carbon loss (%) per country (Fig. 3) showed that Liberia, Nigeria, CAR, and Cameroon reached between 70-90% of their annual carbon loss in April, while Madagascar reached 60% in October. The DRC, Gabon, Republic of the Congo, Equatorial Guinea, and Ghana have a more gradual monthly increase of cumulative carbon loss with less contrasting seasonality effects. Monthly patterns of carbon losses between the two years analyzed resulted in a correlation coefficient of 0.94 for the CAR, 0.92 for the DRC, 0.91 for Madagascar, 0.90 for Gabon, and 0.83 for Cameroon (Supplementary Fig. 2). For the Republic of the Congo, the two years correlated 0.51. Knowing the peak months of carbon loss for each country and that these patterns are repeatable from one year to another can contribute to better target and prioritize enforcement activities, as well as predicting future patterns and early reporting of annual forest carbon losses.Fig. 3: Annual cumulative carbon loss (%) for both years analyzed, 2019 and 2020.Africa’s total cumulative carbon loss is shown with a black line. The 10 topmost emitting countries out of 23 countries analyzed are shown and represented by distinct colored lines.Full size imageSeveral hotspots of carbon losses can be seen in Fig. 1. The high spatial and temporal details of our analysis are shown in Fig. 4, where several local examples with different drivers of forest disturbances are shown, like logging roads, selective logging, mining, oil palm plantations, urban expansion, and small-holder agriculture. This kind of information, coupled with auxiliary datasets (e.g., legal concessions, protected areas) can identify the legality of forest disturbance21.Fig. 4: Local examples of approx. 10 × 10 km in extent showing different spatiotemporal patterns and drivers of carbon loss.The first column shows the carbon loss, the second column the associated uncertainty, the third column the day-of-the-year when the loss occurred, and the last column shows the monthly distribution of carbon loss and associated uncertainty for each local example. The center coordinates of each location are shown in the third column as latitude and longitude. Exact locations are shown in Supplementary Fig. 3. a Logging roads and selective logging in the Central African Republic, b mining of gold and titanium in the Republic of the Congo, c development of an oil palm plantation in Cameroon, d forest disturbance related to building a new capital city in Equatorial Guinea, and e small-scale agriculture expansion at the edge of the forest in the DRC.Full size imageImplications of rapid monitoring of local carbon lossNear-real time alerts combined with biomass maps result in spatially explicit forest carbon loss, unlike global tabular statistics of national data22,23. We provide new insights into the spatiotemporal dynamics of carbon loss with consistent assessment of accuracy that could enable transparency and completeness for countries reporting on their REDD + progress to the UNFCCC24. We provide monthly carbon loss estimates that could play a key role in local, national, and international forest initiatives for global carbon policy goals25. Such a system can be implemented with minimal costs and is based on open-source datasets and Google Earth Engine cloud computing platform26, thus enabling cost-effective national monitoring of forest carbon loss7. Providing rapid reporting on the location, time, and amount of carbon lost across Africa’s primary humid forest will help undertake immediate action to protect and conserve carbon-rich threatened forests. Furthermore, countries will be able to predict and estimate their annual carbon loss before a reporting period ends, thus having the opportunity to adjust their practices to meet their country-specific commitments for climate change mitigation initiatives.Limitations and future improvementsWe used the RADD alerts (Radar for Detecting Deforestation)18 with a minimum mapping unit (MMU) of 0.2 ha as accuracy estimates were available for this MMU. Events smaller than 0.2 ha would add to the total carbon loss but are by nature associated with higher uncertainties18. The implications of the RADD alerts using a global humid tropical forest product as a forest baseline for 201816,27,28 are twofold. First, the global nature of this product might result in inconsistencies at the local level18. Second, because the forest cover loss information used to generate the forest baseline is based on optical Landsat data, persistent cloud cover in the second half of 2018 in some areas led to missed reporting of forest disturbances, thus being detected at the beginning of 2019 by the RADD alerts. This possible overestimation of carbon loss at the start of 2019 is not an issue for a near-real-time alerting system since later months are not affected. Furthermore, the alerts do not distinguish between human-induced disturbances and natural forest disturbances18. When a new forest disturbance alert is detected, it will be confirmed or rejected within 90 days by subsequent Sentinel-1 images18. That is why our carbon loss reporting separates between high and low confidence alerts for the last three months of 2020, which is common for most forest disturbance alerting products18,29. We separated all the alerts into core and boundary pixels. Core alerts represent complete tree cover removal and we assumed complete carbon loss within a pixel. For boundary alerts, we assumed a 50% carbon loss since these mainly represent forest disturbances with partial tree cover removal. Detecting and quantifying the level of degradation remains challenging and future developments will minimize this uncertainty by providing variable percentages of degraded forest30. The timeliness and spatial details of future forest disturbance alerting products will improve with the availability of open access long-wavelength radar data from near-future satellite missions (e.g., NISAR L-band SAR in 202331), by using a combination of optical and radar forest disturbance alert products, and integration with high-resolution satellite products.We relied on an aboveground biomass baseline map from 201832, prior to RADD alerts starting from 2019. Biomass estimation for the tropical moist forests is based on ALOS-2 PALSAR-2 L-band satellite and its usage needs to account for the local biases, especially underestimating AGB values higher than 250 Mg ha−1 (ref. 32). Although we reduced this underestimation by adjusting the AGB map based on ground field data, more research is needed on providing up-to-date high-resolution aboveground carbon estimates33 that could further increase the accuracy of local carbon loss estimation. Radar-based estimation of forest carbon stocks is challenging over mountainous terrain and is less accurate in complex canopies3 and future integration of radar and optical satellite data will provide more robust estimates33. Nevertheless, new spaceborne missions (e.g., GEDI34, BIOMASS35) will provide an unprecedented amount of forest structure samples that will improve the algorithms and thus the final accuracy of aboveground biomass estimates.We focused on exploring and analyzing local carbon losses and showing high temporal and spatial patterns of carbon losses. We showed the country statistics to emphasize the temporal dynamics of carbon losses and compare the temporal profiles across our study region. Our approach was not to provide stratified area estimations36 associated with forest disturbances but we used this concept in the sense that we had a stratified sample of higher quality reference data18 to estimate the omission and commission errors and consider those in our uncertainty estimation on the pixel level. The analysis showed that omission and commission errors are small and rather balanced, and thus do not result in a major area bias for the forest disturbances. The uncertainties of the aboveground biomass product32 were adjusted for known regional biases using regional forest biomass plot data sources. With this approach, the original aboveground biomass map bias was partly corrected using a model-based approach deemed to be an alternative to a sample-based approach whenever country data are unavailable37. Our uncertainty analysis and error reduction showed that we expect only minor bias in the forest disturbance and the biomass data and the remaining uncertainties are propagated in our pixel-based uncertainty layer. More

Kraaijeveld, A. R. & Godfray, H. C. J. Trade-off between parasitoid resistance and larval competitive ability in Drosophila melanogaster. Nature 389(6648), 278–280 (1997).ADS 
CAS 
Article 

Google Scholar 
Lochmiller, R. L. & Deerenberg, C. Trade-offs in evolutionary immunology: Just what is the cost of immunity?. Oikos 88(1), 87–98 (2000).Article 

Google Scholar 
Zuk, M. & Stoehr, A. M. Immune defense and host life history. Am. Nat. 160(4), S9–S22 (2002).Article 

Google Scholar 
McKean, K. A. & Nunney, L. Increased sexual activity reduces male immune function in Drosophila melanogaster. Proc. Natl. Acad. Sci. 98(14), 7904–7909 (2001).ADS 
CAS 
Article 

Google Scholar 
Schwenke, R., Lazzaro, B. P. & Wolfner, M. F. Reproduction–immunity trade-offs in insects. Annu. Rev. Entomol. 61(1), 239–256. https://doi.org/10.1146/annurev-ento-010715-023924 (2016).CAS 
Article 
PubMed 

Google Scholar 
McNamara, K. B., Wedell, N. & Simmons, L. W. Experimental evolution reveals trade-offs between mating and immunity. Biol. Lett. 9(4), 20130262. https://doi.org/10.1098/rsbl.2013.0262 (2013).Article 
PubMed 
PubMed Central 

Google Scholar 
Nystrand, M. & Dowling, D. K. Effects of immune challenge on expression of life-history and immune trait expression in sexually reproducing metazoans—a meta-analysis. BMC Biol. 18(1), 135. https://doi.org/10.1186/s12915-020-00856-7 (2020).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Lawniczak, M. K. N. et al. Mating and immunity in invertebrates. Trends Ecol. Evol. 22(1), 48–55 (2007).Article 

