In this new post, Functional Ecology announces that a ‘Replication Statement’ is now mandatory for research articles to be considered for publication. Senior Editor for Functional Ecology, Dustin Marshall, provides some background information on this new requirement and a handful examples to assist authors.

Check out our updated Author Guidelines if you are looking to submit a manuscript to Functional Ecology.

Rationale

Measuring independent replicates at the appropriate scale lies at the heart of good experimental design. However, identifying how many replicates have been measured, and at what scales, is increasingly difficult – especially as some modern statistical analyses lack easily accessible indicators of replication. Furthermore, replicates can be measured at multiple scales simultaneously, so it is not always easy to identify which scale of replication is used for a given inference. Therefore, in the interests of promoting best practice in experimental design, we now require authors to include a statement regarding key elements of replication and experimental design when submitting a manuscript to Functional Ecology.

Requirement

In the Materials and Methods section (before the description of the data analyses), authors MUST state i) the scale at which they seek to make inferences (for example at the level of species, populations or experimental units); ii) the scale(s) at which their treatment or factor of interest is(are) applied; and iii) the number of replicates for each level of treatment or factor. We have provided a table template for including this information, which the editors will use to decide whether the authors’ inferences are supported and should therefore be peer-reviewed. Manuscripts without this table will be returned to authors.

Please note that the one exemption to this requirement is if your paper is theory-based.

Scale of inference	Scale at which the factor of interest is applied	Number of replicates at the appropriate scale

Examples

Example 1.

Authors wish to understand how viviparity affects the evolution of body size among species, the measure 25 individuals at of each species, for 5 vivparous species and 6 oviparous species. In this example, the scale of inference is the species level, the factor of interest is viviparity/oviparity and differs at the scale of species, and the unit of replication at the appropriate level is 5 and 6 for each factor level.

Scale of inference	Scale at which the factor of interest is applied	Number of replicates at the appropriate scale
Species	Species	5 oviparous species, 6 viviparous species

Example 2.

Authors wish to examine the effects of temperature and food on the growth rate of a species of fish. The authors have 24 tanks, each containing 100 fish; the fish in 12 tanks receive a high food regime, the fish in the other 12 tanks receive a low food regime. Within each of the food regimes, 6 tanks are at 25°C and 6 are at 20°C.  In this example, the level of inference is among individuals. The factors of interest are food and temperature and the unit of replication is tank, hence, they have 6 replicate tanks of each temperature/food combination. The individual fish in this example are subsamples, which help with the precision of the estimate at the tank level but not independent replicates.

Scale of inference	Scale at which the factor of interest is applied	Number of replicates at the appropriate scale
Individuals	Tanks	6 of each combination

Example 3.

Authors wish to examine whether a species of plant growing on mountaintops has more of a specific protein than species growing in valleys. They sample 40 mountaintop sites and 30 valley sites, they collect 5 individuals from each site, pool individuals from each of 10 sites to obtain enough material and then extract protein from 5 technical replicates for each pooled sample. In this example, pooling individuals from multiple sites means that the samples from each site are no longer independent and hence the pooled sample is the appropriate scale of replication. The technical replicates likewise do not contribute to the number of independent replicates.

Scale of inference	Scale at which the factor of interest is applied	Number of replicates at the appropriate scale
Site type	Pooled sites	4 mountaintop,  3 valley

Example 4.

Wish to understand how vertebrate herbivory and nutrient addition affect the abundance of leaf mites on a species of plant. They create 4 nutrient addition plots and 4 control plots. Within each plot, they plant 6 plants, 3 of which are surrounded by a herbivore excluding cage, 3 of which have only a cage control. They sample 5 leaves per plant and count the number of mites per leaf. In this example, nutrients are applied at the scale of plot, herbivory exclusion is applied at the scale of plant. So the authors have 4 replicates of each nutrient level and 24 replicates of each herbivory level.

Scale of inference	Scale at which the factor of interest is applied	Number of replicates at the appropriate scale
Plant	Plot	4, 4
Plant	Plant	24, 24

If you have any questions, please direct all queries to the Editorial Office