Google Scholar 
Ahtiainen, J. J., Alatalo, R. V., Kortet, R. & Rantala, M. J. A trade-off between sexual signalling and immune function in a natural population of the drumming wolf spider Hygrolycosa rubrofasciata. J. Evol. Biol. 18(4), 985–991. https://doi.org/10.1111/j.1420-9101.2005.00907.x (2005).CAS 
Article 
PubMed 

Google Scholar 
Simmons, L. W., Zuk, M. & Rotenberry, J. T. Immune function reflected in calling song characteristics in a natural population of the cricket Teleogryllus commodus. Anim. Behav. 69, 1235–1241. https://doi.org/10.1016/j.anbehav.2004.09.011 (2005).Article 

Google Scholar 
Spencer, K. A., Buchanan, K. L., Leitner, S., Goldsmith, A. R. & Catchpole, C. K. Parasites affect song complexity and neural development in a songbird. Proc. R. Soc. B 272(1576), 2037–2043. https://doi.org/10.1098/rspb.2005.3188 (2005).Article 
PubMed 
PubMed Central 

Google Scholar 
Rantala, M. J., Koskimaki, J., Taskinen, J., Tynkkynen, K. & Suhonen, J. Immunocompetence, developmental stability and wingspot size in the damselfly Calopteryx splendens L. Proc R Soc B 267(1460), 2453–2457 (2000).CAS 
Article 

Google Scholar 
Clotfelter, E. D., Ardia, D. R. & McGraw, K. J. Red fish, blue fish: Trade-offs between pigmentation and immunity in Betta splendens. Behav. Ecol. 18(6), 1139–1145. https://doi.org/10.1093/beheco/arm090 (2007).Article 

Google Scholar 
Rantala, M., Jokinen, I., Kortet, R., Vainikka, A. & Suhonen, J. Do pheromones reveal male immunocompetence?. Proc. R. Soc. B 269, 1681–1685 (2002).Article 

Google Scholar 
Worden, B., Parker, P. & Pappas, P. Parasites reduce attractiveness and reproductive success in male grain beetles. Anim. Behav. 59, 543–550 (2000).CAS 
Article 

Google Scholar 
Barthel, A., Staudacher, H., Schmaltz, A., Heckel, D. G. & Groot, A. T. Sex-specific consequences of an induced immune response on reproduction in a moth. BMC Evol. Biol. 15(1), 282. https://doi.org/10.1186/s12862-015-0562-3 (2015).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Sadd, B. et al. Modulation of sexual signalling by immune challenged male mealworm beetles (Tenebrio molitor L.): Evidence for terminal investment and dishonesty. J. Evol. Biol. 19(2), 321–325. https://doi.org/10.1111/j.1420-9101.2005.01062.x (2006).CAS 
Article 
PubMed 

Google Scholar 
Chemnitz, J., Bagrii, N., Ayasse, M. & Steiger, S. Variation in sex pheromone emission does not reflect immunocompetence but affects attractiveness of male burying beetles—a combination of laboratory and field experiments. Sci. Nat. 104(7), 53. https://doi.org/10.1007/s00114-017-1473-5 (2017).CAS 
Article 

Google Scholar 
Johansson, B. G. & Jones, T. M. The role of chemical communication in mate choice. Biol. Rev. 82(2), 265–289. https://doi.org/10.1111/j.1469-185X.2007.00009.x (2007).Article 
PubMed 

Google Scholar 
Rantala, M. J., Kortet, R., Kotiaho, J. S., Vainikka, A. & Suhonen, J. Condition dependence of pheromones and immune function in the grain beetle Tenebrio molitor. Funct. Ecol. 17(4), 534–540 (2003).Article 

Google Scholar 
Niven, J. E. & Laughlin, S. B. Energy limitation as a selective pressure on the evolution of sensory systems. J. Exp. Biol. 211(11), 1792. https://doi.org/10.1242/jeb.017574 (2008).CAS 
Article 
PubMed 

Google Scholar 
Stöckl, A. et al. Differential investment in visual and olfactory brain areas reflects behavioural choices in hawk moths. Sci. Rep. 6(1), 26041. https://doi.org/10.1038/srep26041 (2016).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Elgar, M. A. et al. Insect antennal morphology: The evolution of diverse solutions to odorant perception. Yale J. Biol. Med. 91(4), 457–469 (2018).PubMed 
PubMed Central 

Google Scholar 
Symonds, M. R. E., Johnson, T. L. & Elgar, M. A. Pheromone production, male abundance, body size, and the evolution of elaborate antennae in moths. Ecol. Evol. 2(1), 227–246. https://doi.org/10.1002/ece3.81 (2012).Article 
PubMed 
PubMed Central 

Google Scholar 
Chapman, R. F. Chemoreception: The significance of receptor numbers. In Advances in Insect Physiology (eds Berridge, M. J. et al.) 247–356 (Academic Press, Cambridge, 1982).
Google Scholar 
Symonds, M. R. E. & Elgar, M. A. The evolution of pheromone diversity. Trends Ecol. Evol. 23(4), 220–228. https://doi.org/10.1016/j.tree.2007.11.009 (2008).Article 
PubMed 

Google Scholar 
Wyatt, T. Pheromones and Animal Behaviour: Communication by Smell and Taste (Cambridge University Press, Cambridge, 2003).Book 

Google Scholar 
Elgar, M. A., Johnson, T. L. & Symonds, M. R. E. Sexual selection and organs of sense: Darwin’s neglected insight. Anim. Biol. 69(1), 63–82. https://doi.org/10.1163/15707563-00001046 (2019).Article 

Google Scholar 
Wang, Q. et al. 2018 Antennal scales improve signal detection efficiency in moths. Proc. R. Soc. B 285, 20172832. https://doi.org/10.1098/rspb.2017.2832 (1874).CAS 
Article 

Google Scholar 
Johnson, T. L., Symonds, M. & Elgar, M. Sexual selection on receptor organ traits: Younger females attract males with longer antennae. Sci. Nat. 104, 1–6 (2017).CAS 
Article 

Google Scholar 
Xu, J. & Wang, Q. Male moths undertake both pre- and in-copulation mate choice based on female age and weight. Behav. Ecol. Sociobiol. 63(6), 801–808. https://doi.org/10.1007/s00265-009-0713-x (2009).MathSciNet 
Article 

Google Scholar 
Fricke, C., Adler, M. I., Brooks, R. C. & Bonduriansky, R. The complexity of male reproductive success: Effects of nutrition, morphology, and experience. Behav. Ecol. 26(2), 617–624. https://doi.org/10.1093/beheco/aru240 (2015).Article 

Google Scholar 
Bernays, E. A. & Chapman, R. F. Phenotypic plasticity in numbers of antennal chemoreceptors in a grasshopper: Effects of food. J. Comp. Physiol. 183(1), 69–76. https://doi.org/10.1007/s003590050235 (1998).CAS 
Article 

Google Scholar 
Johnson, T. L., Symonds, M. R. E. & Elgar, M. A. 2017 Anticipatory flexibility: Larval population density in moths determines male investment in antennae, wings and testes. Proc. R. Soc. B 284(1866), 2017–2087. https://doi.org/10.1098/rspb.2017.2087 (1866).Article 

Google Scholar 
Pomiankowski, A. & Møller, A. P. A resolution of the lek paradox. Proc. R. Soc. Lond. B 260(1357), 21–29. https://doi.org/10.1098/rspb.1995.0054 (1995).ADS 
Article 

Google Scholar 
Cardé, R. & Baker, T. Sexual communication with pheromones. In Chemical Ecology of Insects (eds Bell, W. & Cardé, R.) (Chapman and Hall, London, 1984).
Google Scholar 
Kokko, H. & Wong, B. B. M. What determines sex roles in mate searcing?. Evolution 61(5), 1162–1175. https://doi.org/10.1111/j.1558-5646.2007.00090.x (2007).Article 
PubMed 

Google Scholar 
Alberts, A. Constraints on the design of chemical communication systems in terrestrial vertebrates. Am. Nat. 139, S62–S89 (1992).Article 

Google Scholar 
van Dongen, S., Matthysen, E., Sprengers, E. & Dhondt, A. A. Mate selection by male winter moths Operophtera brumata (Lepidoptera, Geometridae): Adaptive male choice or female control?. Behaviour 135, 29–42 (1998).Article 

Google Scholar 
Henneken, J., Goodger, J. Q. D., Jones, T. M. & Elgar, M. A. Diet-mediated pheromones and signature mixtures can enforce signal reliability. Front. Ecol. Evol. https://doi.org/10.3389/fevo.2016.00145 (2017).Article 

Google Scholar 
Harari, A. R., Zahavi, T. & Thiéry, D. Fitness cost of pheromone production in signaling female moths. Evolution 65(6), 1572–1582. https://doi.org/10.1111/j.1558-5646.2011.01252.x (2011).Article 
PubMed 

Google Scholar 
Pham, H. T., McNamara, K. B. & Elgar, M. A. Socially cued anticipatory adjustment of female signalling effort in a moth. Biol. Lett. 16(12), 20200614. https://doi.org/10.1098/rsbl.2020.0614 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Morgan, F. D. & Cobbinah, J. R. Oviposition and establishment of Uraba lugens (Walker), the gum leaf skeletoniser. Aust. For. 40(1), 44–55. https://doi.org/10.1080/00049158.1977.10675665 (1977).Article 

Google Scholar 
Pham, H. T., McNamara, K. B. & Elgar, M. A. Age-dependent chemical signalling and its consequences for mate attraction in the gumleaf skeletonizer moth, Uraba lugens. Anim. Behav. 173, 207–213. https://doi.org/10.1016/j.anbehav.2020.12.010 (2021).Article 

Google Scholar 
McNamara, K. B., van Lieshout, E., Jones, T. M. & Simmons, L. W. Age-dependent trade-offs between immunity and male, but not female, reproduction. J. Anim. Ecol. 82(1), 235–244. https://doi.org/10.1111/j.1365-2656.2012.02018.x (2012).Article 
PubMed 

Google Scholar 
Simmons, L. W. Resource allocation trade-off between sperm quality and immunity in the field cricket, Teleogryllus oceanicus. Behav. Ecol. 23(1), 168–173. https://doi.org/10.1093/beheco/arr170 (2012).Article 

Google Scholar 
Triseleva, T. A. & Safonkin, A. F. Variation in antennal sensory system in different phenotypes of large fruit-tree tortrix Archips podana Scop (Lepidoptera: Tortricidae). Biol Bull 33(6), 568–572. https://doi.org/10.1134/s1062359006060069 (2006).Article 

Google Scholar 
Rasband, W. S. ImageJ (National Institutes of Health, Maryland USA, 2009).
Google Scholar 
R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, Austria, 2013).
Google Scholar 
Sanes, J. R. & Hildebrand, J. G. Origin and morphogenesis of sensory neurons in an insect antenna. Dev. Biol. 51(2), 300–319. https://doi.org/10.1016/0012-1606(76)90145-7 (1976).CAS 
Article 
PubMed 

Google Scholar 
Gill, K. P., Wilgenburg, E. V., Macmillan, D. L. & Elgar, M. A. Density of antennal sensilla influences efficacy of communication in a social insect. Am. Nat. 182(6), 834–840. https://doi.org/10.1086/673712 (2013).Article 
PubMed 

Google Scholar 
Jayaweera, A. & Barry, K. L. Male antenna morphology and its effect on scramble competition in false garden mantids. Sci. Nat. 104(9), 75. https://doi.org/10.1007/s00114-017-1494-0 (2017).CAS 
Article 

Google Scholar 
Greenfield, M. D. Moth sex pheromones: An evolutionary perspective. Fla Entomol. 64(1), 4–17. https://doi.org/10.2307/3494597 (1981).Article 

Google Scholar 
McNamara, K. B., van Lieshout, E. & Simmons, L. W. The effect of maternal and paternal immune challenge on offspring immunity and reproduction in a cricket. J. Evol. Biol. 27(6), 1020–1028. https://doi.org/10.1111/jeb.12376 (2014).CAS 
Article 
PubMed 

Google Scholar 
Foster, S. P. & Anderson, K. G. 2020 Sex pheromone biosynthesis, storage and release in a female moth: Making a little go a long way. Proc. R. Soc. B 287, 20202775. https://doi.org/10.1098/rspb.2020.2775 (1941).CAS 
Article 

Google Scholar 
Gibb, A. R. et al. Major sex pheromone components of the Australian gum leaf skeletonizer Uraba lugens: (10E,12Z)-hexadecadien-1-yl acetate and (10E,12Z)-hexadecadien-1-ol. J. Chem. Ecol. 34(9), 1125–1133. https://doi.org/10.1007/s10886-008-9523-2 (2008).CAS 
Article 
PubMed 

Google Scholar 
Kerr, A. M., Gershman, S. N. & Sakaluk, S. K. Experimentally induced spermatophore production and immune responses reveal a trade-off in crickets. Behav. Ecol. 21(3), 647–654. https://doi.org/10.1093/beheco/arg035 (2010).Article 

Google Scholar 
Ahmed, A. M., Baggott, S. L., Maingon, R. & Hurd, H. The costs of mounting an immune response are reflected in the reproductive fitness of the mosquito Anopheles gambiae. Oikos 97(3), 371–377 (2002).Article 

Google Scholar 
Hurd, H. Host fecundity reduction: A strategy for damage limitation?. Trends Parasitol. 17(8), 363–368. https://doi.org/10.1016/S1471-4922(01)01927-4 (2001).CAS 
Article 
PubMed 

Google Scholar 
Adamo, S. A. Evidence for adaptive changes in egg laying in crickets exposed to bacteria and parasites. Anim. Behav. 57(1), 117–124. https://doi.org/10.1006/anbe.1998.0999 (1999).CAS 
Article 
PubMed 

Google Scholar 
Arnqvist, G. & Nilsson, T. The evolution of polyandry: Multiple mating and female fitness in insects. Anim. Behav. 60, 145–164 (2000).CAS 
Article 

Google Scholar 
Parker, G. A., Lessells, C. M. & Simmons, L. W. Sperm competition games: A general model for precopulatory male-male competition. Evolution 67(1), 95–109. https://doi.org/10.1111/j.1558-5646.2012.01741.x (2013).Article 
PubMed 

Google Scholar 
Simmons, L. W., Lüpold, S. & Fitzpatrick, J. L. Evolutionary trade-off between secondary sexual traits and ejaculates. Trends Ecol. Evol. 32(12), 964–976. https://doi.org/10.1016/j.tree.2017.09.011 (2017).Article 
PubMed 

Google Scholar 
Parker, G. A. & Pizzari, T. Sperm competition and ejaculate economics. Biol. Rev. 85(4), 897–934. https://doi.org/10.1111/j.1469-185X.2010.00140.x (2010).Article 
PubMed 

Google Scholar 
Katsuki, M. & Lewis, Z. A trade-off between pre- and post-copulatory sexual selection in a bean beetle. Behav. Ecol. Sociobiol. 69(10), 1597–1602. https://doi.org/10.1007/s00265-015-1971-4 (2015).Article 

Google Scholar 
Gage, M. J. G. Continuous variation in reproductive strategy as an adaptive response to population-density in the moth Plodia interpunctella. Proc. R. Soc. B 261(1360), 25–30 (1995).ADS 
Article 

Google Scholar 
Shiel, B. P., Sherman, C. D. H., Elgar, M. A., Johnson, T. L. & Symonds, M. R. E. Investment in sensory structures, testis size, and wing coloration in males of a diurnal moth species: Trade-offs or correlated growth?. Ecol. Evol. 5(8), 1601–1608. https://doi.org/10.1002/ece3.1459 (2015).Article 
PubMed 
PubMed Central 

Google Scholar 
Rolff, J. Bateman’s principle and immunity. Proc. R. Soc. B 269(1493), 867–872. https://doi.org/10.1098/rspb.2002.1959 (2002).Article 
PubMed 
PubMed Central 

Google Scholar 
Calabrese, E. J. & Baldwin, L. A. Hormesis: A generalizable and unifying hypothesis. Crit. Rev. Toxicol. 31(4–5), 353–424. https://doi.org/10.1080/20014091111730 (2001).CAS 
Article 
PubMed 

Google Scholar 
Calabrese, E. J. & Mattson, M. P. How does hormesis impact biology, toxicology, and medicine?. NPJ Aging Mech. Dis. 3(1), 13. https://doi.org/10.1038/s41514-017-0013-z (2017).Article 
PubMed 
PubMed Central 

Google Scholar  More

Geslin, B. et al. Massively introduced managed species and their consequences for plant–pollinator interactions. Adv. Ecol. Res. 57, 147–199 (2017).
Google Scholar 
Huryn, V. M. B. Ecological impacts of introduced honey bees. Q. R. Biol. 72, 275–297 (1997).
Google Scholar 
Stout, J. C. & Morales, C. L. Ecological impacts of invasive alien species on bees. Apidologie 40, 388–409 (2009).
Google Scholar 
Hung, K.-L.J., Kingston, J. M., Albrecht, M., Holway, D. A. & Kohn, J. R. The worldwide importance of honey bees as pollinators in natural habitats. Proc. R. Soc. Ser. B 285, 20172140 (2018).
Google Scholar 
Paini, D. R. Impact of the introduced honey bee (Apis mellifera) (Hymenoptera: Apidae) on native bees: A review. Austral Ecol. 29, 399–407 (2004).
Google Scholar 
Moritz, R. F. A., Hartel, S. & Neumann, P. Global invasions of the western honey bee (Apis mellifera) and the consequences for biodiversity. Ecoscience 12, 289–301 (2005).
Google Scholar 
Paini, D. R. & Roberts, J. D. Commercial honey bees (Apis mellifera) reduce the fecundity of an Australian native bee (Hylaeus alcyoneus). Biol. Cons. 123, 103–112 (2005).
Google Scholar 
Munoz, I. & De la Rua, P. Wide genetic diversity in old world honey bees threatened by introgression. Apidologie 52, 200–217 (2021).
Google Scholar 
Williams, I. H. The dependences of crop production within the European Union on pollination by honey bees. Agric. Zool. Rev. 6, 229–257 (1994).
Google Scholar 
Thompson, C. E., Biesmeijer, J. C., Allnutt, T. R., Pietravalle, S. & Budge, G. E. Parasite pressures on feral honey bees (Apis mellifera sp.). PLoS One 9, e105164 (2014).ADS 
PubMed 
PubMed Central 

Google Scholar 
Belsky, J. & Joshi, N. K. Impact of biotic and abiotic stressors on managed and feral bees. Insects 10, 233 (2019).PubMed Central 

Google Scholar 
Medina-Flores, C. A., Guzman-Novoa, E., Hamiduzzaman, M. M., Arechiga-Flores, C. F. & Lopez-Carlos, M. A. Africanized honey bees (Apis mellifera) have low infestation levels of the mite Varroa destructor in different ecological regions in Mexico. Genet. Mol. Res. 13, 7282–7293 (2014).CAS 
PubMed 

Google Scholar 
Portman, Z. M., Tepedino, V. J., Tripodi, A. D., Szalanski, A. L. & Durham, S. L. Local extinction of a rare plant pollinator in Southern Utah (USA) associated with invasion by Africanized honey bees. Biol. Invasions 20, 593–606 (2018).
Google Scholar 
Santos, G. M. D. et al. Invasive Africanized honeybees change the structure of native pollination networks in Brazil. Biol. Invasions 14, 2369–2378 (2012).
Google Scholar 
Chapman, R. E. & Bourke, A. F. G. The influence of sociality on the conservation biology of social insects. Ecol. Lett. 4, 650–662 (2001).
Google Scholar 
Aizen, M. A. et al. When mutualism goes bad: Density-dependent impacts of introduced bees on plant reproduction. New Phytol. 204, 322–324 (2014).
Google Scholar 
Breeze, T. D. et al. Agricultural policies exacerbate honeybee pollination service supply-demand mismatches across Europe. PLoS One 9, e82996 (2014).ADS 
PubMed 
PubMed Central 

Google Scholar 
Baum, K. A. et al. Spatial distribution of Africanized honey bees in an urban landscape. Landsc. Urban Plan. 100, 153–163 (2011).
Google Scholar 
Ratnieks, F. L. W., Piery, M. A. & Cuadriello, I. The natural nest and nest density of the africanized honey-bee (Hymenoptera, Apidae) near Tapachula, Chiapas, Mexico. Can. Entomol. 123, 353–359 (1991).
Google Scholar 
Baum, K. A., Rubink, W. L., Pinto, M. A. & Coulson, R. N. Spatial and temporal distribution and nest site characteristics of feral honey bee (Hymenoptera: Apidae) colonies in a coastal prairie landscape. Environ. Entomol. 33, 727–739 (2004).
Google Scholar 
Rangel, J. et al. Africanization of a feral honey bee (Apis mellifera) population in South Texas: Does a decade make a difference?. Ecol. Evol. 6, 2158–2169 (2016).PubMed 
PubMed Central 

Google Scholar 
Oldroyd, B. P., Thexton, E. G., Lawler, S. H. & Crozier, R. H. Population demography of Australian feral bees (Apis mellifera). Oecologia 111, 381–387 (1997).ADS 
CAS 
PubMed 

Google Scholar 
Arundel, J. et al. Remarkable uniformity in the densities of feral honey bee Apis mellifera Linnaeus, 1758 (Hymenoptera: Apidae) colonies in South Eastern Australia. Austral Entomol. 53, 328–336 (2014).
Google Scholar 
Remm, J. & Lõhmus, A. Tree cavities in forests—The broad distribution pattern of a keystone structure for biodiversity. For. Ecol. Manag. 262, 579–585 (2006).
Google Scholar 
Lindenmayer, D., Crane, M., Blanchard, W., Okada, S. & Montague-Drake, R. Do nest boxes in restored woodlands promote the conservation of hollow-dependent fauna?. Restor. Ecol. 24, 244–251 (2016).
Google Scholar 
New South Wales Department of Planning, Industry and Environment 2003. https://www.environment.nsw.gov.au/topics/animals-and-plants/threatened-species/nsw-threatened-species-scientific-committee/determinations/final-determinations/2000-2003/competition-from-feral-honeybees-key-threatening-process-listing (accessed 22 Feb 2021).Goldingay, R. L., Rohweder, D. & Taylor, B. D. Nest box contentions: Are nest boxes used by the species they target?. Ecol. Manag. Restor. 21, 115–122 (2020).
Google Scholar 
Lindenmayer, D. B. et al. Are nest boxes a viable alternative source of cavities for hollow-dependent animals? Long-term monitoring of nest box occupancy, pest use and attrition. Biol. Cons. 142, 33–42 (2009).
Google Scholar 
Lindenmayer, D. B. et al. The anatomy of a failed offset. Biol. Conserv. 210, 286–292 (2017).
Google Scholar 
Macak, P. V. Nest boxes for wildlife in Victoria: An overview of nest box distribution and use. Vic. Nat. 137, 4–14 (2020).
Google Scholar 
Le Roux, D. S. et al. Effects of entrance size, tree size and landscape context on nest box occupancy: Considerations for management and biodiversity offsets. For. Ecol. Manag. 366, 135–142 (2016).
Google Scholar 
Berris, K. K. & Barth, M. PVC nest boxes are less at risk of occupancy by feral honey bees than timber nest boxes and natural hollows. Ecol. Manag. Restor. 21, 155–157 (2020).
Google Scholar 
Jaffe, R. et al. Estimating the density of honeybee colonies across their natural range to fill the gap in pollinator decline censuses. Conserv. Biol. 24, 583–593 (2010).PubMed 

Google Scholar 
Utaipanon, P., Schaerf, T. M. & Oldroyd, B. P. Assessing the density of honey bee colonies at ecosystem scales. Ecol. Entomol. 44, 291–304 (2019).
Google Scholar 
Utaipanon, P., Holmes, M. J., Chapman, N. C. & Oldroyd, B. P. Estimating the density of honey bee (Apis mellifera) colonies using trapped drones: Area sampled and drone mating flight distance. Apidologie 50, 578–592 (2019).CAS 

Google Scholar 
Williamson, E. M. Reliability of honey bee hive density estimates using drone sampling: does relative hive size or distance affect a colony’s drone contribution? Honours Thesis, The University of Adelaide (2020).Benson, J. S. The effect of 200 years of European settlement on the vegetation and flora of New South Wales. Cunninghamia 2, 343–370 (1991).
Google Scholar 
New South Wales Office of Environment and Heritage 2015. Upgraded NSW woody vegetation extent for 2011. http://data.auscover.org.au/xwiki/bin/view/Product+pages/nsw+5m+woody+extent+and+fpc (accessed 13 May 2020).R Core Team. R: A Language and Environment for Statistical Computing. (R Foundation for Statistical Computing, 2020). www.R-project.org (accessed 12 January 2021).Burnham, K. P. & Anderson, D. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach (Springer, 2002).MATH 

Google Scholar 
Albert, A. & Anderson, J. A. On the existence of maximum likelihood estimates in logistic regression models. Biometrika 71, 1–10 (1984).MathSciNet 
MATH 

Google Scholar 
Firth, D. Bias reduction of maximum likelihood estimates. Biometrika 80, 27–38 (1993).MathSciNet 
MATH 

Google Scholar 
Kosmidis, I., Pagui, E. C. K. & Sartori, N. Mean and median bias reduction in generalized linear models. Stat. Comput. 30, 43–59 (2020).MathSciNet 
MATH 

Google Scholar 
Anderson, D. R. Model Based Inference in the Life Sciences: A Primer on Evidence (Springer Science & Business Media, 2007).
Google Scholar 
Barton, K. MuMIn: Multi-model inference. R package version 1.43.17 (2016).Hijmans, R. J. Raster: Geographic Data Analysis and Modeling. R package version 3.4-5 (2020).Kassambara, A. ggpubr: ‘ggplot2’ Based Publication Ready Plots. R package version 0.3.0 (2018).Kosmidis, I. brglm2: Bias Reduction in Generalized Linear Models. R package version 0.6.2 (2020).Kosmidis, I., Schumacher, D. detectseparation: Detect and Check for Separation and Infinite Maximum Likelihood Estimates. R package version 0.1 (2020).Pebesma, E. Simple features for R: Standardized support for spatial vector data. R J. 10, 439–446 (2018).
Google Scholar 
Pateiro-Lopez, B., Rodriguez-Casal, A. Alphahull: Generalization of the Convex Hull of a Sample of Points in the Plane. R package version 2.2 (2019).Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer, 2009).MATH 

Google Scholar 
Wickham, H. The split-apply-combine strategy for data analysis. J. Stat. Softw. 40, 1–29 (2011).
Google Scholar 
Wickham, H. Forcats: Tools for working with categorical variables (factors). R package version 0.5.0 (2018).Wickham, H., François, R., Henry, L., Müller, K. dplyr: A Grammar of Data Manipulation. R package version 1.0.0 (2021).Birtchnell, M. J. & Gibson, M. Long-term flowering patterns of melliferous Eucalyptus (Myrtaceae) species. Aust. J. Bot. 54, 745–754 (2006).
Google Scholar 
Steinhauer, N. et al. Drivers of colony losses. Curr. Opin. Insect Sci. 26, 142–148 (2018).PubMed 

Google Scholar 
Cunningham, S. A., Heard, T. & FitzGibbon, F. The future of pollinators for Australian Agriculture. Aust. J. Agric. Res. 53, 893–900 (2002).
Google Scholar 
Hinson, E. M., Duncan, M., Lim, J., Arundel, J. & Oldroyd, B. P. The density of feral honey bee (Apis mellifera) colonies in South East Australia is greater in undisturbed than in disturbed habitats. Apidologie 46, 403–413 (2015).
Google Scholar 
McIntyre, S. Ecological and anthropomorphic factors permitting low-risk assisted colonization in temperate grassy woodlands. Biol. Conserv. 144, 1781–1789 (2011).
Google Scholar 
Steffan-Dewenter, I. & Kuhn, A. Honeybee foraging in differentially structured landscapes. Proc. R. Soc. B Biol. Sci. 270, 569–575 (2003).
Google Scholar 
Wintle, B. A. et al. Global synthesis of conservation studies reveals the importance of small habitat patches for biodiversity. Proc. Natl. Acad. Sci. U.S.A. 116, 909–914 (2019).CAS 
PubMed 

Google Scholar 
Arthur, A. D., Li, J., Henry, S. & Cunningham, S. A. Influence of woody vegetation on pollinator densities in oilseed Brassica fields in an Australian temperate landscape. Basic Appl. Ecol. 11, 406–414 (2010).
Google Scholar 
Lindenmayer, D. B. et al. New policies for old trees: Averting a global crisis in a keystone ecological structure. Conserv. Lett. 7, 61–69 (2014).
Google Scholar 
Crane, M. J., Lindenmayer, D. B. & Cunningham, R. B. The value of countryside elements in the conservation of a threatened arboreal marsupial Petaurus norfolcensis in agricultural landscapes of south-eastern Australia—the disproportional value of scattered trees. PLoS One 9, e107178 (2014).ADS 
PubMed 
PubMed Central 

Google Scholar 
Gibbons, P., Lindenmayer, D. B., Barry, S. C. & Tanton, M. T. Hollow selection by vertebrate fauna in forests of southeastern Australia and implications for forest management. Biol. Conserv. 103, 1–12 (2002).
Google Scholar 
Seeley, T. D. & Morse, R. A. The nest of the honey bee (Apis mellifera L.). Insectes Soc. 23, 495–512 (1976).
Google Scholar 
Hung, K. L. J., Ascher, J. S., Davids, J. A. & Holway, D. A. Ecological filtering in scrub fragments restructures the taxonomic and functional composition of native bee assemblages. Ecology 100, e02654 (2019).PubMed 

Google Scholar 
Cockle, K. L., Martin, K. & Drever, M. C. Supply of tree-holes limits nest density of cavity-nesting birds in primary and logged subtropical Atlantic forest. Biol. Conserv. 143, 2851–2857 (2010).
Google Scholar 
Heard, T. Stingless bees. In Australian Native Bees: A Practical Hand Book 106–139 (NSW Department of Primary Industries, 2016).Geoscience Australia 2006. GEODATA TOPO 250K. Commonwealth of Australia. http://pid.geoscience.gov.au/dataset/ga/63999 (accessed 11 December 2020). More

Sabine, C. L. et al. The oceanic sink for anthropogenic CO2. Science 305, 367–371 (2004).CAS 
PubMed 

Google Scholar 
Landschützer, P. et al. The reinvigoration of the Southern Ocean carbon sink. Science 349, 1221–1224 (2015).PubMed 

Google Scholar 
Dunne, J. P., Sarmiento, J. L. & Gnanadesikan, A. A synthesis of global particle export from the surface ocean and cycling through the ocean interior and on the seafloor. Global Biogeochem. Cycles 21, https://doi.org/10.1029/2006GB002907 (2007).Buesseler, K. O., Boyd, P. W., Black, E. E. & Siegel, D. A. Metrics that matter for assessing the ocean biological carbon pump. Proc. Natl Acad. Sci. USA 117, 9679–9687 (2020).CAS 
PubMed 
PubMed Central 

Google Scholar 
de Baar, H. J. On iron limitation of the Southern Ocean: experimental observations in the Weddell and Scotia Seas. Mar. Ecol. Prog. Ser. 65, 105–122 (1990).
Google Scholar 
Twining, B. S. & Baines, S. B. The trace metal composition of marine phytoplankton. Annu. Rev. Mar. Sci. 5, 191–215 (2013).
Google Scholar 
Martin, J. H., Fitzwater, S. E. & Gordon, R. M. Iron deficiency limits phytoplankton growth in Antarctic waters. Glob. Biogeochem. Cycles 4, 5–12 (1990).CAS 

Google Scholar 
Boyd, P. W. et al. Mesoscale iron enrichment experiments 1993–2005: synthesis and future directions. science 315, 612–617 (2007).CAS 
PubMed 

Google Scholar 
Sunda, W. Feedback interactions between trace metal nutrients and phytoplankton in the ocean. Front. Microbiol. 3, 204 (2012).PubMed 
PubMed Central 

Google Scholar 
Martin, J. H. Glacial‐interglacial CO2 change: the iron hypothesis. Paleoceanography 5, 1–13 (1990).
Google Scholar 
Martin, J. H., Gordon, R. M. & Fitzwater, S. E. Iron in Antarctic waters. Nature 345, 156 (1990).CAS 

Google Scholar 
Moore, C. M. et al. Processes and patterns of oceanic nutrient limitation. Nat. Geosci. 6, 701–710 (2013).CAS 

Google Scholar 
Behrenfeld, M. J. & Milligan, A. J. Photophysiological expressions of iron stress in phytoplankton. Annu. Rev. Mar. Sci. 5, 217–246 (2013).
Google Scholar 
Greene, R. M., Geider, R. J., Kolber, Z. & Falkowski, P. G. Iron-induced changes in light harvesting and photochemical energy conversion processes in eukaryotic marine algae. Plant Physiol. 100, 565–575 (1992).CAS 
PubMed 
PubMed Central 

Google Scholar 
Raven, J. A., Evans, M. C. & Korb, R. E. The role of trace metals in photosynthetic electron transport in O2-evolving organisms. Photosynthesis Res. 60, 111–150 (1999).CAS 

Google Scholar 
Raven, J. A. Predictions of Mn and Fe use efficiencies of phototrophic growth as a function of light availability for growth and of C assimilation pathway. N. Phytologist 116, 1–18 (1990).CAS 

Google Scholar 
Wolfe-Simon, F., Grzebyk, D., Schofield, O. & Falkowski, P. G. The role and evolution of superoxide dismutases in algae 1. J. Phycol. 41, 453–465 (2005).CAS 

Google Scholar 
Middag, R. D., De Baar, H. J. W., Laan, P., Cai, P. V. & Van Ooijen, J. C. Dissolved manganese in the Atlantic sector of the Southern Ocean. Deep Sea Res. Part II: Topical Stud. Oceanogr. 58, 2661–2677 (2011).CAS 

Google Scholar 
Buma, A. G., De Baar, H. J., Nolting, R. F. & Van Bennekom, A. J. Metal enrichment experiments in the Weddell‐Scotia Seas: effects of iron and manganese on various plankton communities. Limnol. Oceanogr. 36, 1865–1878 (1991).CAS 

Google Scholar 
Middag, R., de Baar, H. J., Klunder, M. B. & Laan, P. Fluxes of dissolved aluminum and manganese to the Weddell Sea and indications for manganese co‐limitation. Limnol. Oceanogr. 58, 287–300 (2013).CAS 

Google Scholar 
Browning, T. J. et al. Strong responses of Southern Ocean phytoplankton communities to volcanic ash. Geophys. Res. Lett. 41, 2851–2857 (2014).CAS 

Google Scholar 
Wu, M. et al. Manganese and iron deficiency in Southern Ocean Phaeocystis Antarctica populations revealed through taxon-specific protein indicators. Nat. Commun. 10, 1–10 (2019).
Google Scholar 
Browning, T. J., Achterberg, E. P., Engel, A. & Mawji, E. Manganese co-limitation of phytoplankton growth and major nutrient drawdown in the Southern Ocean. Nat. Commun. 12, 884 (2021).CAS 
PubMed 
PubMed Central 

Google Scholar 
Viljoen, J. J. et al. Links between the phytoplankton community composition and trace metal distribution in summer surface waters of the Atlantic southern ocean. Front. Mar. Sci. 6, 295 (2019).
Google Scholar 
Arrigo, K. R. Marine microorganisms and global nutrient cycles. Nature 437, 349–355 (2005).CAS 
PubMed 

Google Scholar 
De Baar, H. J. W. von Liebig’s law of the minimum and plankton ecology (1899–1991). Prog. Oceanogr. 33, 347–386 (1994).
Google Scholar 
Saito, M. A., Goepfert, T. J. & Ritt, J. T. Some thoughts on the concept of colimitation: three definitions and the importance of bioavailability. Limnol. Oceanogr. 53, 276–290 (2008).CAS 

Google Scholar 
Pausch, F., Bischof, K. & Trimborn, S. Iron and manganese co-limit growth of the Southern Ocean diatom Chaetoceros debilis. PLos ONE 14, e0221959 (2019).CAS 
PubMed 
PubMed Central 

Google Scholar 
Hopkinson, B. M. et al. Iron limitation across chlorophyll gradients in the southern Drake Passage: phytoplankton responses to iron addition and photosynthetic indicators of iron stress. Limnol. Oceanogr. 52, 2540–2554 (2007).CAS 

Google Scholar 
Trimborn, S., Hoppe, C. J., Taylor, B. B., Bracher, A. & Hassler, C. Physiological characteristics of open ocean and coastal phytoplankton communities of Western Antarctic Peninsula and Drake Passage waters. Deep Sea Res. Part I: Oceanographic Res. Pap. 98, 115–124 (2015).CAS 

Google Scholar 
Rijkenberg, M. J. et al. The distribution of dissolved iron in the West Atlantic Ocean. PLoS ONE 9, e101323 (2014).PubMed 
PubMed Central 

Google Scholar 
Prézelin, B. B., Hofmann, E. E., Mengelt, C. & Klinck, J. M. The linkage between upper circumpolar deep water (UCDW) and phytoplankton assemblages on the west Antarctic Peninsula continental shelf. J. Mar. Res. 58, 165–202 (2000).
Google Scholar 
Varela, M., Fernandez, E. & Serret, P. Size-fractionated phytoplankton biomass and primary production in the Gerlache and south Bransfield Straits (Antarctic Peninsula) in Austral summer 1995–1996. Deep Sea Res. Part II: Topical Stud. Oceanogr. 49, 749–768 (2002).CAS 

Google Scholar 
Hoffmann, L. J., Peeken, I. & Lochte, K. Effects of iron on the elemental stoichiometry during EIFEX and in the diatoms Fragilariopsis kerguelensis and Chaetoceros dichaeta. Biogeosciences 4, 569–579 (2007).CAS 

Google Scholar 
Blanco-Ameijeiras, S. et al. Exopolymeric substances control microbial community structure and function by contributing to both C and Fe nutrition in Fe-limited Southern Ocean provinces. Microorganisms 8, 1980 (2020).CAS 
PubMed Central 

Google Scholar 
Church, M. J., Hutchins, D. A. & Ducklow, H. W. Limitation of bacterial growth by dissolved organic matter and iron in the Southern Ocean. Appl. Environ. Microbiol. 66, 455–466 (2000).CAS 
PubMed 
PubMed Central 

Google Scholar 
Obernosterer, I., Fourquez, M. & Blain, S. Fe and C co-limitation of heterotrophic bacteria in the naturally fertilized region off the Kerguelen Islands. Biogeosciences 12, 1983–1992 (2015).
Google Scholar 
Fourquez, M., Obernosterer, I., Davies, D. M., Trull, T. W. & Blain, S. Microbial iron uptake in the naturally fertilized waters in the vicinity of the Kerguelen Islands: phytoplankton–bacteria interactions. Biogeosciences 12, 1893–1906 (2015).
Google Scholar 
Fourquez, M. et al. Microbial competition in the subpolar southern ocean: an Fe–C Co-limitation experiment. Front. Mar. Sci. 6, 776 (2020).
Google Scholar 
Boyd, P. W. et al. A mesoscale phytoplankton bloom in the polar Southern Ocean stimulated by iron fertilization. Nature 407, 695–702 (2000).CAS 
PubMed 

Google Scholar 
De Baar, H. J. et al. Synthesis of iron fertilization experiments: from the iron age in the age of enlightenment. J. Geophys. Res. Oceans 110, https://doi.org/10.1029/2004JC002601 (2005).Smetacek, V. & Naqvi, S. W. A. The next generation of iron fertilization experiments in the Southern Ocean. Philos. Trans. R. Soc. A: Math., Phys. Eng. Sci. 366, 3947–3967 (2008).CAS 

Google Scholar 
Geider, R. J. & La Roche, J. The role of iron in phytoplankton photosynthesis, and the potential for iron-limitation of primary productivity in the sea. Photosynthesis Res. 39, 275–301 (1994).CAS 

Google Scholar 
van Leeuwe, M. A. & Stefels, J. Effects of iron and light stress on the biochemical composition of Antarctic Phaeocystis sp. (Prymnesiophyceae). II. Pigment composition. J. Phycol. 34, 496–503 (1998).
Google Scholar 
Hoffmann, L. J., Peeken, I., Lochte, K., Assmy, P. & Veldhuis, M. Different reactions of Southern Ocean phytoplankton size classes to iron fertilization. Limnol. Oceanogr. 51, 1217–1229 (2006).CAS 

Google Scholar 
Koch, F., Beszteri, S., Harms, L. & Trimborn, S. The impacts of iron limitation and ocean acidification on the cellular stoichiometry, photophysiology, and transcriptome of Phaeocystis antarctica. Limnol. Oceanogr. 64, 357–375 (2019).CAS 

Google Scholar 
Koch, F. & Trimborn, S. Limitation by Fe, Zn, Co, and B12 results in similar physiological responses in two antarctic phytoplankton species. Front. Mar. Sci. 6, 514 (2019).
Google Scholar 
Peers, G. & Price, N. M. A role for manganese in superoxide dismutases and growth of iron‐deficient diatoms. Limnol. Oceanogr. 49, 1774–1783 (2004).CAS 

Google Scholar 
Cefarelli, A. O. et al. Diversity of the diatom genus Fragilariopsis in the Argentine Sea and Antarctic waters: morphology, distribution and abundance. Polar Biol. 33, 1463–1484 (2010).
Google Scholar 
Marchetti, A. & Harrison, P. J. Coupled changes in the cell morphology and elemental (C, N, and Si) composition of the pennate diatom Pseudo-nitzschia due to iron deficiency. Limnol. Oceanogr. 52, 2270–2284 (2007).CAS 

Google Scholar 
Boyd, P. W. et al. Microbial control of diatom bloom dynamics in the open ocean. Geophys. Res. Lett. 39, https://doi.org/10.1029/2012GL053448 (2012).Behrenfeld, M. J. & Kolber, Z. S. Widespread iron limitation of phytoplankton in the South Pacific Ocean. Science 283, 840–843 (1999).CAS 
PubMed 

Google Scholar 
Strzepek, R. F., Hunter, K. A., Frew, R. D., Harrison, P. J. & Boyd, P. W. Iron‐light interactions differ in Southern Ocean phytoplankton. Limnol. Oceanogr. 57, 1182–1200 (2012).CAS 

Google Scholar 
Klunder, M. B. et al. Dissolved Fe across the Weddell Sea and Drake Passage: impact of DFe on nutrient uptake. Biogeosciences 11, 651–669 (2014).
Google Scholar 
Trimborn, S. et al. Iron sources alter the response of Southern Ocean phytoplankton to ocean acidification. Mar. Ecol. Prog. Ser. 578, 35–50 (2017).CAS 

Google Scholar 
Smith, W. O. & Lancelot, C. Bottom-up versus top-down control in phytoplankton of the Southern Ocean. Antarct. Sci. 16, 531–539 (2004).
Google Scholar 
Schoffman, H., Lis, H., Shaked, Y. & Keren, N. Iron–nutrient interactions within phytoplankton. Front. Plant Sci. 7, 1223 (2016).PubMed 
PubMed Central 

Google Scholar 
Meijers, A. J. S. The Southern Ocean in the coupled model intercomparison project phase 5. Philos. Trans. R. Soc. A: Math., Phys. Eng. Sci. 372, 20130296 (2014).CAS 

Google Scholar 
Hauck, J. et al. On the Southern Ocean CO2 uptake and the role of the biological carbon pump in the 21st century. Glob. Biogeochem. Cycles 29, 1451–1470 (2015).CAS 

Google Scholar 
Cabanes, D. J. et al. Using Fe chemistry to predict Fe uptake rates for natural plankton assemblages from the Southern Ocean. Mar. Chem. 225, 103853 (2020).CAS 

Google Scholar 
Cutter, G. A. et al. Sampling and sample-handling protocols for GEOTRACES Cruises, Version 3.0 (2017).Gerringa, L. J. A., De Baar, H. J. W. & Timmermans, K. R. A comparison of iron limitation of phytoplankton in natural oceanic waters and laboratory media conditioned with EDTA. Mar. Chem. 68, 335–346 (2000).CAS 

Google Scholar 
Hoppe, C. J. et al. Iron limitation modulates ocean acidification effects on Southern Ocean phytoplankton communities. PLoS ONE 8, e79890 (2013).PubMed 
PubMed Central 

Google Scholar 
Hathorne, E. C. et al. Online preconcentration ICP-MS analysis of rare earth elements in seawater. Geochemistry, Geophysics, Geosystems 13, https://doi.org/10.1029/2011GC003907 (2012).Rapp, I., Schlosser, C., Rusiecka, D., Gledhill, M. & Achterberg, E. P. Automated preconcentration of Fe, Zn, Cu, Ni, Cd, Pb, Co, and Mn in seawater with analysis using high-resolution sector field inductively-coupled plasma mass spectrometry. Analytica Chim. Acta 976, 1–13 (2017).CAS 

Google Scholar 
Utermöhl, H. Zur vervollkommnung der quantitativen phytoplankton-methodik: Mit 1 Tabelle und 15 abbildungen im Text und auf 1 Tafel. Int. Ver. f.ür. theoretische und Angew. Limnologie: Mitteilungen 9, 1–38 (1958).
Google Scholar 
Edler, L. Recommendations on Methods for Marine Biological Studies in the Baltic Sea. Phytoplankton and Chlorophyll (Publication-Baltic Marine Biologists BMB (Sweden), 1979).Tomas, C. R. & Haste, G. R. Identifying Marine Phytoplankton (Academic Press, 1997).Olson, R. J., Zettler, E. R., Chisholm, S. W. & Dusenberry, J. A. in Particle Analysis in Oceanography 351–399 (Springer, 1991).Koch, F., Sanudo-Wilhelmy, S. A., Fisher, N. S. & Gobler, C. J. Effect of vitamins B1 and B12 on bloom dynamics of the harmful brown tide alga, Aureococcus anophagefferens (Pelagophyceae). Limnol. Oceanogr. 58, 1761–1774 (2013).CAS 

Google Scholar 
Welschmeyer, N. A. Fluorometric analysis of chlorophyll a in the presence of chlorophyll b and pheopigments. Limnol. Oceanogr. 39, 1985–1992 (1994).CAS 

Google Scholar 
Oxborough, K. et al. Direct estimation of functional PSII reaction center concentration and PSII electron flux on a volume basis: a new approach to the analysis of Fast Repetition Rate fluorometry (FRRf) data. Limnol. Oceanogr.: Methods 10, 142–154 (2012).
Google Scholar 
Schlitzer, R. Ocean Data View. (2015). More

Many women in science, technology, engineering and mathematics (STEM) need to make decisions about marital name change, and have to consider how this might affect their publication record and future career. Mentorship that considers race, ethnicity, culture, religion and parenting, as well as a centralized system to dynamically and retroactively streamline name change, will promote agency and choice for women navigating STEM careers, writes Bala Chaudhary.Women, whether in same-sex or heterosexual relationships, still predominantly make decisions regarding marital name change1. In science, technology, engineering and mathematics (STEM) fields, as the proportion of female researchers rises, more women are considering the potential effects of marital name change on their careers. The stakes are high, as relationship status and name discrimination contribute to gender2 and racial3 inequities in faculty hiring. The shifting demographics of students and a greater proportion of STEM undergraduates engaging in research and publishing has also led to more scientists questioning decisions around name changes. Dual-scientist couples considering sharing a last name may wonder about gendered assessments of their contributions to work. Women occasionally ask for advice on this topic using social-media platforms such as Twitter. Community members chime in with myriad options: keep your name, change your name, hyphenate, add a middle name, couples choose a new name, keep separate personal and legal names, and so on. There is no single correct approach for this personal decision, so online discussions and testimonials4 are invaluable resources for women with few immediate role models. More

The LCA approach includes the potential effects of depriving humans and ecosystems of water resources, as well as the specific potential effects of pollutants affecting water and thus the environment49. Water stress is commonly defined as the ratio of total freshwater consumption to the level of its hydrological availability. ISO 14046 presents a new concept, i.e., WF, which is associated with the LCA approach. The standard’s “water scarcity footprint” refers to the potential impacts associated with the quantitative aspect of water use50. Figure 2 shows the WF per cultivation area of conventional and organic carrot production. In general, there are significant differences in the total value of the WF in question. For conventional carrot production technology, it is 10.25 m3 ha−1, while for organic technology, it is only 1.96 m3 ha−1. In the case of conventional production, treatments using significant amounts of chemicals have the greatest impact on the WF, i.e., fertilization (mainly mineral) (WF = 6.85 m3 ha−1), and chemical plant protection (WF = 1.19 m3 ha−1). The analysis of WF in organic farming showed that its highest value (WF = 0.84 m3 ha−1) concerns the harvesting of carrots, while soil preparation ranks second (WF = 0.45 m3 ha−1). A slightly lower WF of 0.38 m3 ha−1 was recorded in the case of transporting the harvested carrots to the farm buildings. It can therefore be concluded that in organic farming, it is (diesel) fuel consumption that has the greatest impact on WF level.Figure 2Water footprint in conventional and organic carrot production (m3 ha−1).Full size imageIn terms of production volume, in conventional technology, the WF is 0.196 m3 t−1. On the other hand, in organic technology the value of WF is approx. four times lower and amounts to 0.049 m3 t−1 of harvested carrots. For comparison, the WF in tomato production is 160 m3 per 1 tonne of produce51. Such a high value results mainly from irrigation of the plants.In order to explain in detail the impact of individual agricultural treatments on the water deficit in carrot production, a detailed WF analysis was carried out for the treatments that demonstrated the highest values. In the case of conventional technology, it was fertilization (Fig. 3). Upon analyzing Fig. 3, it can be observed that the use of urea, and hence nitrogen, has the greatest impact on WF with regard to fertilization. Most nitrogen mineral fertilizers have a negative impact on the environment, causing ozone depletion in the stratosphere, groundwater pollution, global warming, and water eutrophication52,53. The largest water footprint associated with the use of mineral fertilizers in conventional cultivation is mainly due to a very energy-intensive fertilizer production process. Depending on the type of fertilizer and the technology used, the production process involves machines and equipment for cleaning, grinding, drying, sieving, extruding, granulating, packing, pumping, evaporation (crystallization) and transport. The vast majority of these treatments are powered by electricity. In contrast, conventional power production, regardless of the technology and fuel used (nuclear, natural gas, or coal), is characterized by very high water consumption. Mineral fertilizers are also a material whose consumed mass is relatively high compared to other production materials (seeds, pesticides, and diesel fuel). These two factors mentioned above have a decisive impact on the largest water footprint associated with the use of mineral fertilizers in conventional carrot cultivation. Processes requiring the use of machinery, i.e., fertilizer spreaders (1%) and the consumption of diesel fuel (0.1%) have the lowest impact on the level of fertilization-induced WF. Such a low impact of diesel fuel results mainly from its relatively low consumption during fertilization, most often using very efficient centrifugal spreaders. For comparison, WF related only to the use of carrot irrigation water is 20 m3 t−1 of harvested crops54.Figure 3Water footprint related to carrot fertilization in conventional production (m3 ha−1).Full size imageIn the case of organic technology, WF of harvesting was analyzed in detail (Fig. 4). Carrots were excavated with harvesters, which cut the aboveground parts, cleaned the roots and collected them in a hopper. Sometimes the excavation was preceded by mowing the carrot leaves with mowers. Carrot harvesters are machines that require farm tractors with high-power combustion engines, and the harvesting procedure itself is very time-consuming, hence such a large impact of fuel consumption on WF in carrot harvesting. Despite the above, the share of diesel consumption in the total value of WF related to carrot harvesting is only 11%. However, when comparing the WF related to fuel consumed during harvesting and during fertilization, it can be noticed that in the case of harvesting, WF is approx. 15 times higher.Figure 4Water footprint related to carrot harvest in organic production (m3 ha−1).Full size imageIn LCA, the potential effects of water pollution have traditionally been addressed in impact categories such as (eco) toxicity, acidification, and eutrophication42,43. In the WF analysis, the impact of water consumption is generally related to specific goals within a given conservation area, such as: Human Health, Ecosystems Quality and Resources43. The impact of water consumption on human health is expressed in DALY and is obtained by modeling the cause-effect chain of water scarcity (lack of irrigation water) leading to malnutrition. Ecosystem quality is assessed by modeling the cause-effect chain of freshwater consumption with the quality of the terrestrial ecosystem, based on the number of species disappearing each year (species * year). On the other hand, the impact of water consumption in the resources category is assessed by modeling the cause-effect chain of freshwater consumption in relation to the depletion of water resources, along with the cost ($) of extracting an additional cubic meter of water46. The data in Table 2 shows WF in conventional carrot production related to the three impact categories, and Fig. 5 shows its structure. The total impact of individual processes in the Human Health category is 1.15E−05 DALY, in the Ecosystem Quality category—1.53E−07 species * year, and in the Resources category—2.97 $ surplus. For comparison, WF in the above-mentioned impact areas per 1 ha of tomatoes is, respectively: Human Health—5.00E−03 DALY, Ecosystem Quality—2.50E−05 species * year55. When analyzing Fig. 5, it can be observed that in all impact categories, fertilization has the greatest environmental impact, the share of which in individual categories is at approx. 67.0–67.7%. Chemical plant protection ranks second, the impact of which in the three categories ranges from 11.9 to 12.6%. In addition to the treatments related to fertilizers and chemicals, treatments associated with high consumption of diesel fuel, i.e., soil preparation and harvest, have a significant impact on the value of individual categories in carrot production. This confirms the results of many studies, i.e. that the extraction, production and, above all, the use of diesel fuel bring significant damage to the environment56,57.Table 2 Environmental impact related to the use of water in conventional carrot production per area unit (ha).Full size tableFigure 5The structure of WF in individual impact categories in conventional carrot production.Full size imageBearing in mind that carrot yield range in conventional cultivation is 43–65 t ha−1, the total impact of individual processes per 100 tons of harvested carrots is as follows: Human Health: 2.17E−05 DALY, Ecosystem Quality: 2.88E−07 species * year and Resources: 5.57 $ surplus.Upon comparing the obtained results with the research presented in the literature and conducted with a similar methodology, it can be concluded that for the production of 100 tons of tomatoes, the total environmental footprint for the above-mentioned impact areas, including factors other than water, is respectively: Human Health: 2.7E−01 DALY, Ecosystems Quality: 1.45E−03 Species * year, Resources: 1.05E + 06 $58. On the other hand, the WF for green beans, per 100 tons of harvest was reported as follows: Human Health: from 2.00E−2 to 1.08E−1 DALY, Ecosystem Quality: from1.10E−3 to 1.80E−3 species * year, Resources: from 1.90E+2 to 1.40E+3 $ surplus59.Detailed WF results for the fertilization process in conventional carrot production are presented in Fig. 6. Among the individual factors shaping the environmental impact, what stands out is the consumption of urea, i.e. nitrogen (44.9–47.0% of the total impact in individual categories) and of phosphorus fertilizers, the impact of which is at 31.4–32.4%.Figure 6Structure of WF of the fertilization process in conventional carrot production as per individual impact categories.Full size imageIn endpoint analysis, the impact of water use is generally related to specific endpoints in a given conservation area: Human Health, Ecosystems Quality or Resources43. The data in Table 3 shows WF in organic carrot production as per the three impact categories, and Fig. 7 shows its structure. The total WF values in each category are as follows: in the Human Health category—2.11E−06 DALY, in the Ecosystem Quality category—3.00E−08 species * year and in the Resources category—0.56 $ surplus. The above results are over five times lower compared to the footprint in conventional production (Table 2), and therefore it can be concluded that organic production not only enables the production of healthy carrot, but also has a very positive impact on the broadly understood environment. Upon analyzing the data from Tables 2 and 3, it can be observed that the environmental impact of fertilization treatment in organic production is over thirty times lower compared to the impact of fertilization in conventional production. Moreover, the fact that no pesticides are used means that the impact of chemical plant protection treatments is 0. Upon analyzing Fig. 7, it can be observed that the largest share in the total value of WF in individual impact categories is that of carrot harvest, from 41.9% (Ecosystem Quality) to 43.1% (Resources). The reason for such a significant environmental impact of carrot harvesting technology is the use of complex harvesters, as explained in Fig. 8. When calculating WF, both direct water consumption in a technology is taken into account, as well as indirect, related e.g., to the production of agricultural equipment used in the technology. The complexity of the machinery, the type of materials it is made of and the type of technology used in its production determine the WF.Table 3 Environmental impact related to the use of water in organic carrot production per area unit (ha).Full size tableFigure 7The structure of WF in individual impact categories in organic carrot production.Full size imageFigure 8Structure of WF of the harvest process in organic carrot production as per individual impact categories.Full size imageThe methodology for calculating WF is very diverse and includes many methods. Moreover, the results of research on WF related to the production of vegetable species presented in the literature often differ in terms of the analyzed system boundaries, production technology, irrigation, etc. Therefore, the possibility of a broad discussion of the results of WF of conventional and organic carrot production is limited.The detailed structure of WF of the carrot harvesting process in organic farming is shown in Fig. 8. The use of machines, i.e. harvesters, has a decisive share (90.3–96.6%) in the total value of individual impact categories. The reason for this significant impact was explained above. Relatively high consumption of fuel during harvester operation contributes little to the water footprint structure. The use of diesel fuel has the highest impact on damage caused in the Ecosystem Quality category, with the lowest impact in the Human Health category. More